Alcohol Intake and Cardiovascular
and Gastrointestinal Diseases
Do other factors matter?
Janne Schurmann Tolstrup
Copenhagen 2011
Alcohol Intake and Cardiovascular
and Gastrointestinal Diseases
Do other factors matter?
By
Janne Schurmann Tolstrup
Copenhagen, 2011
National Institute of Public Health
Faculty of Health Sciences
University of Southern Denmark
Denne afhandling er den 1. marts 2011 af Akademisk Råd, Det
Sundhedsvidenskabelige Fakultet, Syddansk Universitet antaget til forsvar for den medicinske doktorgrad.
Ole Skøtt
h.a.dec.
Forsvaret finder sted den 30. marts 2011 kl 15.00 i lokale
1.1.02, Statens Institut for Folkesundhed, Syddansk Universitet, Øster Farimagsgade 5a, 1353 København K.
This review is based on the following publications:
1.
Tolstrup J, Jensen MK, Tjønneland A, Overvad K, Mukamal KJ, Grønbæk M: Prospective study of alcohol
drinking patterns and coronary heart disease in
women and men. BMJ 2006;332:1244-1248.
2.
Tolstrup JS, Grønbæk M, Nordestgaard BG: Alcohol
Intake, Myocardial Infarction, Biochemical Risk Factors, and Alcohol Dehydrogenase Genotypes. Circulation: Cardiovascular Genetics 2009a;2:507-514.
3.
Hvidtfeldt UA, Tolstrup JS, Jakobsen MU, Heitmann
BL, Grønbæk M, O'Reilly E, Balter K, Goldbourt U,
Hallmans G, Knekt P, Liu S, Pereira M, Pietinen P,
Spiegelman D, Stevens J, Virtamo J, Willett WC, Rimm
EB, Ascherio A: Alcohol intake and risk of coronary
heart disease in younger, middle-aged, and older
adults. Circulation 2010;121:1589-1597.
4.
Tolstrup JS, Nordestgaard BG, Rasmussen S, TybjærgHansen A, Grønbæk M: Alcoholism and alcohol drinking habits predicted from alcohol dehydrogenase
genes. Pharmacogenomics J 2007;8:220-227.
5.
Tolstrup JS, Hansen JL, Grønbæk M, Vogel U, Tjonneland AM, Joensen A, Overvad K: Alcohol drinking habits, alcohol dehydrogenase genotypes and risk of
acute coronary syndrome. Scand J Pub Health 2010;
38:489-94
6.
Tolstrup JS, Jensen MK, Tjønneland A, Overvad K,
Grønbæk M: Drinking pattern and mortality in middle-aged men and women. Addiction 2004;99:323330.
7.
Mukamal KJ, Tolstrup JS, Friberg J, Jensen G,
Grønbæk M: Alcohol consumption and risk of atrial
fibrillation in men and women: the Copenhagen City
Heart Study. Circulation 2005b;112:1736-1742.
8.
Tolstrup JS, Grønbæk M, Tybjærg-Hansen A, Nordestgaard BG: Alcohol intake, alcohol dehydrogenase
genotypes, and liver damage and disease in the Danish general population. Am J Gastroenterol
2009b;104:2182-2188.
9.
Kristiansen L, Grønbæk M, Becker U, Tolstrup JS:
Risk of pancreatitis according to alcohol drinking habits: a population-based cohort study. Am J Epidemiol 2008;168:932-937.
10. Tolstrup JS, Kristiansen L, Becker U, Grønbæk M:
Smoking and risk of acute and chronic pancreatitis
among women and men: a population-based cohort
study. Arch Intern Med 2009c;169:603-609.
-2-
PREFACE
My interest in epidemiology started during my appointment as
a research assistant at the Institute of Preventive Medicine in
2001. Here, the stimulating scientific environment under the
direction of Thorkild IA Sørensen and Morten Grønbæk created
basis for my first epidemiological training and interest. Later,
at the National Institute of Public Health, my work became
more focused on studying health effects of alcohol, a work that
was carried out on the background of the alcohol epidemiological studies performed by especially Morten. I wish to thank
Morten for an increasing number of years of fruitful and inspiring collaboration.
While preparing the papers for this thesis, I have had the
opportunity to work with some excellent and creative persons
(Anne Tjønneland, Berit Heitmann, Kenneth Mukamal, Kim
Overvad, Majken Karoline Jensen, Søren Rasmussen, Thorkild IA
Sørensen, Ulrik Becker). For this, I am grateful.
I thank Louise Kristiansen and Ulla Arthur Hvidtfeldt,
talented former students of mine, who first-authored two of
the papers for this thesis.
I am most indebted to Børge Nordestgaard for his continuous
support in my scientific education. Børge is a role model in the
performance of structured, scientific work and he seems able
to glimpse the bigger picture of every complexity.
Ulrik Becker is thanked for commenting on this thesis and
for generously sharing his knowledge within alcohol and
gastroenterology.
I wrote a major part of this thesis at San Cataldo, in the
beautiful surroundings of the Amalfi coast. This can highly be
recommended to anyone who wishes for concentration and
peace, while being in the pleasant company of others.
Thanks to my colleagues at the National Institute of Public
Health for contributing to an open and stimulating workplace.
I owe thanks to participants as well as to initiators of the
cohort study included in this thesis, especially the Copenhagen
City Heart Study and the Diet, Cancer and Health Study, who
have generously made their data available for my work.
Finally, I wish to thank my family and friends for their
never ending support and continuous interest in my research.
-3-
CONTENTS
1. INTRODUCTION ........................................................................................ 5
Alcohol and biomarkers for liver damage .................................... 11
Alcohol and liver disease ................................................................. 11
ADH genotypes, biomarkers and liver disease ..................... 12
Acute and chronic pancreatitis .......................................................... 12
Alcohol, drinking pattern and pancreatitis ............................. 12
Smoking – an independent risk factor for pancreatitis? .. 13
2. CARDIOVASCULAR DISEASE ............................................................. 5
Coronary heart disease............................................................................ 5
Alcohol and cardiovascular biomarkers ..................................... 5
Genetic variation in alcohol dehydrogenase ............................ 6
The influence of sex .............................................................................. 7
Does the effect of alcohol on CHD depend upon age? .......... 8
The influence of drinking pattern .................................................. 8
Is wine more healthy than beer and spirits? ............................ 9
Atrial Fibrillation ......................................................................................10
Is the risk of AF according to alcohol
threshold-shaped? .........................................................................10
4. CONCLUSIONS .......................................................................................... 14
5. PERSPECTIVES ........................................................................................ 14
6. SUMMARY .................................................................................................. 14
7. DANSK RESUMÉ ...................................................................................... 15
3. GASTROINTESTINAL DISEASE .......................................................11
Liver cirrhosis ............................................................................................11
8. LIST OF CITED LITERATURE .......................................................... 16
ABBREVIATIONS
ADH alcoholdehydrogenase
ALDH acetaldehydedehydrogenase
AF
atrial fibrillation
ALT alanine aminotransferase
AST aspartate aminotransferase
CHD coronary heart disease
CI
confidence interval
-GT
-glutamyl transpeptidase
HDL high-density lipoprotein
IRD
incidence rate difference
LDL low-density lipoprotein
Note: 1 drink corresponds to 12 grams of pure alcohol
-4-
1. INTRODUCTION
impact on risk and thereby increase the possibility of primary
prevention against alcohol-related diseases in the future, the
aim is to review the data for differential effects of alcohol
according to modifying factors on these diseases.
Alcohol drinking has been part of human civilization for millenniums. It is used worldwide, and in most Western societies,
nondrinkers constitute the minority. The level of intake is
high, averaging 15 drinks per week for each adult European
man and women (WHO 2004). For many, the use of alcohol is
linked with pleasure and sociability, but its use potentially has
harmful consequences; alcohol contributes substantially to the
global burden of disease and 4% of all deaths are attributable
to alcohol, more so among men than among women (6.3% vs
1.1%) (Rehm et al. 2009a). While this ranks alcohol high on
the list of avoidable risk factors, alcohol in the context of public health is not as simple as is the case of for instance tobacco
smoking, where there seems to be no level of risk-free consumption and the risk of morbidity and mortality increases by
every cigarette. In contrast, the association between alcohol
and all-cause mortality is J-shaped; the nadir of the J reflects a
relatively lower risk of coronary heart disease (CHD) among
light to moderate drinkers compared with abstainers (Corrao
et al. 2000) and the ascending leg of the J is reflecting an increased risk of alcohol-related diseases such as liver cirrhosis,
pancreatitis, upper gastrointestinal cancers, cardiomyopathy,
and polyneuropathy among excessive alcohol users.
The evidence that alcohol intake is causally related to a decreased risk of CHD is substantial and comprises abundant
and consistent results from epidemiological studies in independent study populations, summarized elsewhere (Hill 2005;
Rehm et al. 2003; Klatsky 2003; Maclure 1993), and experimental evidence from clinical trials of the biological mechanism through changes in lipids and haemostatic factors (Rimm
et al. 1999). Further, none of the studies that have addressed
various kinds of bias have left concerns about a non-causal
explanation of the association between alcohol and CHD (Mukamal et al. 2001).
Hence, while the causality between alcohol and disease
appears evident, several factors may influence the strength of
alcohol’s effect, beneficial or harmful: the cardioprotective
effect of alcohol may not apply to everybody or all circumstances and not all individuals who drink heavily develop alcohol-related diseases. Specifically, the effect of alcohol may
depend on the drinkers’ age and sex, and on the pattern of
which the alcohol is ingested; steady or binge. Other lifestyle
habits, especially smoking, may be interacting with alcohol
and cause a risk that differs in magnitude from what is expected from the risk associated with the individual exposures.
Also, the risk of alcohol-related diseases may depend on certain genetic factors, especially genes that produce enzymes
affecting the rate of alcohol degradation.
For the harmful effects, little research has been conducted
into effects of modifying factors, maybe because the incidence
of these diseases is lower than for CHD and hence the statistical power to perform stratified analysis is more limited. Further, the evidence of an association between alcohol and outcome stems to a large degree from case-control and clinical
studies, and results obtained on the basis of general population cohorts are sparse.
While alcohol contributes to more than 60 types of disease
and injury (Gutjahr et al. 2001), the focus in this thesis is on
CHD, atrial fibrillation, liver cirrhosis and pancreatitis. With
the ambition to improve the understanding of the individual
2. CARDIOVASCULAR DISEASE
An estimated 16.7 million or approximately 30% of global
deaths result from various forms of cardiovascular disease. Of
these, 43% are due to CHD (Mackay et al. 2004). Unlike liver
cirrhosis and pancreatitis, the etiology of cardiovascular disease is multifactorial with several risk factors, including smoking, unhealthy diet, physical inactivity, hypertension and obesity. In contrast, moderate alcohol drinking is associated with a
reduced risk of CHD. On the other hand, consumption of alcohol in excessive amounts may be associated with increased
risk of atrial fibrillation.
Coronary heart disease
Relative risk of CHD
Consistently, a lower risk of CHD was observed in light and
moderate alcohol drinkers in the Diet, Cancer and Health
Study (Figure 1), the Copenhagen City Heart Study, and in each
of the eight study cohorts of the Pooling Project of Diet and
Coronary Disease (Hvidtfeldt et al. 2010; Tolstrup et al. 2009a;
Tolstrup et al. 2006). The maximal benefit was obtained at
around 1-2 drinks per day for women and 2-3 drinks per day
for men; for higher amounts of alcohol intake, there seemed to
be no further gain. Other studies have even reported an increase in risk at high intakes, indicating a J-shaped association
between alcohol intake and CHD (Corrao et al. 2000).
Figure 1. Alcohol intake is inversely associated with risk of CHD
Data from the Diet, Cancer and Health Study (from Tolstrup et al.
2006).
Alcohol and cardiovascular biomarkers
The beneficial effect of alcohol is primarily thought to be mediated through an increase in HDL and a decrease in fibrinogen (Rimm et al. 2007; Mukamal et al. 2005a; Rimm et al. 1999;
Langer et al. 1992; Criqui et al. 1987). For those biomarkers,
we observed that increasing alcohol intake was associated
-5-
Women
(n=4,105)
(n=5,479)
(n=4,105)
(n=5,479)
p<0.0001†
p<0.0001*
80
75
p<0.0001†
Systolic Blood Pressure
(mmHg)
Men
85
4.0
LDL Cholesterol
(mmol/L)
Women
3.8
3.6
3.4
p<0.0001†
ADH1C·2 produce enzymes with a 2.5-fold difference (Bosron
et al. 1986) (Figure 4).
Despite the large difference in in vitro activity of the
ADH1B·1 and ADH1B·2 enzymes, differences in alcohol degradation rate in human studies are modest and not even consistently observable (Neumark et al. 2004; Mizoi et al. 1994).
However, support for an in vivo effect comes from East Asian
studies that consistently report that ADH1B fast metabolizers
experience more unpleasant symptoms such as flushing when
drinking alcohol compared with ADH1B slow metabolizers. A
finding that probably can be attributed to the higher levels of
acetaldehyde that are produced in individuals with fast alcohol degradation (Figure 3) (Yokoyama et al. 2003; Carr et al.
2002; Takeshita et al. 2001; Takeshita et al. 1996). Further, we
found in 9,080 Caucasians that the ADH1B fast genotype was
associated with a lower alcohol intake, which further supports
a functional role of this variation (Tolstrup et al. 2007). For
ADH1C, no effect on alcohol degradation rate was found in a
human study (Whitfield 1994).
In Caucasians, frequencies of the most active alleles
(ADH1B·2 and ADH1C·1) are 2% and 58%, much different from
the allele distribution in East Asians where corresponding
frequencies are 70% and 93% (Tolstrup et al. 2007; Zintzaras
et al. 2006). Hence, most studies on ADH1B genotypes are
conducted in East Asians and most studies on ADH1C genotypes are conducted in Caucasians. The fast alleles are in linkage disequilibrium, meaning the majority of individuals who
are hetero- or homozygous for the fast ADH1B·2 will also be
hetero- or homozygous for the fast ADH1C·1 genotype.
It has been suggested that slow metabolizers may have
lower risks of CHD compared with fast metabolizers, because
they have alcohol in the blood for a longer period of time.
Findings in accordance with this hypothesis were obtained in
the Physicians’ Health Study, where moderate alcohol consumption was associated with a decreased risk. However, the
ADH1C genotype modified this association: as compared with
nondrinkers, the relative risk among ADH1C slow metabolizers who drank 8 or more drinks per week was much lower
than among the ADH1C intermediate and fast metabolizers in
the same alcohol category (p=0.01 for ADH1C-alcohol interaction) (Hines et al. 2001) (Table 1). In support of biological
plausibility of this result, the plasma HDL level was significantly higher among the slow metabolizers. However, results from
studies that followed, mainly of smaller size, were less convincing (Ebrahim et al. 2008; Heidrich et al. 2007; Younis et al.
2005; Djousse et al. 2005b) (Table 1). The ADH1B genotypes,
which are very unevenly distributed among Caucasians, were
p<0.0001*
135
130
125
120
p=0.23‡
p=0.006‡
1.8
1.6
1.4
1.2
p<0.0001*
3.0
p<0.0001*
2.1
Fibrinogen
(g/L)
HDL Cholesterol
(mmol/L)
3.2
2.4
140
2.0
p<0.0001*
Triglycerides
(mmol/L)
Diastolic Blood Pressure
(mmHg)
90
Men
1.8
1.5
p<0.0001*
p<0.0001*
2.8
2.6
2.4
2.2
2.0
1.2
28+
21-27
14-20
7-13
1-6
<1
35+
28-35
21-27
14-20
7-13
1-6
<1
28+
21-27
14-20
7-13
1-6
<1
35+
28-35
21-27
14-20
7-13
1-6
<1
Alcohol intake (drinks/week)
Figure 2. Alcohol intake and level of cardiovascular biomarkers
*P-value for linear trend, †P-value for trend for alcohol intake >21
drinks/week, ‡P-value for U-shape. Data from the Copenhagen City
Heart Study (modified from Tolstrup et al. 2009).
with increasing HDL cholesterol and decreasing fibrinogen
with no threshold, ie. even low amounts of alcohol intake were
associated with increasing HDL and decreasing fibrinogen
(Figure 2). These findings are consistent with the effects of
alcohol intake in the low range on the risk of CHD. We also
observed that alcohol intake at higher levels (>14/21 drinks
per week for women/men) was associated with higher systolic and diastolic blood pressure and with increased levels of
non-fasting triglycerides, which, on the other hand, are associated with increased cardiovascular risk (Nordestgaard et al.
2007; Lewington et al. 2002) and in support of a J-shaped
association between alcohol and risk of CHD.
Genetic variation in alcohol dehydrogenase
Alcohol is passively absorbed from the stomach and duodenum and is distributed within the body’s water compartment.
The rate of alcohol degradation is 30 to 90 minutes per drink
and it is influenced by sex, frequency of alcohol intake, age and
genetic factors (Ramchandani et al. 2001). Alcohol degradation
is mainly catalyzed by alcohol dehydrogenase (ADH), which is
rate limiting for the reaction (Figure 3). At least 7 human loci
encode alcohol degrading enzymes, of which class I (ADH1A,
ADH1B and ADH1C) is characterized by enzyme products with
high affinity for ethanol (subsequently referred to as alcohol),
while other ADHs mainly degrade other types of alcohols
(Osier et al. 2002). Functional genetic variation is found in
ADH1B and ADH1C: in vitro studies have shown that alleles
ADH1B·2 and ADH1B·1 produce enzymes with a 40-fold difference in alcohol degradation rate, and alleles ADH1C·1 and
Alcohol
ADH
Acetaldehyde
ALDH
ADH gene cluster (~380 kb)
4q
ADH4
ADH1 C/B/A
ADH5
ADH1C
Enzyme
·1
Frequency
0.58
Max velocity (U/min) 87
·2
0.42
35
ADH2
ADH1B
·1
0.98
9.2
ADH3
ADH1A
·2
0.02
400
Figure 4. The ADH gene cluster. The ADH class I genes (ADH1A,
ADH1B, ADH1C) have high affinity for ethonol. Genetic variation
affecting enzyme activity is found in ADH1B and ADH1C
(modified from Osier at al. 2002 and Bosron et al. 1988).
Acetate + H2O
Figure 3. Alcohol degradation
ADH; alcohol dehydrogenase, ALDH; acetaldehyde dehydrogenase.
-6-
N Alcohol
events dr/wk
Study
Tolstrup et al. 2010
770
Tolstrup et al. 2009a
628
Ebrahim et al. 2008
Heidrich et al. 2007
145
72
Younis et al. 2005*
220
Djoussé et al. 2005
Hines et al. 2001
132
395
1-6
7-20
21+
1-13
14+
3-23
1-8
8+
1-5
5+
>0
1-8
8+
ADH1C
1/1 (fast)
1 (reference)
0.88 (0.56-1.39)
0.97 (0.59-1.59)
0.99 (0.70-1.40)
0.80 (0.53-1.23)
7.9 (5.2-12)
0.56 (0.19-1.61)
1.06 (0.50-2.25)
0.70 (0.40-1.22)
0.57 (0.33-0.98)
0.74 (0.41-1.32)
1.11 (0.67-1.84)
0.62 (0.34-1.13)
1/2 (intermed.)
2/2 (slow)
1.38 (0.87-2.19) 1.10 (0.59-2.08)
0.97 (0.62-1.51) 0.91 (0.52-1.58)
0.73 (0.45-1.19) 0.84 (0.46-1.54)
0.98 (0.71-1.37) 0.83 (0.55-1.25)
0.82 (0.56-1.19) 0.88 (0.55-1.42)
6.2 (4.2-9.1)
11.3 (7.1-18)
0.83 (0.34-2.07)
0.36 (0.16-0.80)
0.56 (0.32-0.99) 0.66 (0.31-1.38)
0.77 (0.47-1.26) 0.68 (0.36-1.27)
0.60 (0.34-1.08) 0.53 (0.24-1.18)
0.66 (0.40-1.08) 1.02 (0.55-1.88)
0.68 (0.40-1.15) 0.14 (0.04-0.45)
Relative risk of CHD
not accounted for in any of these studies. We retested the
hypothesis in the Copenhagen City Heart Study; in this material, the power to detect an ADH1C-alcohol interaction of an
effect size similar to that observed in the Hines et al. study was
97%.
However, we found no difference in the risk of myocardial
infarction according to the ADH1C genotype among either light
to moderate or heavy drinkers (Figure 5, p=0.49 for interaction) (Tolstrup et al. 2009a). Adjusting results for the ADH1B
genotypes or, alternatively, excluding ADH1B fast metabolizers
from the analysis had no influence on results. The statistical
power to perform separate analyses for the ADH1B genotype
was insufficient. Similar results were obtained from the Diet,
Cancer and Health Study; there was no difference in the risk of
acute coronary syndrome according to ADH1C genotype in
strata of the amount or frequency of alcohol intake (Tolstrup et
al. 2010).
We also tested whether ADH1B and ADH1C genotypes
were independently associated with biomarkers of cardiovascular disease (systolic and diastolic blood pressure, HDL cholesterol, LDL cholesterol, triglycerides and fibrinogen) and
whether associations between alcohol and the biomarkers
were modified by genotypes, but found no evidence of either
of the hypotheses: ADH1B and ADH1C genotypes were not
consistently associated with any of the biomarkers among
men or women, and there was no sign of interaction between
genotypes and alcohol intake on any of the cardiovascular
biochemical risk factors (all P-values >0.05) (Tolstrup et al.
2009a).
Taken together, the lower risk of CHD among light to moderate drinkers does not seem to be modified by the ADH1C
genotype to any major extent. First, despite the quite strong
1/1 (fast)
2/1 (intermediate)
2/2 (slow)
1.0
0.5
0.25
<1
0.95
0.49
0.19
0.07
0.49
Table 1. Summary of previous findings of ADH1C,
alcohol and CHD
Presented numbers are relative risks (95% CI) compared with ADH1C·1/1 nondrinkers (for Ebrahim et al.
incidence rates are presented).
P represents P-value for the ADH1C-alcohol interaction.
*P-value for alcohol-ADH1C interaction was 0.02 after
post hoc regrouping alcohol into 1-2 and 2+ drinks per
week (the corresponding risk estimates are not
shown).
0.61
0.01
interaction observed by Hines et al., later studies have not
replicated this finding (Table 1). Second, other studies using
biomarkers, mainly HDL, as outcome, are not demonstrating
any modification between alcohol and ADH1C genotype on the
level of the respective biomarker either (Tolstrup et al. 2009a;
Ebrahim et al. 2008; Younis et al. 2005; Djousse et al. 2005a;
Whitfield et al. 2003). Third, in vivo effects of ADH1C variations
on the rate of alcohol degradation, demonstrating that individuals with ADH1C slow genotype are clearing alcohol more
slowly than individuals with the faster genotype, have been
difficult to assess in experimental studies (Whitfield et al.
2001; Whitfield 1994; Couzigou et al. 1991).
The power to detect interactions between genetic variations and environmental exposures increases steeply as the
relative risks associated with either exposure approaches
unity. The maximal effect of alcohol on the risk of CHD is about
0.7 (relative to nondrinkers), and the effect of the genotype in
the absence of alcohol is expected to be null (since ADH genes
have no known pleiotropic effects). In ours as well as the other
studies, the statistical power to detect ADH1C-alcohol interactions on the risk of CHD was most likely insufficient. Hence, it
cannot be excluded that the ADH1C genotypes play a minor
role for the association between alcohol intake and CHD; a
formal meta-analysis of the studies in Table 1 could maybe
answer this question.
In conclusion, the importance of the ADH1B and ADH1C
genotypes for the association between alcohol and the risk of
developing CHD, if any, is limited.
ADH1C genotype
2.0
P
1-13
14+
Alcohol intake (drinks/week)
Figure 5. Relative risk of CHD according to ADH1C genotype and
alcohol intake in men and women Data from the Copenhagen City
Heart Study (from Tolstrup et al, 2009a).
-7-
The influence of sex
For a given alcohol intake, blood alcohol levels will be higher
for women than for men because women have smaller body
sizes and relatively more body fat. This means that any dose of
alcohol will imply a higher effective dose for women than for
men, and women are, thus, generally more sensitive to any
alcohol-related effect. Further, women differ from men in
alcohol pharmacokinetics and in effects of alcohol on sex hormones (Register et al. 2002; Mumenthaler et al. 1999; Gavaler
et al. 1993; Colditz et al. 1987).
In line with this, a meta-analysis by Corrao and coworkers
indicate differences in the cardioprotective effect of alcohol
between men and women (Corrao et al. 2000): in women, the
measured risk of CHD decreased up to 6 drinks per week,
showed evidence of protective effects up to 18 drinks per
week, and reached statistical significance of harmful effects at
30 drinks per week. Conversely, the measured risk in men was
decreasing up to 15 drinks per week, showed evidence of
protective effects up to 51 drinks per week and reached statistical significance of harmful effects at 67 drinks per week.
Hence, both the protective and the detrimental effects of alcohol on CHD are achieved at lower levels in women than in
men.
between drinkers and nondrinkers (Abbott et al. 2002). Hence,
authors concluded that the association between alcohol intake
and CHD risk attenuated with age, which is somewhat in contrast to the findings from the Pooling Project on Diet and Coronary Disease even though the 75+ year-olds were not analyzed separately.
Age is associated with drinking pattern; for instance,
young people tend to binge drink more often than older individuals. Among Australians, the highest proportion of those
with occasional heavy drinking was found among the young
men and women (18-24 years old); with increasing age, this
proportion decreased continuously over the entire age span
(Livingston et al. 2009). This is similar to findings in the United
States, where frequent heavy drinking comprised 27% of 18to 29-year-old men and only 4% of 65+ year-olds (Hilton
1987). The Pooling Project on Diet and Coronary Disease did
not include drinking pattern, which is a limitation of the results. However, if the youngest participants binge drink more
often than older participants, their risk is probably conservatively estimated because binge drinking does not seem to be
associated with a decreased risk of CHD.
In conclusion, young adults, middle-aged and older individuals seem to have the same relative effect of alcohol intake
on CHD risk. So even though the etiology of CHD may differ
according to age, this indicates that the mechanism through
which alcohol seems to protect against CHD also applies for
young adults. However, younger adults are at low risk of CHD
and the absolute beneficial effect is small.
Does the effect of alcohol on CHD depend upon age?
CHD etiology in young adults may differ from that of older
individuals; for instance, relatively more cases of CHD among
younger adults can be attributed to genetic causes (Zdravkovic
et al. 2002; Marenberg et al. 1994). Hence, this age group may
not benefit from alcohol drinking as much as older individuals.
Since the incidence of CHD is very low in men younger than 40
years and in women younger than 50 years, the statistical
power to compare effects of alcohol in different age strata is
insufficient in most data sets. By using pooled data from eight
cohort studies from North America and Europe, the Pooling
Project on Diet and Coronary Disease, we tested if associations
between alcohol and CHD in young adults (39-50 years), middle-aged (50-60 years) and older individuals (60+ years) were
similar (Hvidtfeldt et al. 2010). Overall, associations between
alcohol intake and CHD risk seemed similar by age group
(Figure 6, left panel). This visual impression was confirmed by
a statistically nonsignificant test for interaction (P=0.25 for
men and 0.34 for women). In all age groups, a lower CHD
incidence was observed among moderately drinking men and
women than among nondrinkers. However, a smaller CHD
incidence rate difference (IRD) between nondrinkers and
moderate drinkers was observed among younger men
(IRD=45 per 100,000; 90% CI: 8-84) than among middle-aged
(64 per 100,000; 24-102) and older men (89 per 100,000; 44140) (Figure 6, right panel). Among women, similar results
were obtained.
The association between alcohol and CHD risk across age
groups is sparsely studied. In the Honolulu Heart Program,
including 18,456 men, authors observed a lower risk among
drinkers aged 45-74 years as compared with nondrinkers;
among older men (75+ years) there was no difference in risk
The influence of drinking pattern
In epidemiologic studies where alcohol exposure is summarized into a single measure of average amount, individuals
who drink small amounts on a number of weekdays are categorized with individuals who drink the same weekly amount
on a Saturday night. Evidence has emerged that these evidently dissimilar drinking patterns are associated with different
risks of morbidity and mortality.
Drinking pattern has been defined in various ways such as
drinking with meals, in weekends only, to intoxication, to a
<50 years
39-50
years
<50 years
39-50
years
500
1.50
Incidence of coronary heart disease per 100,000
1.00
0.75
50-60 years
Incidence per 100,000
1.50
1.00
0.75
0.50
60+ years
1.50
1.00
0.75
0.50
80
300
60
200
40
100
20
0
0
50-60 years
500
100
400
80
300
60
200
40
100
20
0
0
60+ years
500
100
400
80
300
60
200
40
100
20
0
0.
12
3. .9
017
.4
17
.4
+
35
+
0
0
0.
12.
9
3.
01
17 7.4
.5
-3
4.
9
7.
4
17
.4
+
3.
01
0
Alcohol intake (drinks/week)
Graphs by agegroup
.9
0.
12
35
+
.9
3.
017
.4
17
.5
-3
4.
9
0
0
0.
12
Relative risk of coronary heart disease
Hazard ratio
0.50
100
400
Alcohol intake (drinks/week)
Graphs by agegroup
-8-
Figure 6. Relative and absolute
risks of coronary heart disease in
men and women by age groups.
Black dots and bars represent
men, and white dots and bars
represent women.
Data from the Pooling Project on
Diet and Coronary Disease (modified from Hvidtfeldt et al. 2010).
certain blood alcohol level, more than a certain amount per
session (6 drinks, 13 drinks, ½ a bottle of spirits, etc.), and
amount and frequency have been combined. A common feature of these approaches is that alcohol exposure is described
in more than one dimension. Thus, in these studies the main
focus is not to compare nondrinkers and drinkers, but rather
to compare drinkers characterized by different drinking patterns.
In the Diet, Cancer and Health Study, we compared mortality risk among participants with different drinking frequencies. We found that the risk was higher among women drinking 7 drinks per week and among men drinking 14 drinks per
week if this amount was taken on one day of the week compared with distributing the same amount on more days of the
week (data not shown) (Tolstrup et al. 2004). In other studies,
results consistently imply an increased mortality risk of drinking large amounts of alcohol per session (Makela et al. 2005;
Laatikainen et al. 2003; Malyutina et al. 2002; Rehm et al.
2001; Trevisan et al. 2001).
Some of the excess mortality risk that is associated with a
binge-like drinking pattern seems to be due to an increased
risk of CHD. In the Male Health Professionals, Mukamal and
coworkers showed that the risk of CHD seemed to be more
strongly related to frequency than to amount of alcohol intake
among men (Mukamal et al. 2003). In the Diet, Cancer and
Health Study, we observed similar results for the men: relative
risks were generally lowest for the most frequent intake within similar categories of amount (Table 2). For example, among
men drinking on average 7-13 drinks per week, relative risks
of CHD were 0.89 (95% CI: 0.62-1.29) for drinking alcohol on
1 days per week, 0.81 (95% CI: 0.67-0.98) for 2-4 days a
week, and 0.66 (95% CI: 0.52-0.83) for 5-7 days per week (P
for trend = 0.0001). For the same category of drinking frequency, relative risks for increasing amount of alcohol intake
tended to be similar. Among women, the tendencies in results
were opposite; it seemed that a higher amount was associated
with the lowest risk within similar categories of drinking
frequency (Table 2) (Tolstrup et al. 2006). Generally, the impact of drinking pattern among women has been less studied
than among men, but a case-control study with comparable
measures of amount and frequency of alcohol intake also
found a tendency that amount was more strongly associated
with CHD risk than frequency among women (Mukamal et al.
Amount
(drinks/week)
Men
1-6
7-13
14-20
21+
P for trend
Women
1-6
7-13
14+
P for trend
1
Frequency (days/week)
2-4
5-7
2005a). The statistical power among women was lower than
among men, and results in women should, therefore, be taken
with caution.
Most studies trying to separate the effect of occasions of
heavy drinking from the effect of the average alcohol intake
found independently increased risk of CHD among both men
(Trevisan et al. 2004; Laatikainen et al. 2003; Murray et al.
2002; Malyutina et al. 2002) and women associated with binging (Dorn et al. 2007; Laatikainen et al. 2003; Murray et al.
2002; Hammar et al. 1997). Furthermore, in a meta-analysis,
the dose-response relation between alcohol intake and CHD
risk was significantly different in irregular and regular drinkers; in irregular drinkers who consumed alcohol for 2 days a
week or less often, a J-shaped curved was obtained, with an
increasing risk at intakes over 11 drinks per week. In contrast,
in regular drinkers who consumed alcohol on 3 days of the
week or more often, a protective effect was observed, even at
high amounts of alcohol intake (Bagnardi et al. 2008). This
result is based on pooled data from two studies only (Tolstrup
et al. 2006; McElduff et al. 1997) and should, thus, be taken
with caution. Furthermore, sex-specific effects were not accounted for. Nevertheless, should this finding represent causality, difference in drinking pattern by study population can
explain why some studies report an increased risk of CHD at
high intake of alcohol (a J-shaped curve), while others find no
increase in risk with increasing intake (sometimes referred to
as L-shaped).
In conclusion, drinking pattern has independent effects on
CHD risk. Data suggest that drinking frequency is inversely
and independently associated with risk of CHD, at least among
men. On the other hand, occasional heavy drinking (binge
drinking) is associated with increased CHD risk.
Is wine more healthy than beer and spirits?
It has been suggested that wine drinking has an especially
beneficial effect – an idea that originated from the observation
that the incidence of CHD in France was low despite a high
prevalence of smoking and fat intake, the so-called French
paradox (Criqui et al. 1994; Renaud et al. 1992). Biological
explanations for a more cardioprotective effect of wine compared with beer and spirits include the fact that substances in
wine (besides the ethanol) have been shown to inhibit platelet
aggregation and to slow down oxidation of low-density lipoprotein (Ramprasath et al. 2010; Pace-Asciak et al. 1996; Renaud et al. 1996; Pace-Asciak et al. 1995; Frankel et al. 1993).
In support of this appealing theory, it was found in the Copenhagen City Heart Study, that wine drinkers had a much lower
mortality than beer and spirit drinkers (Grønbæk et al. 1995).
However, quite some evidence indicates that the apparent
favorable effect of wine is an artifact resulting from characteristics of the drinker rather than of the drink itself. First, in an
overview of 10 cohort studies most of them did not demonstrate a more pronounced cardioprotective effect of red wine
compared to other beverages (Rimm et al. 1996). Differences
in results from country to country according to wine, beer and
spirits may be attributable to socio-economic or behavioral
characteristics of wine, beer and spirits drinkers. In line with
this, studies have shown that wine drinkers are more likely to
report optimal health, score higher in intelligent tests and to
eat a more healthy diet than beer and spirits drinkers (Johan-
P for
trend
1.00*
0.89
1.10
1.00
0.25
0.80
0.81
0.91
0.67
0.22
0.70
0.66
0.68
0.63
<0.0001
0.02
0.0001
0.001
<0.0001
1.00*
0.67
0.65
0.002
0.78
0.74
0.27
<0.0001
1.32
0.82
0.72
0.0003
0.57
0.12
0.01
Table 2. Relative risks of CHD as a function of amount and frequency of alcohol intake. *Reference group. Data from the Diet,
Cancer and Health Study (modified from Tolstrup et al. 2006).
-9-
sen et al. 2006; Barefoot et al. 2002; Mortensen et al. 2001;
Grønbæk et al. 1999; Tjønneland et al. 1999; Idler et al. 1997).
In conclusion, some studies support the clinical and experimental suggestions of an effect of wine in addition to a light
to moderate alcohol intake on cardiovascular disease, although the differences may be small and to a large extent may
be due to confounding.
Atrial Fibrillation
Anecdotal evidence proposes that acute high doses of alcohol
can trigger atrial fibrillation (AF) in otherwise cardiac-healthy
individuals, a syndrome known as holiday heart and first described by Ettinger, and later supported by others (Thornton
1984; Lowenstein et al. 1983; Ettinger et al. 1978). Further,
while AF in individuals younger than 45 years is rare, an alcoholic binge can cause this (Krishnamoorthy et al. 2009; Koul et
al. 2005). The relationship of habitual alcohol intake with risk
of incident AF is less certain.
Men
Relative risk of atrial fibrillation
Relative risk of atrial fibrillation
Is the risk of AF according to alcohol threshold-shaped?
In at least four case-controls studies, the risk of AF was significantly increased among the most heavily drinking patients,
while the risk associated with moderate drinking was comparable to abstention (Djousse et al. 2004; Ruigomez et al. 2002;
Koskinen et al. 1987). In the Copenhagen City Heart Study, we
studied prospectively the association between alcohol and
incident atrial fibrillation among 16,415 men and women; a
total of 1,071 AF cases occurred during follow-up (Mukamal et
al. 2005b). Among men, alcohol intake throughout the moderate range was not associated with risk of atrial fibrillation,
but consumption of 35 or more drinks per week was associated with a relative risk of 1.45 (95% CI: 1.02-2.04). No
increased risk was observed among moderately drinking
women (Figure 7). In accordance with this result, in the Women’s Health Study, the risk of AF was not increased among
women drinking up to 18 drinks per week; at higher amounts,
the relative risk was 1.60 (95% CI: 1.13-2.25).
The above results all indicate a threshold-shaped association between habitual alcohol intake and AF risk; no increased
risk at light to moderate intakes and an increased risk only at a
certain level. In the Copenhagen City Heart Study, to explore
this hypothesis further we used linear spines; the risk of AF
increased notably at a threshold of >35 drinks per week, with
a flat relationship at lower levels of intake.
In contrast to the above results, an analysis of the Diet,
Cancer, and Health cohort found 25% to 46% higher risks of
AF associated with average intake of just >12 drinks per week.
However, no increased risk was observed among women at
any intake (Frost et al. 2004). In two different cohorts of 65+
year-old men and women, no increase and inverse association,
respectively, were found in the risk of AF according to alcohol
intake (Mukamal et al. 2007; Psaty et al. 1997). In both these
populations, the level of alcohol use was low and binge drinking was uncommon; hence these results are not necessarily in
contrast to the idea of a threshold-shaped association between
alcohol and AF risk.
Heavier drinkers could sustain a higher risk of AF by different mechanisms. We assessed a series of potential mediators of the association of heavy alcohol intake with risk of AF,
including blood pressure, incident CHD during follow-up, and
incident congestive heart failure during follow-up. As expected, all three were strongly and independently associated
with risk of AF. However, adjustment for these factors had
little effect on the alcohol AF association. For example, the
relative risk of AF among men who consumed 35 or more
drinks per week was 1.45 in the basic model and 1.63 (95% CI:
1.15-2.31) with adjustment for all three potential mediators,
indicating that these factors do not mediate to any major extent the relation between alcohol and AF (Figure 7). Other
plausible explanations include the fact that chronic heavy
alcohol use itself has toxic effects on the pericardium, as supported by animal studies (Piano et al. 1999), or chronic heavy
drinking could have direct proarrhythmic effects; explanations
that can be further addressed in experimental studies. Third,
heavier drinkers are likely to have repeated exposure to episodic heavy drinking (ie. binge drinking), which could increase
the risk of triggering a single episode of AF. Difference in
drinking pattern could potentially explain the difference in
level of thresholds among the various studies; in study populations where the habit of heavy drinking occasions is common,
the observed threshold in amount of alcohol intake will be
lower than in study populations were binge drinking is rare.
We found no clear evidence that drinking frequency was independently related with risk (data not shown). We did not have
information on other dimensions of drinking pattern to perform further tests.
In conclusion, most studies have shown that moderate alcohol drinking is unrelated with risk of AF among men and
women, while more heavy consumption is associated with a
higher AF risk. This relationship does not appear to be related
to the adverse effects of heavy drinking on CHD or blood pressure.
2.0
1.0
0.5
<1
1-6
7-13
14-20
21-27
28-34
35+
Women
2.0
1.0
0.5
<1
Alcohol intake (drinks/week)
1-6
7-13
14-20
21+
Alcohol intake (drinks/week)
Figure 7. Alcohol and risk of atrial fibrillation among men and women. White dots represent multivariate adjusted relative risks; black dots
represent relative risks that are further adjusted for potential mediators (use of blood pressure medication, systolic blood pressure and incident diagnoses of CHD and congestive heart failure). Data from the Copenhagen City Heart Study (modified from Mukamal et al. 2005).
- 10 -
3. GASTROINTESTINAL DISEASE
Liver cirrhosis and pancreatitis are generally considered to be
the major alcohol-related diseases of the gastrointestinal tract.
The impact of alcohol on cirrhosis morbidity and mortality has
been scientifically recognized for decades. In populations
where hepatitis infections are uncommon, ~80% of cirrhosis
cases are attributable to excessive alcohol consumption, constituting 16% of deaths attributable to alcohol worldwide
(Rehm et al. 2009a).
As early as 1878, alcohol was proposed as a risk factor for
pancreatitis (Friedrich 1878), which has been confirmed in
clinical series (Voirol et al. 1980; Sarles et al. 1979; Edlund et
al. 1968), and it is well known that a large fraction of pancreatitis cases is caused by alcohol. However, epidemiologic studies on the quantitative aspect of the association between
alcohol intake and pancreatitis have until recently been
sparse, which is why the impact of alcohol induced pancreatitis on global morbidity and mortality has not been estimated
(Rehm et al. 2009a; Rehm et al. 2009b).
Liver cirrhosis
Excessive alcohol intake associates with biochemical signs of
liver damage and with liver disease, which has mostly been
demonstrated in case-control studies (Corrao et al. 2004).
Whether this is also the case for low to moderate levels of
alcohol intake in individuals in the general population is less
certain. Liver cirrhosis may take several years to develop;
intermediate states from pathologically healthy organ to manifest disease exist in all ranges, but are often difficult to study
in the general population because clinical symptoms have not
yet arisen. Biomarkers for liver damage are appealing alternatives, since they are immediate and sensitive measures, objectively evaluating the current state of the organ.
Alcohol and biomarkers for liver damage
Extreme levels of alanine aminotransferase (ALT), albumin,
alkaline phosphatase, bilirubin, coagulation factors II, VII and
X, -glutamyl transpeptidase ( -GT) and mean erythrocyte
volume are generally indicative of acute liver damage and are
used routinely for clinical purposes (Burtis et al. 2006; Marks
et al. 2002). Among participants of the Copenhagen City Heart
Study 1991-94 and 2001-03, these biomarkers were obtained
as part of the clinical examination. By combining the biomarker levels with questionnaire information of the participants’
self-reported alcohol intake, we found that increasing alcohol
intake associated with increasing levels of ALT, GT, albumin,
bilirubin, coagulation factors and erythrocyte volume, and
with decreasing levels of alkaline phosphatase (Tolstrup et al.
2009b) (Figure 8). As illustrated, the difference in levels of the
various biomarkers between low, moderate, and excessive
alcohol intake were mostly small and probably not indicative
of subclinical liver damage; however, mechanistically it is
nevertheless an interesting finding that biomarkers were
influenced at low intakes of alcohol with no apparent threshold effect. These biochemical tests have not previously been
correlated with usual alcohol intake in a large population
sample.
It was somewhat surprising that alcohol associated with
increased levels of coagulation factors II+VII+X, equivalent to
decreased prothrombin time. Chronic damage of hepatocytes
Figure 8. Alcohol intake and biomarkers for liver damage
P-values are for linear trend. Data from the Copenhagen City Heart
Study (from Tolstrup et al. 2009b).
by alcohol usually leads to reduced capacity to produce proteins, as seen in alcoholic liver cirrhosis, while in less severe
forms of alcoholic liver disease, prothrombin time and, thus,
levels of coagulation factors II+VII+X may not be affected
(Burtis et al. 2006; Marks et al. 2002). However, our observation agrees with previous epidemiological data which demonstrate that heavy alcohol consumption leads to a more procoagulant state with higher levels of coagulation factor VII (Lee et
al. 2003). Alternatively, since albumin also increased in response to increasing alcohol intake, the mechanism behind the
increased levels of coagulation factors may be yet unknown.
The increased levels of ALT, -GT, mean erythrocyte volume, and bilirubin in individuals with the highest alcohol
intake observed in the present study are in accordance with
those expected, that is, increased alcohol intake associates
with increased biochemical signs of liver cell damage.
In conclusion, we observed that increasing alcohol intake
from none through moderate and excessive intake leads to
stepwise increasing signs of liver damage with no threshold
effect. The mechanism behind some of these findings is not yet
understood.
Alcohol and liver disease
As expected, increasing amount of alcohol intake was strongly
associated with increasing risk of alcoholic liver cirrhosis and
with a risk of alcoholic liver disease overall. The highest risk
was observed among the most heavily drinking participants:
as compared with nondrinkers, the risk of alcoholic liver cirrhosis was 13 (95% CI: 4.6-37) and the risk of alcoholic liver
disease overall was 4.1 (95% CI: 2.5-7.0) for drinking more
than 28 drinks per week (both P’s for trend<0.0001).
- 11 -
Alcohol, drinking pattern and pancreatitis
In case-control and clinical studies, patients with pancreatitis
are found to have a higher alcohol intake (Yadav et al. 2009;
Lin et al. 2001; Talamini et al. 1999; Yen et al. 1982; Durbec et
al. 1978). Knowledge of the quantitative association between
alcohol drinking habits in the general population is sparse. To
asses the association between amount, type and frequency of
alcohol intake and risk of pancreatitis, we used data on 17,905
men and women in the Copenhagen City Heart Study. A high
alcohol intake was similarly associated with a risk of both
acute and chronic pancreatitis for men and women (Kristiansen et al. 2008). The risk of total pancreatitis increased with
increasing alcohol intake, but was, however, only statistically
significant above 35 drinks per week (Figure 10). To test for a
threshold-like relationship, we remodeled the association
between alcohol and pancreatitis using fractional polynomials;
however, results did not indicate a threshold. The curves flattened out at high intakes, but the test for linearity did not
provide evidence for departure. The relative risk of total
pancreatitis increased with 1.13 for every additional drink per
day (95% CI: 1.06-1.21). In a recent metaanalysis, the corres8
Relative risk of pancreatitis
ADH genotypes, biomarkers and liver disease
Production of toxic acetaldehyde within cells is thought to be
one of the causes of liver damage and disease in response to
alcohol intake (Witt et al. 2007; Adachi et al. 2005; Lieber
2004). The ADH1B and ADH1C genotypes could lead to different intracellular acetaldehyde concentrations in response to
alcohol intake, and thus to differences in damage and disease
of the liver (Zintzaras et al. 2006; Cichoz-Lach et al. 2006; Sun
et al. 2005; Verlaan et al. 2004; Matsumoto et al. 1996). We
tested this hypothesis in the Copenhagen City Heart Study.
ADH1B and ADH1C genotypes were not associated with
any of the biomarkers for liver damage (ALT, albumin, alkaline
phosphatase, bilirubin, coagulation factors II, VII and X, -GT
and mean erythrocyte volume), nor did ADH1B and ADH1C
genotypes interact with alcohol on the level of any of these
biomarkers (all P-values >0.05) (data not shown). Further,
ADH1B and ADH1C genotypes were not associated with a risk
of alcoholic liver disease, nor was the association between
alcohol intake and liver disease modified by ADH1C genotype
(Figure 9); we did not have enough statistical power to stratify
similarly for ADH1B genotype.
This study and most other studies (Zintzaras et al. 2006)
have not been able to demonstrate any effect of ADH1B and
ADH1C genotypes on the risk of alcoholic liver disease. Also, in
support of this finding, we could not even demonstrate an
influence of these genotypes on biomarkers of liver damage in
individuals in the general population.
In conclusion, ADH1B and ADH1C genotypes are not associated with alcoholic liver damage or disease.
Acute and chronic pancreatitis
The distinction between acute and chronic pancreatitis is not
clear and disagreement as to whether they really are distinct
diseases or whether repeated attacks of acute pancreatitis
eventually lead to chronic pancreatitis exists (Apte et al. 2005;
DiMagno et al. 2005). The most common causes of pancreatitis
are thought to be gallstones and excessive alcohol intake.
Relative risk of alcoholic liver disease
6
5
4
3
2
1
ADH1C 1/1
ADH1C 1/2
ADH1C 2/2
*
0
<1
1-13
14-20
21-27
28+
Alcohol intake (drinks per week)
Figure 9. Alcohol intake and risk of alcoholic liver disease by ADH1C
genotype. *Reference group. Data from the Copenhagen City Heart
Study (from Tolstrup et al. 2009).
4
2
1
0.5
0
1-6
7-13
14-20
21-34
35-48
48+
Alcohol intake (drinks/week)
Figure 10. Alcohol and risk of pancreatitis
Data from the Copenhagen City Heart Study (modified from Kristiansen et al. 2009)
ponding relative risk was found to be 1.2 (95% CI: 1.1-1.2)
(Irving et al. 2009).
Drinking frequency did not seem to be independently associated with pancreatitis (data not shown). For analyses of
the type of alcohol (beer, wine, and spirits) and risk of pancreatitis, we found an increased risk especially associated with
a high beer intake. This result could hypothetically be due to a
protective effect of the antioxidants in wine. However, the
validity of these results is severely limited by the low statistical power to perform this analysis.
Limitations include the fact that the diagnosis of acute and
chronic pancreatitis from the Danish Hospital Discharge Register has not been validated; misclassification between the two
diagnoses may occur, as the symptoms and diagnostic criteria
are overlapping and the two diseases can co-exist (Kloppel et
al. 1995). Such misclassification would result in similarly
observed associations between alcohol and risk of acute and
chronic pancreatitis, as it appeared to be the case for the obtained results.
In conclusion, alcohol in the general population is associated with a risk of pancreatitis. Drinking pattern did not
- 12 -
Smoking – an independent risk factor for pancreatitis?
In the medical literature, smoking is generally not mentioned
as a preventable cause of pancreatitis, even though quite a
few, mostly case-control, studies suggest this (Law et al. 2010;
Yadav et al. 2009; Lindkvist et al. 2008; Morton et al. 2004; Lin
et al. 2000; Talamini et al. 1999; Talamini et al. 1996; Bourliere
et al. 1991; Lowenfels et al. 1987; Yen et al. 1982). Also, evidence from experimental studies suggest that smoking is
associated with pancreas damage (Wittel et al. 2008; Wittel et
al. 2006; Doi et al. 1995; Chowdhury et al. 1995). In most populations, smoking is strongly associated with alcohol drinking,
and an independent effect of smoking can be difficult to assess,
especially from a case-control design. We aimed at assessing
the independent effects of smoking on the risk of pancreatitis
in the general population.
In the Copenhagen City Heart Study, we observed doseresponse associations between smoking and risk of acute and
chronic pancreatitis in men and women (Tolstrup et al. 2009c).
For total pancreatitis, the relative risk as compared with never
smokers was 2.5 (95% CI: 1.5-3.9) and 3.3 (95% CI: 1.9-5.8)
for men and women who smoked 15-24 and 25 grams of
tobacco per day (Figure 11). For former smokers, we found a
statistically significantly increased risk of pancreatitis (1.7,
95% CI: 1.0-2.7). This relatively high risk in the ex-smokers
did not seem to be due to sick-quitters because omitting the
first years of follow-up had only limited effect on the size of
the hazard ratio. Lung function within smoking categories
ranged in the expected order (decreasing in the order of never-smokers, ex-smokers, and current smokers of 1-14, 15-24
and >25 g/day, respectively). Also, the amount of carbon monoxide in the expired air was similar among never and exsmokers, which indicates that there was not a substantial
fraction of current smokers that was misclassified as exsmokers. We did not have sufficient statistical power to further explore the risk among the ex-smokers.
We performed additional analyses to explore if the risk of
pancreatitis associated with smoking was independent of
alcohol: first, the association between smoking and pancreatitis was confirmed among consistent light-to-moderate alcohol
Relative risk of pancreatitis
8
4
2
1
0.5
Never
Former
1-14 g/day
15-24 g/day
drinkers (participants who in each of the examinations of the
Copenhagen City Heart Study reported to drink maximally 14
drinks per week for women and 21 drinks per week for men),
and second, the association between smoking and the risk of
pancreatitis was stratified according to alcohol intake at the
examination closest to the pancreatitis diagnosis (Figure 12).
7
Relative risk of pancreatitis
seem to be associated with independent risks, but this calls for
further research on the subject.
Never smokers
6
Ex -smokers
Current smokers
5
4
3
2
1
*
0
<7
7 -20
>20
Alcohol intake (drinks/week)
Figure 12. Alcohol intake, smoking status and risk of pancreatitis
*Reference group. Data from the Copenhagen City Heart Study
(from Tolstrup et al. 2009).
An increased risk of pancreatitis with increasing amount of
smoking was consistently observed in both analyses. There
was no evidence of interaction between alcohol and smoking
on the risk of pancreatitis (P =0.70).
Approximately 46% of cases of pancreatitis were attributable to smoking in this cohort with a high proportion of current and former smokers (only 38% were never smokers).
During the last decade, the proportion of smokers in the Danish population, as in most Western populations, has decreased, so the etiologic fraction of pancreatitis according to
smoking has probably decreased as well. However, given the
quite high relative risks of pancreatitis associated with smoking, the fraction of cases attributable to smoking will most
likely still be considerable.
Smoking may also be associated with gallstone disease
(Jørgensen et al. 1990; Jørgensen 1989). If so, observed associations between smoking and pancreatitis could be explained
by an increased risk of gallstones in individuals who smoke,
which, in turn, renders these individuals at high risk of pancreatitis. To test this hypothesis, we performed analyses including gall stones as a time dependent variable. As expected,
gallstones were strongly associated with pancreatitis risk (11,
95% CI: 7.7-15); however, including gallstones had very little
influence on the relative risk of pancreatitis associated with
smoking, indicating that gallstone disease does not mediate
the association between smoking and pancreatitis.
In conclusion, smoking is associated with an increased risk
of pancreatitis and in both men and women. The risk associated with smoking is independent of alcohol and of gallstone
disease; two risk factors suggested as being the main causes of
pancreatitis.
25+ g/day
Smoking status
Figure 11. Smoking and risk of pancreatitis
Data from the Copenhagen City Heart Study (modified from Tolstrup
et al. 2009c).
- 13 -
4. CONCLUSIONS
The low risk of coronary heart disease (CHD) in light to moderate alcohol drinkers was consistently observed in men and
women, was independent of genetic variation in alcohol dehydrogenase genotypes ADH1B and ADH1C, and was of similar
relative size in young adults, middle-aged and older individuals. However, since younger adults are at low risk of CHD, the
absolute beneficial effect was small in the young. In contrast,
drinking pattern had independent effects on CHD risk: a frequent intake appeared to be the most important factor for the
cadioprotective effect of alcohol, whereas occasional heavy
drinking was associated with increased CHD risk.
The risk of atrial fibrillation according to alcohol intake
was consistent with a threshold-like relationship; for light to
moderate alcohol intake there was no increase in the risk,
while more heavy consumption was associated with a higher
risk. This relationship did not appear to be related to the adverse effects of heavy drinking on CHD or blood pressure.
The risk of liver damage and disease was strongly associated with the amount of alcohol intake; we observed stepwise increasing signs of liver damage with increasing alcohol
intake, from none through moderate and excessive intake with
no apparent threshold effect. The risk of liver damage and
disease was not associated with ADH1B and ADH1C genotypes.
The risk of pancreatitis was associated with alcohol in the
general population, a relation that seemed independent of the
alcohol drinking pattern.
Smoking was associated with the risk of developing pancreatitis and this effect was independent of alcohol intake and
gallstone disease. The risk of pancreatitis was increased approximately 3-fold among smokers (as compared with never
smokers), which was similar to the risk associated with heavy
drinking (relative to nondrinking).
5. PERSPECTIVES
For public health purposes, the effects of alcohol on diseases
and conditions should not be considered in isolation. However, the emerging and consistent evidence that the beneficial
effects of alcohol are not attained by episodic binge drinking
and that the risk of coronary heart disease increases among
individuals with a binge-like drinking pattern is of considerable public health relevance. It remains to be tested how the
risk of atrial fibrillation and pancreatitis according to alcohol
depends upon the drinking pattern.
Young adults (39-50 years) seem to have the same relative
benefit of a low to moderate alcohol intake according to CHD
risk as middle-aged and older individuals. Caution should be
taken when interpreting this finding in a public health setting,
because other alcohol-related diseases and conditions in this
age group were not accounted for. For public health purposes,
the association between alcohol and all-cause mortality is
essential because it reflects net loss of life attributable to alcohol consumption and, thus, constitutes the scientific basis for
creating guidelines on sensible drinking. For even younger
individuals, the risk of coronary heart disease is practically
zero and beneficial effects of alcohol are thus negligible. Further, binge drinking is more prevalent among young people
than among older individuals, and hazardous implications of a
binge-like drinking pattern, such as an increased risk of traffic
accidents and injuries, most likely predominate among the
young.
Smoking is a strong risk factor for the development of
pancreatitis and should be recognized as such. For instance,
pancreatitis should be included in overall assessments of the
harmful consequences of smoking on morbidity and mortality
and of the economic consequences to society. In future work,
the effect of changes of smoking behavior should be studied.
Specifically, the effect of the time since quitting smoking is
interesting to study when the excess risk associated with
smoking declines.
Obviously, alcohol has effects on numerous other diseases
and conditions, and many other factors are of importance
besides those considered in this thesis. For instance, individuals with a family history of alcoholism may be better off abstaining completely from drinking alcohol. In an optimal setting, the balance between alcohol’s beneficial and harmful
effects should be achieved. The most important factors in this
equation may very well be to keep the amount of alcohol intake in the light to moderate range and to avoid binge drinking.
6. SUMMARY
Alcohol is used all over the world and in most Western societies, the average intake is high. Alcohol is associated with more
than 60 diseases and globally, 4% of all deaths are attributable
to alcohol.
The aim of the present thesis is to study associations between alcohol intake and risk of coronary heart disease (CHD),
atrial fibrillation, liver cirrhosis and pancreatitis, and more
specifically, to review the data for differential effects of alcohol
according to modifying factors on these diseases.
The thesis is based on the results from 10 epidemiological
studies, conducted in the Copenhagen City Heart Study, the
Diet, Cancer and Health Study and the Pooling Project of Diet
and Coronary Disease. In all study cohorts, a lower risk of CHD
was observed in light and moderate alcohol drinkers. In the
Copenhagen City Heart Study, we also found that increasing
alcohol intake was associated with increasing HDL cholesterol
and non-fasting triglycerides, higher systolic and diastolic
blood pressure and decreasing fibrinogen. In contrast, ADH1B
and ADH1C genotypes were not associated with risk of CHD or
with any of the cardiovascular biomarkers, and there was no
indication that associations between alcohol intake and CHD
and between alcohol intake and biomarkers were modified by
genotypes. The finding that ADH genotypes are not modifying
the association between alcohol and CHD was confirmed in the
Diet, Cancer and Health Study.
In the Pooling Project of Diet and Coronary Disease, we
found that the association between alcohol and relative risk of
CHD was similar in young adults (39-50 years), middle-aged
(50-60 years) and older individuals (60+ years). However,
since the incidence of CHD is low in young adults, the incidence rate difference between nondrinkers and moderate
drinkers was much smaller in young adults than in older individuals, hence, for young adults, the absolute beneficial effect
of alcohol is small.
Alcohol has differential effects on the risk of mortality and
CHD according to drinking pattern. In the Diet, Cancer and
Health Study, we found that for the same weekly amount of
alcohol intake, a non-frequent intake implied a higher risk of
- 14 -
death than a frequent one. For CHD, drinking frequency may
be the primary determinant of the inverse association between alcohol intake and CHD: The risk of CHD was lower
among men who drank alcohol on more days of the week,
compared to men who drank alcohol on fewer days of the
week, independent of the total weekly amount of alcohol intake.
The risk of atrial fibrillation according to alcohol intake
seems to be threshold-shaped; In the Copenhagen City Heart
Study, no increased risk was observed among light to moderate drinkers and an increased risk only among individuals
drinking >35 drinks per week.
Also in the Copenhagen City Heart Study, the risk of liver
cirrhosis was strongly associated with increasing alcohol
intake. Further, we found that increasing alcohol intake associated with biomarkers for liver damage (alanine aminotransferase, albumin, alkaline phosphatase, bilirubin, coagulation
factors II, VII and X, -glutamyl transpeptidase and mean erythrocyte volume). In contrast, ADH1B and ADH1C genotypes
were not associated with a risk of liver cirrhosis or with any of
these biomarkers, and there was no indication that associations between alcohol intake and liver cirrhosis and between
alcohol intake and biomarkers were modified by genotypes.
Finally, we observed that alcohol intake was associated
with an increased risk of pancreatitis. Smoking was also associated with an increased risk of pancreatitis, which was independent of alcohol and of gallstone disease; two risk factors
suggested as being the main causes of pancreatitis.
In conclusion, the results show that the association between alcohol and CHD is independent of genetic variation in
alcohol-degrading enzymes. The alcohol drinking pattern is
independently associated with risk of coronary heart disease
and all-cause mortality. The beneficial effect of alcohol on CHD
is observed among both young adults, middle-aged and elderly, but the magnitude of the absolute beneficial effect is least
among the young adults. Both alcohol and smoking are associated with increased risk of pancreatitis. These results are
important for future studies of the biological effects of alcohol
on health and for public guidelines on alcohol.
7. DANSK RESUMÉ
Alkohol er forbundet med risikoen for mere end 60 sygdomme
og cirka 4 % af alle dødsfald kan tilskrives alkohol. Alkoholindtaget er højt i de fleste vestlige lande.
Formålet med nærværende afhandling er at undersøge
sammenhænge mellem alkoholforbrug og risiko for koronar
hjertesygdom, atrieflimmer, levercirrose og pankreatitis, og
mere specifikt at gennemgå evidensen for at alkohol har forskellige effekter på risikoen for disse sygdomme i forhold til
udvalgte baggrundsfaktorer.
Afhandlingen er baseret på 10 videnskabelige artikler, som
tager udgangspunkt i Østerbroundersøgelsen, Kost, Kræft og
Helbred og the Pooling Project of Diet and Coronary Disease.
Sidstnævnte er en samling af data fra 8 forskellige kohortestudier fra hele verden.
I samtlige datamaterialer observerede vi en lavere risiko
for koronar hjertesygdom blandt personer med et let til moderat alkoholforbrug sammenlignet med afholdende. I Østerbroundersøgelsen fandt vi desuden, at stigende alkoholforbrug
var forbundet med stigende HDL-kolesterol og ikke-fastende
triglycerider, højere systolisk og diastolisk blodtryk og faldende fibrinogenniveau.
Alkohol nedbrydes i leveren af alkoholdehydrogenase
(ADH), som er en klasse af isoenzymer. Der findes genetisk
variation i ADH1B og ADH1C med forskellig alkoholnedbrydningshastighed til følge. Vi undersøgte om sammenhængen
mellem alkohol og koronar hjertesygdom afhang af disse genotyper, men det var ikke tilfældet, ligesom sammenhænge mellem alkohol og kardiovaskulære biomarkører heller ikke afhang af ADH1B og ADH1C genotyperne.
Vi undersøgte sammenhængen mellem alkohol og koronar
hjertesygdom blandt unge voksne (39-50 år), midaldrende
(50-60 år) og ældre individer (60+ år). De relative risici var af
samme størrelse i de tre aldersgrupper, men eftersom incidensen af koronar hjertesygdom er lav blandt unge voksne,
var incidensrate-differenserne mellem moderat drikkende og
afholdende meget mindre hos unge voksne end hos ældre
aldersgrupper. Den absolutte gavnlige effekt af alkohol blandt
unge voksne er derfor lille.
Vi sammenlignede også risikoen for at dø blandt personer
med det samme ugentlige alkoholforbrug men med forskellig
drikkefrekvens. Risikoen for at dø var større blandt kvinder
med et ugentligt forbrug på over 7 genstande og blandt mænd
med et ugentligt forbrug på over 14 genstande som indtog al
alkoholen på én dag sammenlignet med mænd og kvinder med
det samme ugentlige forbrug spredt udover flere af ugens
dage. For koronar hjertesygdom syntes drikkefrekvens at
være vigtig for den inverse sammenhæng mellem alkoholforbrug og koronar hjertesygdom: Risikoen var lavere blandt
mænd, der drak alkohol på flere af ugens dage sammenlignet
med mænd, der drak alkohol på færre af ugens dage. Dette
fund var uafhængigt af det samlede ugentlige alkoholforbrug.
Risikoen for atrieflimren var øget blandt personer, som
drak mere end 35 genstande om ugen. For lavere alkoholforbrug var risikoen for atrieflimmer ikke øget sammenlignet
med risikoen blandt afholdende.
Risikoen for levercirrose var stærkt forbundet med stigende alkoholforbrug. Desuden fandt vi at stigende alkoholforbrug er forbundet med biomarkører for leverskader (alanin
aminotransferase, albumin, alkalisk fosfatase, bilirubin, koagulationsfaktorer II, VII og X, -glutamyl transpeptidase og
erytrocytvolumen). Disse sammenhænge var uafhængige af
ADH1B og ADH1C genotyperne.
Endelig observerede vi, at alkohol var forbundet med en
øget risiko for pankreatitis. Rygning var også forbundet med
en øget risiko for pankreatitis, hvilket var uafhængigt af alkohol og af galdestenssygdom, som er de to risikofaktorer, som
anses for at være de væsentligste årsager til pankreatitis.
I sammenfatning viser resultaterne, at sammenhængen
mellem alkohol og koronar hjertesygdom er uafhængig af
genetisk variation i alkoholnedbrydende enzymer. Alkoholdrikkemønstre har, uafhængigt af det samlede alkoholforbrug,
indflydelse på risiko for koronar hjertesygdom og død. Den
gavnlige effekt af alkohol på hjertesygdom observeres både
blandt unge voksne, midaldrende og ældre, men størrelsen af
den absolutte gavnlige effekt er mindst blandt de yngste. Både
alkohol og rygning er forbundet med øget risiko for pankreatitis. Disse resultater har betydning for såvel fremtidige studier
af biologiske effekter af alkohols betydning for helbred som
for folkesundhedsmæssige budskaber om alkohol.
- 15 -
8. LIST OF CITED LITERATURE
Criqui MH, Cowan LD, Tyroler HA, Bangdiwala S, Heiss G, Wallace RB,
Cohn R: Lipoproteins as mediators for the effects of alcohol consumption and cigarette smoking on cardiovascular mortality: results form
the Lipid Research Clinics Follow-up Study. Am J Epidemiol
1987;126:629-637.
Abbott RD, Curb JD, Rodriguez BL, Masaki KH, Yano K, Schatz IJ, Ross
GW, Petrovitch H: Age-related changes in risk factor effects on the
incidence of coronary heart disease. Ann Epidemiol 2002;12:173-181.
Adachi M & Brenner DA: Clinical syndromes of alcoholic liver disease.
Dig Dis 2005;23:255-263.
Criqui MH & Ringel BL: Does diet or alcohol explain the French paradox? Lancet 1994;344:1719-1723.
Apte MV, Pirola RC, Wilson JS: Molecular mechanisms of alcoholic
pancreatitis. Dig Dis 2005;23:232-240.
DiMagno MJ & Dimagno EP: Chronic pancreatitis. Curr Opin Gastroenterol 2005;21:544-554.
Bagnardi V, Zatonski W, Scotti L, La Vecchia C, Corrao G: Does drinking
pattern modify the effect of alcohol on the risk of coronary heart
disease? Evidence from a meta-analysis. J Epidemiol Community Health
2008;62:615-619.
Barefoot JC, Gronbaek M, Feaganes JR, McPherson RS, Williams RB,
Siegler IC: Alcoholic beverage preference, diet, and health habits in the
UNC Alumni Heart Study. Am J Clin Nutr 2002;76:466-472.
Bosron WF & Li TK: Genetic polymorphism of human liver alcohol and
aldehyde dehydrogenases, and their relationship to alcohol metabolism and alcoholism. Hepatology 1986;6:502-510.
Burtis, CA, Ashwood, ER, and Bruns, DE. Tietz Textbook of Clinical
Chemistry and Molecular Diagnosis. St. Louis: Elsevier Saunders, 2006.
Carr LG, Foroud T, Stewart T, Castelluccio P, Edenberg HJ, Li TK: Influence of ADH1B polymorphism on alcohol use and its subjective effects
in a Jewish population. Am J Med Genet 2002;112:138-143.
Cichoz-Lach H, Partycka J, Nesina I, Celinski K, Slomka M, Wojcierowski J: Genetic polymorphism of alcohol dehydrogenase 3 in alcohol liver
cirrhosis and in alcohol chronic pancreatitis. Alcohol Alcohol
2006;41:14-17.
Corrao G, Rubbiati L, Bagnardi V, Zambon A, Poikolainen K: Alcohol
and coronary heart disease: a meta-analysis. Addiction 2000;95:15051523.
Couzigou P, Fleury B, Groppi A, Iron A, Coutelle C, Cassaigne A, Begueret J: Role of alcohol dehydrogenase polymorphism in ethanol metabolism and alcohol-related diseases. Adv Exp Med Biol 1991;284:263270.
Doi R, Chowdhury P, Nishikawa M, Takaori K, Inoue K, Imamura M,
Rayford PL: Carbachol and cholecystokinin enhance accumulation of
nicotine in rat pancreatic acinar cells. Pancreas 1995;10:154-160.
Dorn JM, Hovey K, Williams BA, Freudenheim JL, Russell M, Nochajski
TH, Trevisan M: Alcohol drinking pattern and non-fatal myocardial
infarction in women. Addiction 2007;102:730-739.
Chowdhury P, Doi R, Tangoku A, Rayford PL: Structural and functional
changes of rat exocrine pancreas exposed to nicotine. Int J Pancreatol
1995;18:257-264.
Corrao G, Bagnardi V, Zambon A, La Vecchia C: A meta-analysis of
alcohol consumption and the risk of 15 diseases. Prev Med
2004;38:613-619.
Djousse L, Levy D, Herbert AG, Wilson PW, D'Agostino RB, Cupples LA,
Karamohamed S, Ellison RC: Influence of alcohol dehydrogenase 1C
polymorphism on the alcohol-cardiovascular disease association
(from the Framingham Offspring Study). Am J Cardiol 2005a;96:227232.
Djousse L, Levy D, Herbert AG, Wilson PW, D'Agostino RB, Cupples LA,
Karamohamed S, Ellison RC: Influence of alcohol dehydrogenase 1C
polymorphism on the alcohol-cardiovascular disease association
(from the Framingham Offspring Study). Am J Cardiol 2005b;96:227232.
Bourliere M, Barthet M, Berthezene P, Durbec JP, Sarles H: Is tobacco a
risk factor for chronic pancreatitis and alcoholic cirrhosis? Gut
1991;32:1392-1395.
Colditz GA, Willett WC, Stampfer MJ, Rosner B, Speizer FE, Hennekens
CH: Menopause and the risk of coronary heart disease in women. N
Engl J Med 1987;316:1105-1110.
Djousse L, Levy D, Benjamin EJ, Blease SJ, Russ A, Larson MG, Massaro
JM, D'Agostino RB, Wolf PA, Ellison RC: Long-term alcohol consumption and the risk of atrial fibrillation in the Framingham Study. Am J
Cardiol 2004;93:710-713.
Durbec JP & Sarles H: Multicenter survey of the etiology of pancreatic
diseases. Relationship between the relative risk of developing chronic
pancreaitis and alcohol, protein and lipid consumption. Digestion
1978;18:337-350.
Ebrahim S, Lawlor DA, Shlomo YB, Timpson N, Harbord R, Christensen
M, Baban J, Kiessling M, Day I, Gaunt T, Davey SG: Alcohol dehydrogenase type 1C (ADH1C) variants, alcohol consumption traits, HDLcholesterol and risk of coronary heart disease in women and men:
British Women's Heart and Health Study and Caerphilly cohorts.
Atherosclerosis 2008;196:871-878.
Edlund Y, Norback B, Risholm L: Acute pancreatitis, etiology and
prevention of recurrence. Follow-up study of 188 patients. Rev Surg
1968;25:153-157.
Ettinger PO, Wu CF, De La Cruz C, Jr., Weisse AB, Ahmed SS, Regan TJ:
Arrhythmias and the "Holiday Heart": alcohol-associated cardiac
rhythm disorders. Am Heart J 1978;95:555-562.
Frankel EN, Kanner J, German JB, Parks E, Kinsella JE: Inhibition of
oxidation of human low-density lipoprotein by phenolic substances in
red wine. Lancet 1993;341:454-457.
- 16 -
Friedrich, N. Chronic pancreatitis. In: Cyclopaedia of the practice of
medicine, New York, William Wood & Co: 1878; 599-608.
Jørgensen T: Gall stones in a Danish population. Relation to weight,
physical activity, smoking, coffee consumption, and diabetes mellitus.
Gut 1989;30:528-534.
Frost L & Vestergaard P: Alcohol and risk of atrial fibrillation or flutter: a cohort study. Arch Intern Med 2004;164:1993-1998.
Jørgensen T, Kay L, Schultz-Larsen K: The epidemiology of gallstones
in a 70-year-old Danish population. Scand J Gastroenterol
1990;25:335-340.
Gavaler JS, Deal SR, Van Thiel DH, Arria A, Allan MJ: Alcohol and estrogen levels in postmenopausal women: the spectrum of effect. Alcoholism: Clinical and Experimental Research 1993;17:786-790.
Klatsky AL: Drink to your health? Sci Am 2003;288:74-81.
Grønbæk M, Deis A, Sorensen TI, Becker U, Schnohr P, Jensen G: Mortality associated with moderate intakes of wine, beer, or spirits. BMJ
1995;310:1165-1169.
Kloppel G & Maillet B: [Development of chronic pancreatitis from
acute pancreatitis: a pathogenetic concept]. Zentralbl Chir
1995;120:274-277.
Grønbæk M, Mortensen EL, Mygind K, Andersen AT, Becker U, Gluud C,
Sørensen TIA: Beer, wine, spirits and subjective health. J Epidemiol
Community Health 1999;53:721-724.
Koskinen P, Kupari M, Leinonen H, Luomanmaki K: Alcohol and new
onset atrial fibrillation: a case-control study of a current series. Br
Heart J 1987;57:468-473.
Gutjahr E, Gmel G, Rehm J: Relation between average alcohol consumption and disease: an overview. Eur Addict Res 2001;7:117-127.
Koul PB, Sussmane JB, Cunill-De Sautu B, Minarik M: Atrial fibrillation
associated with alcohol ingestion in adolescence: holiday heart in
pediatrics. Pediatr Emerg Care 2005;21:38-39.
Hammar N, Romelsjo A, Alfredsson L: Alcohol consumption, drinking
pattern and acute myocardial infarction. A case referent study based
on the Swedish Twin Register. Journal of Internal Medicine
1997;241:125-131.
Krishnamoorthy S, Lip GY, Lane DA: Alcohol and illicit drug use as
precipitants of atrial fibrillation in young adults: a case series and
literature review. Am J Med 2009;122:851-856.
Heidrich J, Wellmann J, Doring A, Illig T, Keil U: Alcohol consumption,
alcohol dehydrogenase and risk of coronary heart disease in the MONICA/KORA-Augsburg cohort 1994/1995-2002. Eur J Cardiovasc Prev
Rehabil 2007;14:769-774.
Kristiansen L, Gronbaek M, Becker U, Tolstrup JS: Risk of pancreatitis
according to alcohol drinking habits: a population-based cohort study.
Am J Epidemiol 2008;168:932-937.
Hill JA: In vino veritas: alcohol and heart disease. Am J Med Sci
2005;329:124-135.
Laatikainen T, Manninen L, Poikolainen K, Vartiainen E: Increased
mortality related to heavy alcohol intake pattern. J Epidemiol Community Health 2003;57:379-384.
Hilton ME: Drinking patterns and drinking problems in 1984: results
from a general population survey. Alcohol Clin Exp Res 1987;11:167175.
Langer RD, Criqui MH, Reed DM: Lipoproteins and blood pressure as
biological pathways for effect of moderate alcohol consumption on
coronary heart disease. Circulation 1992;85:910-915.
Hines LM, Stampfer MJ, Ma J, Gaziano JM, Ridker PM, Hankinson SE,
Sacks F, Rimm EB, Hunter DJ: Genetic variation in alcohol dehydrogenase and the beneficial effect of moderate alcohol consumption on
myocardial infarction. N Engl J Med 2001;344:549-555.
Law R, Parsi M, Lopez R, Zuccaro G, Stevens T: Cigarette smoking is
independently associated with chronic pancreatitis. Pancreatology
2010;10:54-59.
Hvidtfeldt UA, Tolstrup JS, Jakobsen MU, Heitmann BL, Gronbaek M,
O'Reilly E, Balter K, Goldbourt U, Hallmans G, Knekt P, Liu S, Pereira M,
Pietinen P, Spiegelman D, Stevens J, Virtamo J, Willett WC, Rimm EB,
Ascherio A: Alcohol intake and risk of coronary heart disease in
younger, middle-aged, and older adults. Circulation 2010;121:15891597.
Idler EL & Benyamini Y: Self-rated health and mortality: a review of
twenty-seven community studies. J Health Soc Behav 1997;38:21-37.
Irving HM, Samokhvalov AV, Rehm J: Alcohol as a risk factor for pancreatitis. A systematic review and meta-analysis. JOP 2009;10:387392.
Johansen D, Friis K, Skovenborg E, Gronbaek M: Food buying habits of
people who buy wine or beer: cross sectional study. BMJ
2006;332:519-522.
Lee KW & Lip GY: Effects of lifestyle on hemostasis, fibrinolysis, and
platelet reactivity: a systematic review. Arch Intern Med
2003;163:2368-2392.
Lewington S, Clarke R, Qizilbash N, Peto R, Collins R: Age-specific
relevance of usual blood pressure to vascular mortality: a metaanalysis of individual data for one million adults in 61 prospective
studies. Lancet 2002;360:1903-1913.
Lieber CS: Alcoholic fatty liver: its pathogenesis and mechanism of
progression to inflammation and fibrosis. Alcohol 2004;34:9-19.
Lin Y, Tamakoshi A, Hayakawa T, Ogawa M, Ohno Y: Cigarette smoking
as a risk factor for chronic pancreatitis: a case-control study in Japan.
Research Committee on Intractable Pancreatic Diseases. Pancreas
2000;21:109-114.
Lin Y, Tamakoshi A, Hayakawa T, Ogawa M, Ohno Y: Associations of
- 17 -
alcohol drinking and nutrient intake with chronic pancreatitis: findings from a case-control study in Japan. Am J Gastroenterol
2001;96:2622-2627.
Danish adults. Arch Intern Med 2001;161:1844-1848.
Morton C, Klatsky AL, Udaltsova N: Smoking, coffee, and pancreatitis.
Am J Gastroenterol 2004;99:731-738.
Lindkvist B, Appelros S, Manjer J, Berglund G, Borgstrom A: A prospective cohort study of smoking in acute pancreatitis. Pancreatology
2008;8:63-70.
Livingston M & Room R: Variations by age and sex in alcohol-related
problematic behaviour per drinking volume and heavier drinking
occasion. Drug Alcohol Depend 2009;101:169-175.
Mukamal KJ, Conigrave KM, Mittleman MA, Camargo CA, Jr., Stampfer
MJ, Willett WC, Rimm EB: Roles of drinking pattern and type of alcohol
consumed in coronary heart disease in men. N Engl J Med
2003;348:109-118.
Mukamal KJ, Jensen MK, Gronbaek M, Stampfer MJ, Manson JE, Pischon
T, Rimm EB: Drinking frequency, mediating biomarkers, and risk of
myocardial infarction in women and men. Circulation
2005a;112:1406-1413.
Lowenfels AB, Zwemer FL, Jhangiani S, Pitchumoni CS: Pancreatitis in
a native American Indian population. Pancreas 1987;2:694-697.
Lowenstein SR, Gabow PA, Cramer J, Oliva PB, Ratner K: The role of
alcohol in new-onset atrial fibrillation. Arch Intern Med
1983;143:1882-1885.
Mukamal KJ, Psaty BM, Rautaharju PM, Furberg CD, Kuller LH, Mittleman MA, Gottdiener JS, Siscovick DS: Alcohol consumption and risk
and prognosis of atrial fibrillation among older adults: the Cardiovascular Health Study. Am Heart J 2007;153:260-266.
Mackay, J & Mensah, GA. Atlas of Heart Disease and Stroke. Brighton:
Myriad Editions Limited, 2004.
Mukamal KJ & Rimm EB: Alcohol's effects on the risk for coronary
heart disease. Alcohol Res Health 2001;25:255-261.
Maclure M: Demonstration of deductive meta-analysis: ethanol intake
and risk of myocardial infarction. Epidemiol Rev 1993;15:328-351.
Makela P, Paljarvi T, Poikolainen K: Heavy and nonheavy drinking
occasions, all-cause and cardiovascular mortality and hospitalizations:
a follow-up study in a population with a low consumption level. J Stud
Alcohol 2005;66:722-728.
Malyutina S, Bobak M, Kurilovitch S, Gafarov V, Simonova G, Nikitin Y,
Marmot M: Relation between heavy and binge drinking and all-cause
and cardiovascular mortality in Novosibirsk, Russia: a prospective
cohort study. Lancet 2002;360:1448-1454.
Mukamal KJ, Tolstrup JS, Friberg J, Jensen G, Gronbaek M: Alcohol
consumption and risk of atrial fibrillation in men and women: the
Copenhagen City Heart Study. Circulation 2005b;112:1736-1742.
Mumenthaler MS, Taylor JL, O'Hara R, Yesavage JA: Gender differences
in moderate drinking effects. Alcohol Research and Health 1999;23:5564.
Murray RP, Connett JE, Tyas SL, Bond R, Ekuma O, Silversides CK,
Barnes GE: Alcohol volume, drinking pattern, and cardiovascular
disease morbidity and mortality: is there a U-shaped function? Am J
Epidemiol 2002;155:242-248.
Marenberg ME, Risch N, Berkman LF, Floderus B, de Faire U: Genetic
susceptibility to death from coronary heart disease in a study of twins.
N Engl J Med 1994;330:1041-1046.
Marks, V, Cantor, T, Mesko, D, Pullmann, R, and Nosalova, G. Differential diagnosis by laboratory medicine. Springer, 2002.
Matsumoto M, Takahashi H, Maruyama K, Higuchi S, Matsushita S,
Muramatsu T, Okuyama K, Yokoyama A, Nakano M, Ishii H: Genotypes
of alcohol-metabolizing enzymes and the risk for alcoholic chronic
pancreatitis in Japanese alcoholics. Alcohol Clin Exp Res 1996;20:289A292A.
McElduff P & Dobson AJ: How much alcohol and how often? Population based case-control study of alcohol consumption and risk of a
major coronary event. Br Med J 1997;314:1159-1164.
Neumark YD, Friedlander Y, Durst R, Leitersdorf E, Jaffe D, Ramchandani VA, O'Connor S, Carr LG, Li TK: Alcohol dehydrogenase polymorphisms influence alcohol-elimination rates in a male Jewish population. Alcohol Clin Exp Res 2004;28:10-14.
Nordestgaard BG, Benn M, Schnohr P, Tybjaerg-Hansen A: Nonfasting
triglycerides and risk of myocardial infarction, ischemic heart disease,
and death in men and women. JAMA 2007;298:299-308.
Osier MV, Pakstis AJ, Soodyall H, Comas D, Goldman D, Odunsi A,
Okonofua F, Parnas J, Schulz LO, Bertranpetit J, Bonne-Tamir B, Lu RB,
Kidd JR, Kidd KK: A global perspective on genetic variation at the ADH
genes reveals unusual patterns of linkage disequilibrium and diversity. Am J Hum Gen 2002;71:84-99.
Pace-Asciak CR, Hahn S, Diamandis EP, Soleas G, Goldberg DM: The red
wine phenolics trans-resveratrol and quercetin block human platelet
aggregation and eicosanoid synthesis: implications for protection
against coronary heart disease. Clin Chim Acta 1995;235:207-219.
Mizoi Y, Yamamoto K, Ueno Y, Fukunaga T, Harada S: Involvement of
genetic polymorphism of alcohol and aldehyde dehydrogenases in
individual variation of alcohol metabolism. Alcohol Alcohol
1994;29:707-710.
Mortensen EL, Jensen HH, Sanders SA, Reinisch JM: Better psychological functioning and higher social status may largely explain the apparent health benefits of wine: a study of wine and beer drinking in young
Pace-Asciak CR, Rounova O, Hahn SE, Diamandis EP, Goldberg DM:
Wines and grape juices as modulators of platelet aggregation in
healthy human subjects. Clin Chim Acta 1996;246:163-182.
- 18 -
Piano MR, Rosenblum C, Solaro RJ, Schwertz D: Calcium sensitivity and
the effect of the calcium sensitizing drug pimobendan in the alcoholic
isolated rat atrium. J Cardiovasc Pharmacol 1999;33:237-242.
Takeshita T, Mao XQ, Morimoto K: The contribution of polymorphism
in the alcohol dehydrogenase beta subunit to alcohol sensitivity in a
Japanese population. Hum Genet 1996;97:409-413.
Psaty BM, Manolio TA, Kuller LH, Kronmal RA, Cushman M, Fried LP,
White R, Furberg CD, Rautaharju PM: Incidence of and risk factors for
atrial fibrillation in older adults. Circulation 1997;96:2455-2461.
Takeshita T, Yang X, Morimoto K: Association of the ADH2 genotypes
with skin responses after ethanol exposure in Japanese male university students. Alcohol Clin Exp Res 2001;25:1264-1269.
Ramchandani VA, Bosron WF, Li TK: Research advances in ethanol
metabolism. Pathol Biol (Paris) 2001;49:676-682.
Talamini G, Bassi C, Falconi M, Frulloni L, Di F, V, Vaona B, Bovo P, Rigo
L, Castagnini A, Angelini G, Vantini I, Pederzoli P, Cavallini G: Cigarette
smoking: an independent risk factor in alcoholic pancreatitis. Pancreas
1996;12:131-137.
Ramprasath VR & Jones PJ: Anti-atherogenic effects of resveratrol. Eur
J Clin Nutr 2010;
Talamini G, Bassi C, Falconi M, Sartori N, Salvia R, Rigo L, Castagnini A,
Di F, V, Frulloni L, Bovo P, Vaona B, Angelini G, Vantini I, Cavallini G,
Pederzoli P: Alcohol and smoking as risk factors in chronic pancreatitis and pancreatic cancer. Dig Dis Sci 1999;44:1303-1311.
Register TC, Cline JM, Shively CA: Health issues in postmenopausal
women who drink. Alcohol Research and Health 2002;26:299-307.
Rehm J, Gmel G, Sempos CT, Trevisan M: Alcohol-related morbidity
and mortality. Alcohol Res Health 2003;27:39-51.
Thornton JR: Atrial fibrillation in healthy non-alcoholic people after an
alcoholic binge. Lancet 1984;2:1013-1015.
Rehm J, Greenfield TK, Rogers JD: Average volume of alcohol consumption, patterns of drinking, and all-cause mortality: results from
the US National Alcohol Survey. Am J Epidemiol 2001;153:64-71.
Tjønneland A, Grønbæk M, Stripp C, Overvad K: Wine intake and diet
in a random sample of 48763 Danish men and women. Am J Clin Nutr
1999;69:49-54.
Rehm J, Mathers C, Popova S, Thavorncharoensap M, Teerawattananon Y, Patra J: Global burden of disease and injury and economic cost
attributable to alcohol use and alcohol-use disorders. Lancet
2009a;373:2223-2233.
Tolstrup J, Jensen MK, Tjonneland A, Overvad K, Mukamal KJ, Gronbaek M: Prospective study of alcohol drinking patterns and coronary
heart disease in women and men. BMJ 2006;332:1244-1248.
Rehm J & Mathers C: Alcohol and global health: focus on acute pancreatitis needed - Authors' reply. The Lancet 2009b;374:977.
Tolstrup JS, Gronbaek M, Nordestgaard BG: Alcohol Intake, Myocardial
Infarction, Biochemical Risk Factors, and Alcohol Dehydrogenase
Genotypes. Circulation: Cardiovascular Genetics 2009a;2:507-514.
Renaud S & de Lorgeril M: Wine, alcohol, platelets, and the French
paradox for coronary heart disease. Lancet 1992;339:1523-1526.
Tolstrup JS, Gronbaek M, Tybjaerg-Hansen A, Nordestgaard BG: Alcohol intake, alcohol dehydrogenase genotypes, and liver damage and
disease in the Danish general population. Am J Gastroenterol
2009b;104:2182-2188.
Renaud SC & Ruf JC: Effects of alcohol on platelet functions. Clin Chim
Acta 1996;246:77-89.
Rimm E & Moats C: Alcohol and coronary heart disease: drinking
patterns and mediators of effect. Ann Epidemiol 2007;17:S3-S7.
Tolstrup JS, Hansen JL, Gronbaek M, Vogel U, Tjonneland A, Joensen A,
Overvad K: Alcohol drinking habits, alcohol dehydrogenase genotypes
and risk of acute coronary syndrome. Scand J Public Health
2010;38:489-494.
Rimm EB, Klatsky A, Grobbee D, Stampfer MJ: Review of moderate
alcohol consumption and reduced risk of coronary heart disease: is
the effect due to beer, wine or spirits? BMJ 1996;312:731-736.
Tolstrup JS, Jensen MK, Tjonneland A, Overvad K, Gronbaek M: Drinking pattern and mortality in middle-aged men and women. Addiction
2004;99:323-330.
Rimm EB, Williams P, Fosher K, Criqui M, Stampfer MJ: Moderate
alcohol intake and lower risk of coronary heart disease: meta-analysis
of effects on lipids and haemostatic factors. BMJ 1999;319:1523-1528.
Tolstrup JS, Kristiansen L, Becker U, Gronbaek M: Smoking and risk of
acute and chronic pancreatitis among women and men: a populationbased cohort study. Arch Intern Med 2009c;169:603-609.
Ruigomez A, Johansson S, Wallander MA, Rodriguez LA: Incidence of
chronic atrial fibrillation in general practice and its treatment pattern.
J Clin Epidemiol 2002;55:358-363.
Tolstrup JS, Nordestgaard BG, Rasmussen S, Tybjaerg-Hansen A, Gronbaek M: Alcoholism and alcohol drinking habits predicted from alcohol dehydrogenase genes. Pharmacogenomics J 2007;8:220-227.
Sarles H, Cros RC, Bidart JM: A multicenter inquiry into the etiology of
pancreatic diseases. Digestion 1979;19:110-125.
Sun L, Konig IR, Jacobs A, Seitz HK, Junghanns K, Wagner T, Ludwig D,
Jacrobs A, Homann N: Mean corpuscular volume and ADH1C genotype
in white patients with alcohol-associated diseases. Alcohol Clin Exp Res
2005;29:788-793.
Trevisan M, Dorn J, Falkner K, Russell M, Ram M, Muti P, Freudenheim
JL, Nochajaski T, Hovey K: Drinking pattern and risk of non-fatal myocardial infarction: a population-based case-control study. Addiction
2004;99:313-322.
- 19 -
Trevisan M, Schisterman E, Mennotti A, Farchi G, Conti S: Drinking
pattern and mortality: the Italian Risk Factor and Life Expectancy
pooling project. Ann Epidemiol 2001;11:312-319.
Zdravkovic S, Wienke A, Pedersen NL, Marenberg ME, Yashin AI, de
Faire U: Heritability of death from coronary heart disease: a 36-year
follow-up of 20 966 Swedish twins. J Intern Med 2002;252:247-254.
Verlaan M, te Morsche RH, Roelofs HM, Laheij RJ, Jansen JB, Peters WH,
Drenth JP: Genetic polymorphisms in alcohol-metabolizing enzymes
and chronic pancreatitis. Alcohol Alcohol 2004;39:20-24.
Zintzaras E, Stefanidis I, Santos M, Vidal F: Do alcohol-metabolizing
enzyme gene polymorphisms increase the risk of alcoholism and
alcoholic liver disease? Hepatology 2006;43:352-361.
Voirol M, Infante F, Brahime-Reteno O, Raymond L, Hollenweger V,
Loizeau E: Consommation d'alcool , de tabac et de nutriments dans les
affections pancreatiques. Schweiz Med Wochenschr 1980;110:854-855.
Whitfield JB: ADH and ALDH genotypes in relation to alcohol metabolic rate and sensitivity. Alcohol Alcohol Suppl 1994;2:59-65.
Whitfield JB, O'Brien ME, Nightingale BN, Zhu G, Heath AC, Martin NG:
ADH genotype does not modify the effects of alcohol on high-density
lipoprotein. Alcohol Clin Exp Res 2003;27:509-514.
Whitfield JB, Zhu G, Duffy DL, Birley AJ, Madden PA, Heath AC, Martin
NG: Variation in alcohol pharmacokinetics as a risk factor for alcohol
dependence. Alcohol Clin Exp Res 2001;25:1257-1263.
WHO. Global status report on alcohol 2004. Geneva: World Health
Organization, 2004.
Witt H, Apte MV, Keim V, Wilson JS: Chronic pancreatitis: challenges
and advances in pathogenesis, genetics, diagnosis, and therapy. Gastroenterology 2007;132:1557-1573.
Wittel UA, Hopt UT, Batra SK: Cigarette smoke-induced pancreatic
damage-experimental data. Langenbecks Arch Surg 2008;393:581-588.
Wittel UA, Pandey KK, Andrianifahanana M, Johansson SL, Cullen DM,
Akhter MP, Brand RE, Prokopczyk B, Batra SK: Chronic pancreatic
inflammation induced by environmental tobacco smoke inhalation in
rats. Am J Gastroenterol 2006;101:148-159.
Yadav D, Hawes RH, Brand RE, Anderson MA, Money ME, Banks PA,
Bishop MD, Baillie J, Sherman S, DiSario J, Burton FR, Gardner TB,
Amann ST, Gelrud A, Lawrence C, Elinoff B, Greer JB, O'Connell M,
Barmada MM, Slivka A, Whitcomb DC: Alcohol consumption, cigarette
smoking, and the risk of recurrent acute and chronic pancreatitis. Arch
Intern Med 2009;169:1035-1045.
Yen S, Hsieh CC, MacMahon B: Consumption of alcohol and tobacco
and other risk factors for pancreatitis. Am J Epidemiol 1982;116:407414.
Yokoyama T, Yokoyama A, Kato H, Tsujinaka T, Muto M, Omori T,
Haneda T, Kumagai Y, Igaki H, Yokoyama M, Watanabe H, Yoshimizu
H: Alcohol flushing, alcohol and aldehyde dehydrogenase genotypes,
and risk for esophageal squamous cell carcinoma in Japanese men.
Cancer Epidemiol Biomarkers Prev 2003;12:1227-1233.
Younis J, Cooper JA, Miller GJ, Humphries SE, Talmud PJ: Genetic variation in alcohol dehydrogenase 1C and the beneficial effect of alcohol
intake on coronary heart disease risk in the Second Northwick Park
Heart Study. Atherosclerosis 2005;180:225-232.
- 20 -
Cite this article as: BMJ, doi:10.1136/bmj.38831.503113.7C (published 3 May 2006)
Research
Prospective study of alcohol drinking patterns and coronary heart
disease in women and men
Janne Tolstrup, Majken K Jensen, Anne Tjønneland, Kim Overvad, Kenneth J Mukamal, Morten Grønbæk
Abstract
Objective To determine the association between alcohol
drinking patterns and risk of coronary heart disease in women
and men.
Design Population based cohort study.
Setting Denmark, 1993-2002.
Participants 28 448 women and 25 052 men aged 50-65 years,
who were free of cardiovascular disease at entry to the study.
Main outcome measures Incidence of coronary heart disease
occurring during a median follow-up period of 5.7 years.
Results 749 and 1283 coronary heart disease events occurred
among women and men. Women who drank alcohol on at least
one day a week had a lower risk of coronary heart disease than
women who drank alcohol on less than one day a week. Little
difference was found, however, between drinking frequency:
one day a week (hazard ratio 0.64, 95% confidence interval 0.51
to 0.81), 2-4 days a week (0.63, 0.52 to 0.77), five or six days a
week (0.79, 0.61 to 1.03), and seven days a week (0.65, 0.51 to
0.84). For men an inverse association was found between
drinking frequency and risk of coronary heart disease across
the entire range of drinking frequencies. The lowest risk was
observed among men who drank daily (0.59, 0.48 to 0.71)
compared with men who drank alcohol on less than one day a
week.
Conclusions Among women alcohol intake may be the
primary determinant of the inverse association between
drinking alcohol and risk of coronary heart disease whereas
among men, drinking frequency, not alcohol intake, seems
more important.
Introduction
Prospective studies have consistently reported a lower risk of
coronary heart disease among consumers of moderate alcohol
compared with abstainers.1 A few studies have investigated this
association by also including various measures of alcohol drinking patterns. Results consistently imply that the pattern of drinking is important and that steady drinking is more beneficial than
drinking in binges.2–6 In a recent such study among men it was
suggested that drinking frequency is the primary determinant of
the inverse association between alcohol intake and coronary
heart disease, and that alcohol intake is of minor importance.6
Some issues still warrant consideration however; most importantly, data on the importance of drinking patterns among
women are limited and results obtained among men may not
apply to women for different reasons. Firstly, sex differences in
alcohol pharmacokinetics have been reported, suggesting that
men have more efficient first pass metabolisms than women
BMJ Online First bmj.com
whereas women may eliminate alcohol faster than men.7
Secondly, oestrogen has beneficial effects on the cardiovascular
system, and studies have suggested that alcohol increases oestrogen levels.8
We determined the association between alcohol drinking
patterns and coronary heart disease among men and women
participating in a population based cohort study consisting of
middle aged Danish citizens.
Methods
From December 1993 to May 1997, 160 725 Danish men and
women were invited to participate in the diet, cancer, and health
study.9 Eligible cohort members were born in Denmark and had
no previous cancers. Overall, 27 178 men and 29 875 women
agreed to participate (response rate 35%). A detailed food
frequency questionnaire consisting of 192 items was enclosed
with the invitation.10 11 This questionnaire was checked by an
interviewer during a clinic visit, when another questionnaire
concerning lifestyle and background factors was completed.
In the food frequency questionnaire alcohol intake was
reported as the average amount over the preceding year. Total
intake was calculated and converted into number of standard
drinks, defined as containing 12 g of ethanol. Drinking frequency
was reported in the background questionnaire in predefined categories (never, less than once a month, 1-3 times monthly, once
a week, 2-4 times weekly, 5 or 6 times weekly, and daily). We
defined abstainers as those who reported no alcohol intake
(amount) and no drinking occasions (frequency). To increase
homogeneity among abstainers we excluded 786 people who
reported no amount but a frequency greater than zero (or vice
versa). We also excluded people with missing information
(n = 303) or with conflicting answers on amount and frequency
of alcohol intake (n = 97). In all, 53 500 people were eligible for
this study.
Follow-up
We obtained information on coronary heart disease from the
Danish Hospital Discharge Register12 and from the Danish Register of Causes of Death,13 where, respectively, all admissions to
hospital for somatic conditions and causes of death in Denmark
are registered. The hospital register is updated to 2002, whereas
the causes of death register, which contains information on fatal
incidents of coronary heart disease, is updated to 2000. In the
period that was covered by both registers, the causes of death
register contributed information on only 8% of cases. Hence we
decided to end follow-up at January 2002, being aware that
information on some fatal cases would be missed from January
2000 to January 2002.
page 1 of 5
Research
In both registers diagnoses are classified according to the
international classification of diseases, eighth and 10th revisions
(codes for coronary heart disease: ICD-8, 410-414 and ICD-10,
I20-I25). We obtained vital status of the participants from the
National Central Person Register. To minimise the risk of including preclinical cases, we excluded 2367 participants who, at baseline, were registered with any cardiovascular disease (ischaemic
stroke, arrhythmias, congestive heart failure, or peripheral
arteriosclerosis).
We observed participants from enrolment until date of coronary heart event (n = 2113), death from other causes (n = 1483),
emigration (n = 183), loss to follow-up (n = 3), or 1 January 2002,
whichever came first.
Statistical analysis
We calculated risk estimates using Cox proportional hazard
regression models, with delayed entry implemented (SAS/STAT
program software). To ensure maximal adjustment for
confounding by age we used age as the time axis. We adjusted the
risk estimates for known risk factors for coronary heart disease:
length of school education (short, ≤ 7 years; medium, 8-10 years;
long, ≥ 11 years); smoking (never; former; current, 1-14, 15-24,
or > 24 g of tobacco/day); physical activity during leisure time
(dummy variables were coded for each of the following activities:
sports, walking, bicycling, housework, gardening, do it yourself);
body mass index (modelled as linear splines, with knots set at 20
and 25); total intake of fruit, vegetables, and fish; and percentage
of total energy intake from saturated fat (all as continuous
variables). We calculated the total intake of different dietary factors using the software program Food Calc (release 1.3,
www.FoodCalc.dk). Using linear splines with knots set at quintiles
of the covariate in question we evaluated the assumed linearity of
quantitative risk factors. We tested assumptions of the
proportional hazards model but detected no violations.
To test for linear trends we treated the median value within
categories continuously. We did not include abstainers when
testing for trend because they may have different trait and health
status than people who consume alcohol lightly to moderately,14
causing a spuriously high relative risk in this category.
To examine the magnitude of rank correlation between
drinking frequency and amount of alcohol, we calculated Spearman’s correlation coefficient. We tested the interaction between
sex and drinking frequency using a nested log likelihood test
where we compared a model containing the variables as single
terms with a model including the interaction terms.
To tackle the external validity of our results, we compared the
observed with the expected number of cases, on the basis of age
and calendar year specific incidence rates in the general Danish
population. These were calculated using nationwide information
from the Danish Hospital Discharge Register.12
Results
Overall, 53 500 people were eligible for our study: 28 448
women and 25 052 men. Women consumed a median of 5.5
alcoholic drinks a week (fifth to 95th centiles, 0.3-24) and men
11.3 (1.1-47). Drinking frequency was highly correlated with
amount of alcohol intake among both women and men (r = 0.86
and r = 0.78).
Infrequent drinkers (less than one day a week) and daily
drinkers (daily) were more likely to be smokers, to have a lower
intake of fruit and vegetables, and to be less educated than participants in the in between drinking frequencies (table 1). Body
mass index was inversely associated with drinking frequency and
frequent drinkers had the lowest body mass index. These trends
page 2 of 5
were the same for both sexes. Generally, fewer women than men
were current and heavy smokers ( > 25 g of tobacco daily) and
women had more hours of physical activity a week and
consumed more fruit and vegetables.
During follow-up (median 5.7 years, range 0.01-8.10) 749
women and 1283 men developed coronary heart disease. Information on 1933 of these cases came from the Danish Hospital
Discharge Register. Based on incidence rates from the general
population the expected number of cases from this register was
716 women (737 observed) and 1217 men (1196 observed). The
observed number did not differ significantly from the expected
(P > 0.10).
Amount of alcohol intake was inversely associated with coronary heart disease among women and men (figure).
Among women, drinking on at least one day a week was
associated with a lower risk of coronary heart disease than drinking more rarely than one day a week (table 2). Hazard ratios were
similar for drinking on one day a week (0.64, 95% confidence
interval 0.51 to 0.81), 2-4 days a week (0.63, 0.52 to 0.77), five or
six days a week (0.79, 0.61 to 1.03), and seven days a week (0.65,
0.51 to 0.84). A test for trend not including women that were
drinking more rarely than one day a week was statistically insignificant (P = 0.49).
Among men, drinking frequency was inversely associated
with risk of coronary heart disease over the whole range of
drinking frequencies (table 2). Hazard ratios were 0.93 (0.75 to
1.16) for drinking on one day a week, 0.78 (0.66 to 0.94) for 2-4
days a week, 0.71 (0.57 to 0.87) for five or six days a week, and
0.59 (0.48 to 0.71) for seven days a week (P for trend < 0.0001).
The test for linear trend remained statistically significant after
excluding men drinking more rarely than on one day a week
(P < 0.0001).
A statistically significant interaction was found between sex
and drinking frequency on the risk of coronary heart disease
(P = 0.02).
Table 3 lists the hazard ratios of coronary heart disease for
different combinations of alcohol amount and drinking
frequency. Within similar categories of drinking frequency,
women drinking the largest amounts generally had the lowest
risk. For example, among women drinking on 2-4 days a week
the hazard ratio was 0.78 (0.63 to 0.97) for 1-6 drinks a week, 0.74
(0.57 to 0.96) for 7-13 drinks a week, and 0.27 (0.13 to 0.58) for
14 or more drinks a week (P for trend < 0.0001). For men, hazard ratios were generally lowest for the most frequent intake
within similar categories of amount (table 3). For example,
among men drinking on average 7-13 drinks a week, hazard
ratios of coronary heart disease were 0.89 (0.62 to 1.29) for
drinking alcohol on one or less days a week, 0.81 (0.67 to 0.98)
for 2-4 days a week, and 0.66 (0.52 to 0.83) for 5-7 days a week (P
for trend = 0.0001). Within categories of drinking frequency,
hazard ratios tended to be similar.
To examine the possibility that latent baseline symptoms of
coronary heart disease such as angina pectoris might reduce the
frequency of drinking alcohol, thereby biasing the results, we
carried out analyses to compare the association between
drinking frequency and coronary heart disease only including
early cases—that is, cases that occurred within the first two years
of follow-up (n = 200 women and n = 381 men)—with the
association including only later cases (n = 549 women and
n = 902 men). An inverse association was observed in both
groups (data not shown).
BMJ Online First bmj.com
Research
Table 1 Baseline characteristics of 28 448 women and 25 052 men participating in the Danish diet, cancer, and health study according-sex and frequency of
drinking alcohol. Values are medians (5th-95th centiles) unless stated otherwise
Frequency of drinking alcohol (days/week)
Characteristic
Never
<1
1
2-4
5 or 6
7
Women:
Number
Median (range) age (years)
613
7160
4320
9373
3117
3865
57 (50-64)
56 (50-64)
56 (50-64)
55 (50-63)
55 (50-64)
56 (50-64)
Alcohol intake (drinks/week)
0 (0)
1 (0.2-5)
3.5 (1.1-9)
6.6 (3-19)
12.7 (5.9-27)
19.1 (7.3-40)
No (%) current smokers
252 (41.1)
2556 (35.7)
1266 (29.3)
2531 (27.0)
963 (30.9)
1650 (42.7)
No (%) current heavy
smokers*
78 (12.7)
508 (7.1)
251 (5.8)
600 (6.4)
256 (8.2)
502 (13.0)
No (%) educated at school ≤7
years
267 (43.6)
3072 (42.9)
1481 (34.3)
2456 (26.2)
598 (19.2)
877 (22.7)
Physical activity (hours/week)†
14 (4-41)
15 (5-40)
15 (5-38)
15 (6-35)
15 (6-34)
15 (5-38)
Body mass index
25.3 (19-38)
25.7 (20-36)
25.3 (20-34)
24.6 (20-32)
24.1 (20-32)
23.8 (19-32)
Vegetable intake (g/day)
146 (26-445)
157 (44-403)
169 (50-398)
178 (59-380)
186 (65-389)
172 (51-378)
Fruit intake (g/day)
188 (18-644)
207 (36-609)
213 (42-565)
202 (44-547)
195 (38-531)
167 (27-492)
Fish intake (g/day)
30 (3-84)
32 (9-85)
36 (11-85)
37 (13-83)
37 (13-83)
37 (11-86)
Saturated fat (% of total
energy)
13 (8-19)
13 (8-18)
13 (8-17)
12 (8-17)
12 (8-16)
12 (8-16)
Men:
Number
Median (range) age (years)
Alcohol intake (drinks/week)
390
2464
2282
8718
4300
6898
56 (50-64)
56 (50-64)
56 (50-64)
55 (50-63)
55 (50-63)
56 (50-64)
0 (0-0)
1.6 (0.3-8)
4.5 (1.3-13)
8.9 (3.7-25)
18.7 (7-42)
26.3 (9.5-60)
No (%) current smokers
192 (49.2)
1089 (44.2)
806 (35.3)
2973 (34.1)
1518 (35.3)
3249 (47.1)
No (%) current heavy
smokers*
171 (43.8)
690 (28.0)
534 (23.4)
2171 (24.9)
1204 (28.0)
2407 (34.9)
No (%) educated at school ≤7
years
154 (39.5)
1175 (47.7)
895 (39.2)
2851 (32.7)
1213 (28.2)
2228 (32.3)
Physical activity (hours/week)†
12 (2-39)
11 (3-33)
11 (3-30)
11 (4-29)
11 (4-29)
11 (4-31)
Body mass index
25.9 (21-33)
26.5 (22-34)
26.3 (21-33)
26.1 (22-33)
26.1 (22-32)
26.0 (21-33)
Vegetable intake (g/day)
136 (25-393)
130 (38-335)
145 (47-343)
157 (53-341)
162 (55-340)
149 (44-336)
Fruit intake (g/day)
149 (14-589)
148 (20-519)
154 (28-505)
149 (27-471)
139 (24-436)
121 (16-415)
Fish intake (g/day)
34 (2-106)
36 (7-98)
39 (11-96)
42 (14-96)
43 (14-97)
43 (13-100)
Saturated fat (% of total
energy)
15 (9-19)
14 (9-18)
14 (9-18)
13 (9-17)
13 (9-16)
12 (8-16)
*Smoking more than 25 g of tobacco daily.
†Sum of recreational and household activities.
Hazard ratio
Discussion
2.25
Women
P for linear tread = 0.001
1.75
1.25
0.75
Hazard ratio
0.25
2.25
0
1-6
7-13
14-20
21-27
≥28
Men
P for linear tread <0.0001
1.75
1.25
0.75
0.25
0
1-6
7-13
14-20
21-27
28-34
≥35
No of alcoholic drinks per week
Hazard ratios (95% confidence intervals), adjusted for age, smoking, education,
physical activity, body mass index, and total intake of fruit, vegetables, fish, and
saturated fat, for coronary heart disease according to alcohol intake among
women and men. Abstainers were not included in analyses for trend
BMJ Online First bmj.com
The frequency of drinking alcohol is inversely associated with
risk of coronary heart disease among men and this was
independent of alcohol intake. Among women, alcohol intake
and not drinking frequency was inversely associated with
coronary heart disease.
A limitation of our study is that only 35% of the invited people participated and hence caution should be taken when generalising our findings. People who choose to participate may have
a different risk profile and be in better health than those who
decline. However, the observed incidence of coronary heart disease did not differ from that of the general population.
We found that the association between drinking frequency
and coronary heart disease was different for men and women.
The number of cases was substantially lower among women than
among men, however, and hence results for women are less certain and warrant further study.
We cannot exclude the possibility that participants with early
symptoms of coronary heart disease at baseline had reduced
their drinking frequency, explaining the inverse association.
However, this association persisted when we analysed early cases
separately, indicating that the observed association is unlikely to
be explained by this possible bias.
Some unhealthy traits (smoking and a low intake of fruit and
vegetables) were common at both extremes of drinking
frequency. Everyday drinking may be associated with borderline
addictive behaviour, and a strong association between smoking
page 3 of 5
Research
Table 2 Hazard ratios (95% confidence intervals) of coronary heart disease according to drinking frequency among women and men
Variable
Frequency of drinking alcohol (days/week)
Never
<1
1
2-4
5 or 6
7
P for trend*
Women:
No of cases
24
276
95
187
77
90
Adjusted for age
1.01 (0.66 to 1.53)
1.00
0.60 (0.47 to 0.76)
0.56 (0.47 to 0.68)
0.69 (0.54 to 0.89)
0.62 (0.49 to 0.79)
0.0004
Adjusted for multiple
factors†
0.92 (0.61 to 1.41)
1.00
0.64 (0.51 to 0.81)
0.63 (0.52 to 0.77)
0.79 (0.61 to 1.03)
0.65 (0.51 to 0.84)
0.007‡
Men:
No of cases
39
180
140
424
195
305
Adjusted for age
1.38 (0.98 to 1.95)
1.00
0.86 (0.69 to 1.07)
0.69 (0.58 to 0.83)
0.65 (0.53 to 0.79)
0.60 (0.50 to 0.73)
<0.0001
Adjusted for multiple
factors†
1.44 (1.02 to 2.04)
1.00
0.93 (0.75 to 1.16)
0.78 (0.66 to 0.94)
0.71 (0.57 to 0.87)
0.59 (0.48 to 0.71)
<0.0001
*Never drinkers not included in analyses for trend.
†Age, smoking, education, physical activity, body mass index, total intake of fruit, vegetables, fish, and saturated fat.
‡P for trend was 0.49 when women were excluded who never drink or drink on less than one day a week.
and drinking has been observed in many studies.15 For the most
rare drinkers, the unhealthy lifestyle may be explained by the fact
that they were the poorest educated, which probably correlates
with low social status. Also this category may include former
alcoholics. Together, results for the extremes of drinking
frequency are more likely to be residually confounded than
results for the in between drinking frequencies and should be
interpreted with caution. However, at least among men, we found
an inverse association between drinking frequency and coronary
heart disease over the entire range of drinking frequencies.
Drinking patterns in our study were constructed by combining information on average intake with drinking frequency, as
done in another study.6 We have avoided the term “binge drinking,” which is mostly defined as drinking a minimum number of
drinks per occasion and we cannot comment on this with the
present data.
Several explanations may account for a possible interaction
between sex and drinking frequency. One explanation is sex specific drinking habits, such as drinking with meals. We cannot
exclude that men who drink frequently are more likely to drink
with meals, which may contribute to a greater risk reduction
compared with men with a less frequent alcohol intake. The beneficial effect of meal related alcohol intake is, however,
controversial.6 It is unlikely that wine drinking, which may be
more beneficial than drinking beer or spirits,16 is responsible for
our results because it has been shown that wine drinkers in this
cohort drink less often than beer drinkers.17 Differences in alcohol pharmacokinetics between sexes may be another explanation.7
The association between alcohol and coronary heart disease
among women may be modified by menopausal status.
Oestrogens have beneficial effects on the cardiovascular system,
protecting women until menopause, when the incidence rapidly
approaches that among men.18 Moderate alcohol drinking is
thought to increase oestrogen levels.8 Few women in this study
(17%) were premenopausal and our findings may be limited to
postmenopausal women.
The inverse association between alcohol and coronary heart
disease can be explained by several biologically plausible mechanisms, including dose dependent effects on high density lipoprotein levels, lower plasma fibrinogen levels, and reduced platelet
aggregation.19 These potential beneficial effects of alcohol must
be considered along with potential adverse effects of a high
intake, such as high blood pressure and increased triglyceride
levels.20 The question is if the balance between beneficial and
harmful effects is affected by drinking pattern. Heavy weekend
drinkers have been found to have a higher daily blood pressure21
and to have greater between day variability in blood pressure
than heavy daily drinkers.22 23 Results are conflicting as to
whether drinking pattern modifies lipid levels. Some studies
found that only regular drinking can raise high density lipoprotein levels,24 25 whereas others found this among weekend drinkers.26 The presumed lowering effect of alcohol on fibrinogen
levels has been found to be independent of drinking pattern
(daily versus weekend drinking).27 It has not been investigated if
drinking pattern affects the presumed association between alcohol and increased oestrogen levels among women.
Table 3 Hazard ratios (95% confidence intervals) of coronary heart disease according to drinking frequency and amount of alcohol intake among women and
men
Alcohol intake
(drinks/week)
Frequency of drinking alcohol (days/week)
Never
≤1
2-4
P for trend
5-7
Women:
0
1.03 (0.68 to 1.56) (n=24)
—
—
—
—
1-6
—
1.00 (n=360)
0.78 (0.63 to 0.97) (n=114)
1.32 (0.84 to 2.07) (n=20)
0.57
7-13
—
0.67 (0.35 to 1.31) (n=9)
0.74 (0.57 to 0.96) (n=66)
0.82 (0.61 to 1.10) (n=52)
0.12
≥14
—
0.65 (0.16 to 2.61) (n=2)
0.27 (0.13 to 0.58) (n=7)
0.72 (0.57 to 0.92) (n=95)
0.01
—
0.002
<0.0001
0.0003
—
P for trend
Men:
0
1.47 (1.05 to 2.06) (n=39)
—
—
—
1-6
—
1.00 (n=278)
0.80 (0.65 to 0.98) (n=141)
0.70 (0.41 to 1.17) (n=15)
0.02
7-13
—
0.89 (0.62 to 1.29) (n=31)
0.81 (0.67 to 0.98) (n=190)
0.66 (0.52 to 0.83) (n=90)
0.0001
14-20
—
1.10 (0.54 to 2.23) (n=8)
0.91 (0.68 to 1.23) (n=52)
0.68 (0.54 to 0.87) (n=90)
0.001
≥21
—
1.00 (0.32 to 3.13) (n=3)
0.67 (0.48 to 0.93) (n=41)
0.63 (0.53 to 0.74) (n=305)
<0.0001
—
0.25
0.22
<0.0001
—
P for trend
Hazard ratios are adjusted for age, education, smoking, physical activity, body mass index, and total intake of vegetables, fruit, fish, and saturated fat. Number of cases in parentheses.
page 4 of 5
BMJ Online First bmj.com
Research
Heavy alcohol drinking is positively associated with many
problems such as liver diseases, cancers, and road crashes, and
overall mortality is higher among individuals with a high alcohol
intake compared with light consumers, reflecting that the beneficial effects of alcohol on coronary heart disease is by far
exceeded by the detrimental effects of alcohol at these levels.
Also, the beneficial effect of alcohol is probably confined to middle aged or older people.28 Therefore the inverse association
between alcohol intake and coronary heart disease should be
viewed in this context when giving public health advice. In conclusion, we found that drinking frequency seemed to be the main
determinant of the inverse association between alcohol intake
and coronary heart disease among men, which confirms results
from another study.6 For women, amount of alcohol may be
more important than frequency.
We thank the participants of the diet, cancer, and health study.
Contributors: JT contributed to the conception and design of the study, the
analysis and interpretation of data, and wrote the manuscript. MKJ, KJM,
and MG contributed to the conception and design of the study, interpretation of data, and to critically revising the paper. AT and KO contributed to
the design of the study, the acquisition of data, interpretation of data, and
critically revising the paper. All authors approved the final version of the
article. Katja Boll prepared the data file and Søren Rasmussen calculated
incidence rates of coronary heart disease in the general Danish population.
MG is the guarantor.
Funding: This study was supported by grants from the Health Insurance
Foundation, the Ministry of the Interior and Health, the Danish Cancer
Society, and the Danish National Board of Health.
Competing interests: None declared.
Ethical approval: This study was approved by the ethical committees for the
Copenhagen and Aarhus municipalities (KF 01-116/96).
1
2
3
4
Flesch M, Rosenkranz S, Erdmann E, Bohm M. Alcohol and the risk of myocardial infarction. Basic Res Cardiol 2001;96:128-35.
Malyutina S, Bobak M, Kurilovitch S, Gafarov V, Simonova G, Nikitin Y, et al. Relation
between heavy and binge drinking and all-cause and cardiovascular mortality in Novosibirsk, Russia: a prospective cohort study. Lancet 2002;360:1448-54.
McElduff P, Dobson AJ. How much alcohol and how often? Population based
case-control study of alcohol consumption and risk of a major coronary event. BMJ
1997;314:1159-64.
Trevisan M, Dorn J, Falkner K, Russell M, Ram M, Muti P, et al. Drinking pattern and
risk of non-fatal myocardial infarction: a population-based case-control study. Addiction
2004;99:313-22.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
What is already known on this topic
Alcohol intake is inversely associated with risk of coronary
heart disease
In men, for the same weekly amount of alcohol intake,
frequent drinkers have a lower risk of coronary heart
disease than less frequent drinkers
Little is known about drinking pattern and the risk of
coronary heart disease among women
28
Murray RP, Connett JE, Tyas SL, Bond R, Ekuma O, Silversides CK, et al. Alcohol volume, drinking pattern, and cardiovascular disease morbidity and mortality: is there a
U-shaped function? Am J Epidemiol 2002;155:242-8.
Mukamal KJ, Conigrave KM, Mittleman MA, Camargo CA Jr, Stampfer MJ, Willett WC,
et al. Roles of drinking pattern and type of alcohol consumed in coronary heart disease
in men. N Engl J Med 2003;348:109-18.
Mumenthaler MS, Taylor JL, O’Hara R, Yesavage JA. Gender differences in moderate
drinking effects. Alcohol Res Health 1999;23:55-64.
Register TC, Cline JM, Shively CA. Health issues in postmenopausal women who drink.
Alcohol Res Health 2002;26:299-307.
Tjonneland A, Gronbaek M, Stripp C, Overvad K. Wine intake and diet in a random
sample of 48 763 Danish men and women. Am J Clin Nutr 1999;69:49-54.
Overvad K, Tjonneland A, Haraldsdottir J, Ewertz M, Jensen OM. Development of a
semiquantitative food frequency questionnaire to assess food, energy and nutrient
intake in Denmark. Int J Epidemiol 1991;20:900-5.
Tjonneland A, Overvad K, Haraldsdottir J, Bang S, Ewertz M, Jensen OM. Validation of
a semiquantitative food frequency questionnaire developed in Denmark. Int J Epidemiol
1991;20:906-12.
Jürgensen HJ, Frølund C, Gustafsen J, Mosbech H, Guldhammer B, Mosbech J. Registration of diagnoses in the Danish National Registry of patients. Methods Inform Med
1986;25:158-64.
Juel K, Helweg-Larsen K. The Danish registers of causes of death. Dan Med Bull
1999;46:354-7.
Shaper AG. Mortality and alcohol consumption. Non-drinkers shouldn’t be used as
baseline. BMJ 1995;310:325.
Bien TH, Burge R. Smoking and drinking: a review of the literature. Int J Addict
1990;25:1429-54.
Gronbaek M, Becker U, Johansen D, Gottschau A, Schnohr P, Hein HO, et al. Type of
alcohol consumed and mortality from all causes, coronary heart disease, and cancer.
Ann Intern Med 2000;133:411-9.
Gronbaek M, Tjonneland A, Johansen D, Stripp C, Overvad K. Type of alcohol and
drinking pattern in 56,970 Danish men and women. Eur J Clin Nutr 2000;54:174-6.
Colditz GA, Willett WC, Stampfer MJ, Rosner B, Speizer FE, Hennekens CH.
Menopause and the risk of coronary heart disease in women. N Engl J Med
1987;316:1105-10.
Agarwal DP. Cardioprotective effects of light-moderate consumption of alcohol: a
review of putative mechanisms. Alcohol Alcohol 2002;37:409-15.
Puddey IB, Rakic V, Dimmitt SB, Beilin LJ. Influence of pattern of drinking on cardiovascular disease and cardiovascular risk factors—a review. Addiction 1999;94:649-63.
Wannamethee G, Shaper AG. Alcohol intake and variations in blood pressure by day
of examination. J Hum Hypertens 1991;5:59-67.
Marmot MG, Elliott P, Shipley MJ, Dyer AR, Ueshima H, Beevers DG, et al. Alcohol and
blood pressure: the INTERSALT study. BMJ 1994;308:1263-7.
Rakic V, Puddey IB, Burke V, Dimmitt SB, Beilin LJ. Influence of pattern of alcohol
intake on blood pressure in regular drinkers: a controlled trial. J Hypertens
1998;16:165-74.
Gruchow HW, Hoffmann RG, Anderson AJ, Barboriak JJ. Effects of drinking patterns
on the relationship between alcohol and coronary occlusion. Atherosclerosis
1982;43:393-404.
Hojnacki JL, Deschenes RN, Cluette-Brown JE, Mulligan JJ, Osmolski TV, Rencricca NJ,
et al. Effect of drinking pattern on plasma lipoproteins and body weight. Atherosclerosis
1991;88:49-59.
Rakic V, Puddey IB, Dimmitt SB, Burke V, Beilin LJ. A controlled trial of the effects of
pattern of alcohol intake on serum lipid levels in regular drinkers. Atherosclerosis
1998;137:243-52.
Dimmitt SB, Rakic V, Puddey IB, Baker R, Oostryck R, Adams MJ, et al. The effects of
alcohol on coagulation and fibrinolytic factors: a controlled trial. Blood Coagul
Fibrinolysis 1998;9:39-45.
Fuchs CS, Stampfer MJ, Colditz GA, Giovannucci EL, Manson JE, Kawachi I, et al. Alcohol consumption and mortality among women. N Engl J Med 1995;332:1245-50.
(Accepted 7 March 2006)
doi 10.1136/bmj.38831.503113.7C
Centre for Alcohol Research, National Institute of Public Health, Øster
Farimagsgade 5, Dk-1399 Copenhagen, Denmark
Janne Tolstrup research fellow
Majken K Jensen research fellow
Morten Grønbæk professor
What this study adds
Institute of Cancer Epidemiology, Danish Cancer Society, Copenhagen
Anne Tjønneland senior research fellow
Intake may be more important than frequency for the
inverse association between alcohol drinking and risk of
coronary heart disease among women
Department of Clinical Epidemiology, Cardiovascular Research Center, Aarhus
University Hospital, Aalborg
Kim Overvad research director
In men, frequency is more important than alcohol intake
Division of General Medicine and Primary Care, Beth Israel Deaconess Medical
Center, Boston, USA
Kenneth J Mukamal assistant professor
Correspondence to: J Tolstrup [email protected]
BMJ Online First bmj.com
page 5 of 5
Alcohol Intake, Myocardial Infarction, Biochemical Risk Factors, and Alcohol
Dehydrogenase Genotypes
Janne S. Tolstrup, Morten Grønbæk and Børge G. Nordestgaard
Circ Cardiovasc Genet 2009;2;507-514; originally published online Aug 22, 2009;
DOI: 10.1161/CIRCGENETICS.109.873604
Circulation: Cardiovascular Genetics is published by the American Heart Association. 7272 Greenville
Avenue, Dallas, TX 72514
Copyright © 2009 American Heart Association. All rights reserved. Print ISSN: 1942-325X. Online
ISSN: 1942-3268
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circgenetics.ahajournals.org/cgi/content/full/2/5/507
Subscriptions: Information about subscribing to Circulation: Cardiovascular Genetics is online at
http://circgenetics.ahajournals.org/subscriptions/
Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters
Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax:
410-528-8550. E-mail:
[email protected]
Reprints: Information about reprints can be found online at
http://www.lww.com/reprints
Downloaded from circgenetics.ahajournals.org by on October 21, 2009
Alcohol Intake, Myocardial Infarction, Biochemical Risk
Factors, and Alcohol Dehydrogenase Genotypes
Janne S. Tolstrup, MS, PhD; Morten Grønbæk, MD, DrMedSci; Børge G. Nordestgaard, MD, DrMedSci
Background—The risk of myocardial infarction is lower among light-to-moderate alcohol drinkers compared with
abstainers. We tested associations between alcohol intake and risk of myocardial infarction and risk factors and whether
these associations are modified by variations in alcohol dehydrogenases.
Methods and Results—We used information on 9584 men and women from the Danish general population in the
Copenhagen City Heart Study. During follow-up, from 1991 to 2007, 663 incident cases of myocardial infarction
occurred. We observed that increasing alcohol intake was associated with decreasing risk of myocardial infarction,
decreasing low-density lipoprotein cholesterol and fibrinogen, increasing diastolic and systolic blood pressure and
high-density lipoprotein cholesterol, and with U-shaped nonfasting triglycerides. In contrast, ADH1B and ADH1C
genotypes were not associated with risk of myocardial infarction or with any of the cardiovascular biochemical risk
factors, and there was no indication that associations between alcohol intake and myocardial infarction and between
alcohol intake and risk factors were modified by genotypes.
Conclusions—Increasing alcohol intake is associated with decreasing risk of myocardial infarction, decreasing low-density
lipoprotein cholesterol and fibrinogen, increasing diastolic and systolic blood pressure and high-density lipoprotein
cholesterol, and U-shaped nonfasting triglycerides. These associations were not modified by ADH1B and ADH1C are
genotypes. (Circ Cardiovasc Genet. 2009;2:507-514.)
Key Words: blood pressure 䡲 genetics 䡲 cardiovascular diseases 䡲 myocardial infarction 䡲 epidemiology 䡲 alcohol
䡲 cardiovascular risk factors 䡲 alcohol dehydrogenase genes
S
ubstantial epidemiological evidence suggests that alcohol
has beneficial effects on the cardiovascular system.1–5
First, a lower risk of coronary heart disease among light-tomoderate alcohol consumers compared with abstainers is
consistently observed in different populations.6 Second, plausible mediating factors such as increased high-density lipoprotein (HDL) cholesterol levels and reduced low-density
lipoprotein (LDL) cholesterol and plasma fibrinogen levels
have also been identified.7,8 In recent research, focus has been
on identifying possible genetic and environmental modifiers of
the association between alcohol and coronary heart disease.
Specifically, it has been suggested that the effect of alcohol on
the risk of coronary heart disease depends on variations in genes
coding for alcohol-degrading enzymes: individuals carrying
genotypes that code for slow alcohol degradation has persistently higher blood alcohol concentrations that leads to a lower
risk of coronary heart disease compared with individuals with
genotypes that code for fast alcohol degradation.
Clinical Perspective on p 514
Alcohol degradation is mainly catalyzed by different alcohol dehydrogenases (ADHs). In vitro studies have shown that
at ADH1B, alleles ADH1B 䡠 2 and ADH1B 䡠 1 produce enzymes with a 38-fold difference in alcohol degradation rate,
and at ADH1C, alleles ADH1C 䡠 1 and ADH1C 䡠 2 produce
enzymes with a 2.5-fold difference.9 However, the size of in
vivo effects of these variations is much more modest and not
even consistently observeable.10 –15 In whites, the frequency
of the most active alleles (ADH1B 䡠 2 and ADH1C 䡠 1) are
⬇2% and 58%.16
Although the idea that the risk reduction of coronary heart
disease in low-to-moderate drinkers depends on the genetic
capacity for alcohol degradation is both appealing and plausible, the evidence for this modification is sparse and inconsistent. In the first study on the subject, which was conducted
among American physicians, ADH1C slow metabolizers
seemingly had a lower risk of myocardial infarction (MI) than
ADH1C fast metabolizers, but this was confined to individuals in the highest drinking category17; however, this is in
contrast to findings from another study, where this effect was
only observed in men with a very low intake (less than 3
drinks per week).18 In some other studies, no significant
interaction between the ADH1C genotype and alcohol on risk
of coronary heart disease was found.19 –21 The potential
Received April 16, 2009; accepted July 27, 2009.
From the Center for Alcohol Research (J.S.T., M.G.), National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark;
The Copenhagen City Heart Study (M.G., B.G.N.), Bispebjerg Hospital, Copenhagen University Hospital, Copenhagen, Denmark; and Department of
Clinical Biochemistry (J.S.T., B.G.N.), Herlev Hospital, Copenhagen University Hospital, Herlev, Denmark.
Correspondence to Janne S. Tolstrup, Center for Alcohol Research, National Institute of Public Health, Øster Farimagsgade 5a, DK-1399 Copenhagen,
Denmark. E-mail [email protected]
© 2009 American Heart Association, Inc.
Circ Cardiovasc Genet is available at http://circgenetics.ahajournals.org
DOI: 10.1161/CIRCGENETICS.109.873604
507
Downloaded from circgenetics.ahajournals.org
by on October 21, 2009
508
Circ Cardiovasc Genet
October 2009
interaction between ADH1B genotypes and alcohol has not
been addressed among whites, most likely because the
ADH1B 䡠 2 allele is relatively rare in this population.
In this study, we aim at testing the association between
alcohol intake and risk of MI, and whether this association is
modified by ADH1B and ADH1C genotypes. We also tested
the association between alcohol intake and cardiovascular
biochemical risk factors, such as blood pressure, HDL cholesterol, LDL cholesterol, triglycerides, and fibrinogen, and
whether these associations differ according to genotype.
Methods
Study Population
Our data originates from the Copenhagen City Heart Study, which is
a series of studies conducted in the Danish general population.
Examinations consisted of interview, physical examination, and
more especially, blood was given for DNA purification at the
examination that was performed in 1991–1994. Enrollment and
examination procedures have been described in more detail elsewhere.22,23 All participants gave informed consent, and the ethics
committee for Copenhagen and Frederiksberg approved the study
(100.2039/91).
Questionnaire Measures
Amount of usual alcohol intake was reported as weekly consumption
of beer (in bottles), wine (in glasses), and spirits (in units). Assuming
1 drink to be equal to 12 g of pure alcohol, a measure of total weekly
alcohol intake was calculated. Smoking was reported as status
(never, former, or current) and amount of smoking (in number of
daily cigarettes, cheroots, cigars, and pipes). Assuming 1 cigarette to
be equivalent to 1 g of tobacco, 1 cheroot or 1 cigar to be equivalent
to 3 g of tobacco, and 1 pipe to be equivalent to 5 g of tobacco, total
amount of daily smoking was calculated. School education was
reported as number of years of basic schooling and categorized as
⬍8, 8 to 11, and ⬎11 years of education, corresponding to lower
primary school, higher primary school, and secondary school.
Participants were classified as having diabetes, hypertension, or
hypercholesterolemia if they reported a physician-made diagnosis.
Familial predisposition to cardiovascular disease was defined as
having 1 or 2 affected parents before the age 60 years. Among
women, we used information on menstruations and hormone replacement therapy to define their menopausal status in categories of pre and
post-menopausal with and without hormone replacement therapy.
Clinical and Laboratory Measures
Arterial blood pressure was measured in the left arm with the
participant in the sitting position after 5 minutes of rest. Study staff
obtained blood samples and measured height and weight. Total
plasma cholesterol, plasma HDL cholesterol, plasma nonfasting
triglycerides, and plasma fibrinogen were measured using standard
hospital assays (Boehringer Mannheim) subjected to daily internal
quality control assessing assay precision and monthly external
quality control assessing assay accuracy.
The ADH1B 䡠 2 allele (rs1229984, Arg47His in exon 3) and
ADH1C 䡠 2 allele (rs698, Ile349Val in exon 8) were identified by
means of duplex polymerase chain reaction followed by Nanogen
microelectronic chip technology (Nanogen NMW 1000 Nanochip
Molecular Biology Workstation24) using standard conditions (details
available from authors). In a validation study, the accuracy of the
Nanogen method was found to be comparable with restriction
fragment length polymorphism.25
Assessment of MI and Vital Status
Information on MI was obtained from the Danish Patient Registry26
and the Danish Causes of Death Registry,27 where all somatic
hospitalizations and causes of death are registered for every citizen
in the entire country. Diagnoses are classified according to the World
Health Organization’s International Classification of Diseases
(ICD), 8th and 10th revisions. The relevant diagnose codes were
ICD-8 410 and ICD-10 I21 and I22. Participants who had revascularization procedures (and no MI) during follow-up were treated as
unaffected in the analysis ie, their risk time was not censored, and
they did not count as cases (unless they had a later MI event). Most
cases of MI deaths are registered in both registries mentioned earlier,
ie, such patients have also been hospitalized with a valid MI
diagnosis based on characteristic symptoms, electrocardiographic
changes and/or elevated cardiac enzymes. Information on ⬇8% of
the MI death cases is solely contributed from the Danish Causes of
Death Registry. These are likely cases that are found dead in their
apartment. In such cases, the MI diagnosis may be less valid, but
because these cases constitute the minority of MI cases, we do not
expect this to have a major influence on our results.
Vital status of the participants was obtained from the Danish Civil
Registration System, where information on address and vital status
are registered for every Danish citizen.
Statistical Analysis
Of the 17 180 individuals who were invited to the 1991–1994
examination, 10 135 participated (59%). Participants of Asian or
black descent (n⫽161), missing questionnaire data (n⫽74), and
acute MI before baseline (316) were excluded, leaving 9584 individuals. Of these, 8777 had given blood and 8740 were successfully
genotyped for ADH1B and ADH1C. In all analyses, ADH1B 䡠 1/2
was combined with ADH1B 䡠 2/2 because of the low number in the
latter group (n⫽5).
We used the ␹2 test to determine whether the ADH1B and ADH1C
genotypes were in Hardy-Weinberg equilibrium.28 Haplotype frequencies for calculation of linkage disequilibrium coefficients were
estimated by HPlus.29,30 Linkage disequilibrium coefficients Lewontin’s D’ was 0.90 and the correlation coefficient r2 was 0.01
between ADH1B 䡠 2 and ADH1C 䡠 1, both coding for the fast alcohol
degradation enzymatic forms.31,32
Risk estimates for acute MI during follow-up were computed by
means of Cox proportional hazard regression models. Age was used
as the time axis to ensure that the estimation procedure was based on
comparisons of individuals at the same age and hence remove
confounding by age. The observation time for each participant was
the period from the Copenhagen City Heart Study 1991–1994
examination, until date of MI, death from other causes, emigration
outside Denmark, or August 1, 2007, whichever came first. We had
follow-up information on all participants.
Test for linear trend was performed by treating the median within
alcohol categories as a continuous variable and tests for interaction
between genotypes and alcohol were performed by comparing a
model including main effects of alcohol and ADH1C genotypes with
a model also including the interaction terms by a log likelihood test.
Associations between alcohol intake, ADH1B and ADH1C genotypes and blood pressure, HDL, LDL, triglyceride, and fibrinogen
were investigated by general linear models (all log-transformed to
approximate normal distributions). Tests for effects of a low-tomoderate alcohol intake (⬍14 drinks/wk for women and ⬍21
drinks/wk for men) on blood pressure, HDL, LDL, triglyceride, and
fibrinogen levels were performed by modeling alcohol intake with
linear splines with a knot placed at 14 and 21 drinks per week for
women and men, respectively, and testing for statistical significance
for each portion of alcohol intake. Furthermore, statistically significant U-shaped effects for triglycerides were tested by including the
alcohol intake as a linear and a squared term. P values less than 0.05
were considered statistically significant.
Finally, we calculated posthoc power using the Quanto program,33
by assuming effect sizes for the ADH1C-alcohol interaction as in the
study by Hines et al,17 and by using information on distribution on
alcohol intake and genotype and number of MI cases from our study.
Assuming a significance level of 0.05, we had 97% power to pick up
an ADH1C-alcohol interaction of a size similar to what was found in
the Hines study.
Downloaded from circgenetics.ahajournals.org by on October 21, 2009
Tolstrup et al
Alcohol Intake, Risk Factors, and Myocardial Infarction
Table 1. Baseline Characteristics of 5479 Women and 4105
Men Who Participated in the Copenhagen City Heart Study
1991–1994 Examination*
Characteristics
Women
and 10 drinks/wk among men (Table 1). Approximately half
of the participants were current smokers. Frequencies of
ADH1B 䡠 1/2⫹2/2 and of ADH1C 䡠 1/1 (genotypes coding for
the most active enzymes) were 4.5% and 34%. ADH1B and
the ADH1C genotypes were both in Hardy-Weinberg equilibrium (P⫽0.43 and 0.64, respectively).
Men
Behavioral and demographic
Age, y
62 (36, 76)
60 (35, 76)
Alcohol intake, drinks/wk†
3.0 (0, 15)
10 (0, 33)
School education ⱕ7 y, %
33
31
Current smokers, %
41
47
Amount of smoking, g/d‡
15 (5, 20)
20 (7, 30)
Low exercise intensity, %
13
13
Postmenopausal, %
73
Use of hormone replacement
therapy, %
16
Alcohol Intake/ADH1B/ADH1C/Genotypes, and
Risk of MI
Increasing amount of alcohol intake was associated with
decreasing risk of MI among men and women (Figure 1). In
contrast, ADH1B and ADH1C genotypes were not associated
with risk of MI in men or women (Table 2).
The association between alcohol intake and risk of MI was
not modified by ADH1C genotype, because hazard ratios
were comparable within strata of ADH1C genotype (Figure
2); the probability value for interaction between ADH1C
genotype and alcohol intake was 0.56 in a model combining
men and women. Similar results were obtained when analyzing each sex separately. We did not have sufficient statistical
power to perform similar analysis for the ADH1B genotype
because the ADH1B 䡠 2 allele is relatively rare.
Clinical and physiological
Body mass index, kg/m2
24.4 (20.1, 31.4)
25.7 (21.6, 31.2)
Diabetes, %
2.5
4.5
Hypertension, %
12.5
9.9
Hypercholesterolemia, %
10.5
5.9
Laboratory
ADH1B 䡠 1/2⫹2/2 (fast), %
4.3
4.6
ADH1B 䡠 1/1 (slow), %
95.7
95.4
ADH1C 䡠 1/1 (fast), %
34
34
ADH1C 䡠 2/1 (intermediate), %
48
48
ADH1C 䡠 2/2 (slow), %
Plasma total cholesterol, mmol/L§
18
18
4.7 (6.2, 8.1)
4.5 (5.9, 7.5)
Alcohol Intake/ADH1B/ADH1C/Genotypes, and
Cardiovascular Biochemical Risk Factors
Among women, alcohol was statistically significantly associated with increasing levels of diastolic and systolic blood
pressure and HDL cholesterol, and with decreasing levels of
LDL cholesterol and fibrinogen with no apparent threshold
effect (Figure 3). Also among women, the association between alcohol intake and nonfasting triglycerides was significantly U-shaped.
Among men, alcohol was statistically significantly associated with increasing levels of diastolic and systolic blood
pressure but only for alcohol intake of 21 or more drinks/wk.
Also among men, alcohol was associated with decreasing
levels of LDL cholesterol, but again, only for alcohol intake
of 21 or more drinks/wk. There appeared to be a U-shaped
association between alcohol intake and nonfasting triglycer-
*Continuous characteristics are shown as median (10th, 90th percentile).
†One drink corresponds to 12 g of pure alcohol.
‡Median (10th, 90th percentile), among current smokers only.
§Nonfasting.
Results
Baseline Characteristics
Of the 9584 participants, 4105 (57%) were women, and the
median age of participants was 61 years (10 –90%, 35 to 77).
The median alcohol intake was 3.0 drinks/wk among women
Men
2
2
Women
p for trend=0.007
.25
.25
.5
.5
1
1
p for trend=0.01
Hazard ratio
509
<1
1-6
7-13
14+
<1
1-6
7-13
14-20
21+
90
11601
53
6878
72
10883
Alcohol intake (drinks/week)
Cases
Person-years
136
18077
106
25969
45
26
13596 8496
60
5065
95
11325
Figure 1. Hazard ratio of MI according to weekly alcohol intake. The analyses were adjusted for school education, smoking, physical
activity, body mass index, diabetes, hypertension, and hypercholesterolemia. Analyses among women also adjusted for postmenopausal status and use of hormone replacement therapy. Vertical bars indicate 95% CIs for the comparison with participants who drank
⬍1 drink per week.
Downloaded from circgenetics.ahajournals.org by on October 21, 2009
510
Circ Cardiovasc Genet
October 2009
Table 2. Hazard Ratios With 95% CIs of 4925 Women and 3815 Men Who Participated in the Copenhagen
City Heart Study 1991–1994 Examination According to ADH1B and ADH1C Genotypes
ADH1B
ADH1C
1/2⫹2/2
1/1
1/1
1/2
2/2
Fast
Slow
Fast
Intermediate
Slow
Relative enzyme activity
Women
Cases
12
257
83
129
57
2600
58 174
20 639
29 236
10 899
Age adjusted
1 (referent)
0.94 (0.52 to 1.68)
1 (referent)
1.08 (0.82 to 1.43)
1.31 (0.93 to 1.84)
Multivariable adjusted*
1 (referent)
0.89 (0.49 to 1.59)
1 (referent)
1.07 (0.81 to 1.41)
1.35 (0.96 to 1.89)
Multivariable, with adjustment
for alcohol intake
1 (referent)
0.91 (0.51 to 1.64)
1 (referent)
1.08 (0.82 to 1.43)
1.36 (0.97 to 1.91)
Person-years
Hazard ratio (95% CI)
Men
Cases
10
348
118
187
53
2050
41 162
14 896
20 865
7452
Age adjusted
1 (referent)
1.82 (0.97 to 3.42)
1 (referent)
1.12 (0.89 to 1.41)
0.88 (0.64 to 1.22)
Multivariable adjusted*
1 (referent)
1.76 (0.93 to 3.33)
1 (referent)
1.08 (0.86 to 1.37)
0.90 (0.65 to 1.25)
Multivariable, with adjustment
for alcohol intake
1 (referent)
1.85 (0.98 to 3.50)
1 (referent)
1.09 (0.87 to 1.38)
0.90 (0.65 to 1.25)
Person-years
Hazard ratio (95% CI)
*Multivariable adjusted models included ADH1B/ADH1C genotypes, school education, smoking, physical activity, body mass index,
diabetes, hypertension, and hypercholesterolemia. Analyses among women also included postmenopausal and use of hormone
replacement therapy.
ides, but it was not statistically significant (P⫽0.23). Alcohol
was statistically significantly associated with increasing levels of HDL cholesterol and with decreasing levels of fibrinogen with no apparent threshold effect among men.
ADH1B and ADH1C genotypes were not consistently
associated with any of these cardiovascular risk factors
among women or men (Table 3). There was no sign of
interaction between ADH1C and alcohol intake on any of the
cardiovascular biochemical risk factors (all P values ⬎0.05).
Discussion
ADH1C·1/1 (fast)
1
ADH1C·2/1 (intermediate)
.5
ADH1C·2/2 (slow)
.25
Hazard ratio
2
In this general population study, we found that increasing
alcohol intake was associated with a lower risk of MI and that
amount of weekly alcohol from none to low through moderate and excessive intake was associated with stepwise increasing levels of diastolic and systolic blood pressure and
HDL cholesterol, with stepwise decreasing levels of LDL
cholesterol and fibrinogen, and with a U-shaped curve for
nonfasting triglycerides. In contrast, ADH1B and ADH1C
genotypes were not associated with risk of MI or with
cardiovascular risk factors, and there was no indication that
associations between alcohol intake and MI or alcohol intake
and cardiovascular risk factors were modified by genotype.
A moderate alcohol intake is consistently shown to be
associated with a decrease in risk of coronary heart disease.4
The maximal benefit seem to be obtained at ⬇1 to 2 drinks
per day for women and 2 to 3 drinks per day for men; for
higher amounts of alcohol intake, there seem to be no further
gain and some studies have even reported an increase in risk,
indicating a J-shaped association between alcohol intake and
coronary heart disease.4
The beneficial effect of alcohol is primarily thought to be
mediated through an increase in HDL and a decrease in
fibrinogen.7 For these 2 cardiovascular risk factors, we
observed that increasing alcohol intake was associated with
increasing HDL cholesterol and decreasing fibrinogen with
no threshold, ie, even low amounts of alcohol intake was
associated with increasing HDL and decreasing fibrinogen.
1-13
<1
14+
Alcohol intake (drinks/week)
Cases
48 90 37
109 152 45
44 74 28
Figure 2. Hazard ratio of MI according to ADH1C
genotype and weekly alcohol intake in men and
women combined. The analyses were adjusted for
sex, ADH1B genotoype, school education, smoking, physical activity, body mass index, diabetes,
hypertension, hypercholesterolemia, postmenopausal status, and use of hormone replacement
therapy. Vertical bars indicate 95% CIs for the
comparison with participants with the ADH1C 䡠 1/1
genotype who drank ⬍1 drink per week. The probability value was 0.56 for interaction between
ADH1C 䡠 1/1 and alcohol intake (estimated in
nested log-likelihood test).
Downloaded from circgenetics.ahajournals.org by on October 21, 2009
Tolstrup et al
p<0.0001*
Men
(n=4,105)
p<0.0001†
85
80
4.0
2.4
p<0.0001*
p<0.0001†
3.8
3.6
3.4
140
511
Men
(n=4,105)
p<0.0001*
p<0.0001†
135
130
125
120
2.0
Triglycerides
(mmol/L)
LDL Cholesterol
(mmol/L)
75
Women
(n=5,479)
Systolic Blood Pressure
(mmHg)
90
HDL Cholesterol
(mmol/L)
p=0.006‡
p=0.23‡
1.8
1.6
1.4
1.2
3.2
p<0.0001*
p<0.0001*
3.0
Fibrinogen
(g/L)
Diastolic Blood Pressure
(mmHg)
Women
(n=5,479)
Alcohol Intake, Risk Factors, and Myocardial Infarction
2.1
1.8
1.5
p<0.0001*
p<0.0001*
2.8
2.6
2.4
2.2
2.0
1.2
35+ (8%)
28-35 (6%)
21-27 (8%)
14-20 (15%)
7-13 (25%)
1-6 (25 %)
<1 (13%)
28+ (2%)
21-27 (3%)
14-20 (7%)
7-13 (19%)
1-6 (38 %)
<1 (31%)
35+ (8%)
28-35 (6%)
21-27 (8%)
14-20 (15%)
7-13 (25%)
1-6 (25 %)
<1 (13%)
28+ (2%)
21-27 (3%)
14-20 (7%)
7-13 (19%)
1-6 (38 %)
<1 (31%)
Alcohol intake (drinks/week)
Alcohol intake (drinks/week)
Figure 3. Average levels of diastolic blood pressure, systolic blood pressure, LDL cholesterol, triglycerides, HDL cholesterol, and fibrinogen according to alcohol intake. The percent of all women and men in each alcohol drinking category is shown in parenthesis after
the amount of drinks per week. *Probability value for linear trend. †Probability value for trend for alcohol intake more than 21 drinks per
week for men. ‡Probability value for U-shape.
These findings are consistent with the above mentioned
effects of alcohol intake in the low-to-moderate range on the
risk of coronary heart disease. We also observed that alcohol
intake at higher levels (⬎14/21 drinks/wk for women/men)
was associated with higher diastolic and systolic blood
pressure and with increased level of nonfasting triglycerides,
which on the other hand is associated with increased cardiovascular risk34,35 and in support of a J-shaped association
between alcohol intake and risk of coronary heart disease.
If the effect of alcohol on coronary heart disease is
modified by variations in ADH genes it really comes down to
the question of whether the risk is lower among ADH1C 䡠 2/2
(slow metabolizers) than among ADH1C 䡠 1/1 (fast metabolizers) among light-to-moderate drinkers but similar among
nondrinkers. Initially, results in American physicians by
Hines et al17 indicated that this is so (interaction, P⫽0.01;
Table 4). However, the relative risk of 0.14 (95% CI, 0.04 to
0.45) was based on only 5 cases and 37 controls with the
ADH1C 䡠 2/2 genotype (slow metabolizers) in the highest
alcohol category drinking (ⱖ8 drinks/wk); for lower alcohol
intake there seemed to be no difference in risk of MI
according to genotype. Another study also found a significant
interaction (P⫽0.02) between alcohol intake and the ADH1C
genotype, but only after performing post hoc regrouping of
the alcohol categories, and in contrast to the previous report,17
the lower risk was found among individuals with the lowest
alcohol intake (0.7 to 2 drinks/wk).18 In yet other studies,
there was no significant interaction between alcohol intake
and the ADH1C genotype (Table 4).19 –21 Finally, our study
which is the largest so far could not confirm either of the
previous findings; our post hoc power calculations showed
that we had 97% power to pick up a ADH1C-alcohol
interaction of a size similar to what was found in the Hines
study.17 This further supports the overall conclusion that the
association between alcohol intake and risk of coronary heart
disease is not modulated by the ADH1C genotype.
Previously, we showed that ADH1B and ADH1C genotypes were associated with amount of alcohol intake and with
risk of heavy drinking (participants with genotypes coding for
slow alcohol degradation were drinking more and had higher
risk of heavy drinking; for ADH1C, however, these effects
were only modest). Hence, individuals with genotypes coding
for slow alcohol degradation have higher blood alcohol
concentrations due to both the lower activity of the resulting
enzyme and to a higher alcohol intake. These effects both
stem directly from the genotype and should therefore not be
separated.
Downloaded from circgenetics.ahajournals.org by on October 21, 2009
512
Circ Cardiovasc Genet
October 2009
Table 3. Estimated Differences in Diastolic Blood Pressure, Systolic Blood
Pressure, LDL Cholesterol, Triglycerides, HDL Cholesterol, and Fibrinogen
According to ADH1B and ADH1C Genotypes*
ADH1B
Relative enzyme activity
ADH1C
1/1
1/2
2/2
Slow
Intermediate
Slow
Women
Diastolic blood pressure, mm Hg
⫺0.06 (0.94)
0.31 (0.40)
1.08 (0.03)
Systolic blood pressure, mm Hg
2.48 (0.05)
0.72 (0.20)
1.96 (0.01)
⫺0.01 (0.67)
LDL cholesterol, mmol/L
⫺0.07 (0.30)
0.01 (0.57)
HDL cholesterol, mmol/L
0.02 (0.58)
0.01 (0.84)
0.01 (0.92)
⫺0.05 (0.27)
⫺0.01 (0.58)
⫺0.01 (0.76)
0.01 (0.94)
0.01 (0.93)
0.00 (0.99)
Diastolic blood pressure, mm Hg
1.23 (0.18)
0.31 (0.46)
⫺0.50 (0.35)
Systolic blood pressure, mm Hg
2.82 (0.04)
0.72 (0.25)
0.37 (0.65)
LDL cholesterol, mmol/L
0.10 (0.22)
0.03 (0.33)
0.04 (0.34)
⫺0.01 (0.50)
Triglycerides, mmol/L
Fibrinogen, g/L
Men
HDL cholesterol, mmol/L
Triglycerides, mmol/L
0.04 (0.28)
0.00 (0.95)
⫺0.11 (0.10)
0.07 (0.05)
0.07 (0.11)
0.05 (0.28)
0.00 (0.96)
⫺0.03 (0.47)
Fibrinogen, g/L
*Numbers represent estimated differences between the ADH1B 䡠 1/1 and ADH1B 䡠 1/2⫹2/2, and
ADH1C 䡠 1/2 and ADH1C 䡠 1/1, and ADH1C 䡠 2/2 and ADH1C 䡠 1/1, respectively. Numbers in parentheses represent P values.
A limitation of our results is that never-drinkers could not
be separated from ex-drinkers and the nondrinking category
may therefore contain some former alcoholics who due to
their former heavy drinking have preexisting illness. This
could lead to an apparent inverse association between alcohol
intake and MI. However, analyses of alcohol intake and
cardiovascular biochemical risk factors (diastolic and systolic
blood pressure, HDL and LDL cholesterol, nonfasting triTable 4.
glycerides, and fibrinogen) showed that the level of the
respective risk factor in the nondrinking category were in
accordance with an overall dose-response shaped curved
between alcohol intake and the cardiovascular risk factor,
indicating that accumulation of former heavy drinkers in this
category is not a major problem.
Limitations further include that information on alcohol
intake was obtained by self-report and has not been validated.
Summary Findings from Studies of ADH1C Genotype and Associations With Coronary Heart Disease*
ADH1C
Reference
Tolstrup et al†
Heidrich et al20
Younis et al‡
18
Djoussé et al19
Hines et al17
Categorization of Alcohol,
Drinks/wk
No. of Events
1/1
1/2
2/2
Interaction P value
⬍1
175
1
1.38 (0.97 to 1.96)
1.60 (1.04 to 2.47)
0.49
1 to 13
307
0.99 (0.70 to 1.40)
0.98 (0.71 to 1.37)
0.83 (0.55 to 1.25)
ⱖ14
146
0.80 (0.53 to 1.23)
0.82 (0.56 to 1.19)
0.88 (0.55 to 1.42)
ⱕ1
24
1
0.69 (0.31 to 1.55)
1 to 8
13
0.56 (0.19 to 1.61)
0.83 (0.34 to 2.07)
ⱖ8
35
1.06 (0.50 to 2.25)
0.36 (0.16 to 0.80)
⬍1
44
1
0.82 (0.47 to 1.45)
0.64 (0.24 to 1.68)
1 to 5
64
0.70 (0.40 to 1.22)
0.56 (0.32 to 0.99)
0.66 (0.31 to 1.38)
ⱖ5
102
0.57 (0.33 to 0.98)
0.77 (0.47 to 1.26)
0.68 (0.36 to 1.27)
0
56
1
0.67 (0.34 to 1.31)
1.11 (0.50 to 2.48)
⬎0
76
0.74 (0.41 to 1.32)
0.60 (0.34 to 1.08)
0.53 (0.24 to 1.18)
⬍1
117
1
1.01 (0.58 to 1.75)
0.59 (0.28 to 1.23)
1 to 8
191
1.11 (0.67 to 1.84)
0.66 (0.40 to 1.08)
1.02 (0.55 to 1.88)
ⱖ8
87
0.62 (0.34 to 1.13)
0.68 (0.40 to 1.15)
0.14 (0.04 to 0.45)
0.07
0.49
0.61
0.01
*Presented numbers are measures of relative risk (95% CIs) compared with ADH1C 䡠 1/1 nondrinkers.
†Results from the present article.
‡P for alcohol: ADH1C interaction was 0.02 after post hoc regrouping alcohol into 1 to 2 and ⱖ2 drinks/week (the latter risk estimates are not shown in the table).
Downloaded from circgenetics.ahajournals.org by on October 21, 2009
Tolstrup et al
Alcohol Intake, Risk Factors, and Myocardial Infarction
However, associations between increasing alcohol intake
and increasing levels of biomarkers of alcohol intake such
as alanine aminotransferase, aspartate aminotransferase,
and ␥-glutamyl transpeptidase has previously been observed within this cohort.36 Also, because we studied
whites only, our results may not necessarily apply to other
ethnic groups.
Our study had several strengths. First of all, sample size is
large and the wide range of alcohol intake provided statistical
power to study effects of low-to-moderate as well as excessive alcohol consumption on levels of cardiovascular risk
factors and risk of MI. Furthermore, participants were men
and women all from the general population of Danish
descent. Hence, population stratification is unlikely to have
affected our results. We had information on several different
cardiovascular risk factors, which were obtained objectively
from the study participants. Hence, it is unlikely that these
measures are differentially biased according to alcohol intake
or genotype.
In summary, we observed that increasing alcohol intake
associated with decreasing risk of MI, decreasing LDL
cholesterol and fibrinogen, increasing diastolic and systolic
blood pressure and HDL cholesterol, and with U-shaped
nonfasting triglycerides. These associations were not modified by ADH1B and ADH1C genotypes.
Acknowledgments
We thank to the participants in the Copenhagen City Heart Study for
their outstanding cooperation.
Sources of Funding
This research was supported in part by the Danish Graduate School
of Public Health, the Danish Heart Foundation, Chief Physician
Johan Boserup and Lise Boserup Foundation, the Health Insurance
Foundation, the Ministry of the Interior and Health, the Danish
Cancer Society, and the Danish National Board of Health.
Disclosures
None.
References
1. Collins MA, Neafsey EJ, Mukamal KJ, Gray MO, Parks DA, Das DK,
Korthuis RJ. Alcohol in moderation, cardioprotection, and neuroprotection: epidemiological considerations and mechanistic studies. Alcohol
Clin Exp Res. 2009;33:206 –219.
2. Mukamal KJ, Conigrave KM, Mittleman MA, Camargo CA Jr, Stampfer
MJ, Willett WC, Rimm EB. Roles of drinking pattern and type of alcohol
consumed in coronary heart disease in men. N Engl J Med. 2003;348:
109 –118.
3. Stampfer MJ, Colditz GA, Willett WC, Speizer FE, Hennekens CH. A
prospective study of moderate alcohol consumption and the risk of
coronary disease and stroke in women. N Engl J Med. 1988;319:267–273.
4. Corrao G, Rubbiati L, Bagnardi V, Zambon A, Poikolainen K. Alcohol
and coronary heart disease: a meta-analysis. Addiction. 2000;95:
1505–1523.
5. Gronbaek M, Becker U, Johansen D, Gottschau A, Schnohr P, Hein HO,
Jensen G, Sorensen TI. Type of alcohol consumed and mortality from all
causes, coronary heart disease, and cancer. Ann Int Med. 2000;133:
411– 419.
6. Flesch M, Rosenkranz S, Erdmann E, Bohm M. Alcohol and the risk of
myocardial infarction. Basic Res Cardiol. 2001;96:128 –135.
7. Rimm E, Moats C. Alcohol and coronary heart disease: drinking patterns
and mediators of effect. Ann Epi. 2007;17:S3–S7.
513
8. Agarwal DP. Cardioprotective effects of light-moderate consumption of
alcohol: a review of putative mechanisms. Alcohol Alcohol. 2002;37:
409 – 415.
9. Bosron WF, Li TK. Genetic polymorphism of human liver alcohol and
aldehyde dehydrogenases, and their relationship to alcohol metabolism
and alcoholism. Hepatology. 1986;6:502–510.
10. Mizoi Y, Yamamoto K, Ueno Y, Fukunaga T, Harada S. Involvement of
genetic polymorphism of alcohol and aldehyde dehydrogenases in individual variation of alcohol metabolism. Alcohol Alcohol. 1994;29:
707–710.
11. Neumark YD, Friedlander Y, Durst R, Leitersdorf E, Jaffe D, Ramchandani VA, O’Connor S, Carr LG, Li TK. Alcohol dehydrogenase
polymorphisms influence alcohol-elimination rates in a male Jewish population. Alcohol Clin Exp Res. 2004;28:10 –14.
12. Yokoyama T, Yokoyama A, Kato H, Tsujinaka T, Muto M, Omori T,
Haneda T, Kumagai Y, Igaki H, Yokoyama M, Watanabe H, Yoshimizu
H. Alcohol flushing, alcohol and aldehyde dehydrogenase genotypes, and
risk for esophageal squamous cell carcinoma in Japanese men. Cancer
Epidemiol Biomarkers Prev. 2003;12:1227–1233.
13. Carr LG, Foroud T, Stewart T, Castelluccio P, Edenberg HJ, Li TK.
Influence of ADH1B polymorphism on alcohol use and its subjective
effects in a Jewish population. Am J Med Genet. 2002;112:138 –143.
14. Takeshita T, Yang X, Morimoto K. Association of the ADH2 genotypes
with skin responses after ethanol exposure in Japanese male university
students. Alcohol Clin Exp Res. 2001;25:1264 –1269.
15. Takeshita T, Mao XQ, Morimoto K. The contribution of polymorphism in
the alcohol dehydrogenase beta subunit to alcohol sensitivity in a
Japanese population. Hum Genet. 1996;97:409 – 413.
16. Tolstrup JS, Nordestgaard BG, Rasmussen S, Tybjaerg-Hansen A,
Gronbaek M. Alcoholism and alcohol drinking habits predicted from
alcohol dehydrogenase genes. Pharmacogenomics J. 2007;8:220 –227.
17. Hines LM, Stampfer MJ, Ma J, Gaziano JM, Ridker PM, Hankinson SE,
Sacks F, Rimm EB, Hunter DJ. Genetic variation in alcohol dehydrogenase and the beneficial effect of moderate alcohol consumption on
myocardial infarction. N Engl J Med. 2001;344:549 –555.
18. Younis J, Cooper JA, Miller GJ, Humphries SE, Talmud PJ. Genetic
variation in alcohol dehydrogenase 1C and the beneficial effect of alcohol
intake on coronary heart disease risk in the Second Northwick Park Heart
Study. Atherosclerosis. 2005;180:225–232.
19. Djousse L, Levy D, Herbert AG, Wilson PW, D’Agostino RB, Cupples
LA, Karamohamed S, Ellison RC. Influence of alcohol dehydrogenase 1C
polymorphism on the alcohol-cardiovascular disease association. Am J
Cardiol. 2005;96:227–232.
20. Heidrich J, Wellmann J, Doring A, Illig T, Keil U. Alcohol consumption,
alcohol dehydrogenase and risk of coronary heart disease in the
MONICA/KORA-Augsburg cohort 1994/1995–2002. Eur J Cardiovasc
Prev Rehabil. 2007;14:769 –774.
21. Ebrahim S, Lawlor DA, Shlomo YB, Timpson N, Harbord R, Christensen
M, Baban J, Kiessling M, Day I, Gaunt T, Davey SG. Alcohol dehydrogenase type 1C (ADH1C) variants, alcohol consumption traits, HDL-cholesterol and risk of coronary heart disease in women and men: British Women’s
Heart and Health Study and Caerphilly cohorts. Atherosclerosis. 2008;196:
871– 878.
22. The Copenhagen City Heart Study. Osterbroundersogelsen. A book of
tables with data from the first examination (1976 –78) and a five year
follow-up (1981– 83). The Copenhagen City Heart Study Group. Scand J
Social Med (Suppl). 1989;41:1–160.
23. Schnohr P, Jensen G, Lange P, Scharling H, Appleyard M. The
Copenhagen City Heart Study—Østerbroundersøgelsen. Tables with data
from the third examination 1991–94. Eur Heart J (Suppl). 2001;3:
H1–H83.
24. Sosnowski RG, Tu E, Butler WF, O’Connell JP, Heller MJ. Rapid
determination of single base mismatch mutations in DNA hybrids by
direct electric field control. Proc Natl Acad Sci USA. 1997;94:
1119 –1123.
25. Sethi AA, Tybjaerg-Hansen A, Andersen RV, Nordestgaard BG.
Nanogen microelectronic chip for large-scale genotyping. Clin Chem.
2004;50:443– 446.
26. Jürgensen HJ, Frølund C, Gustafsen J, Mosbech H, Guldhammer B,
Mosbech J. Registration of diagnoses in the Danish National Registry of
patients. Meth Inf Med. 1986;25:158 –164.
27. Juel K, Helweg-Larsen K. The Danish registers of causes of death. Dan
Med Bul. 1999;46:354 –357.
28. Cavalli-Sforza L, Bodmer W. The Genetics of Human Populations. New
York: Dover; 1999.
Downloaded from circgenetics.ahajournals.org by on October 21, 2009
514
Circ Cardiovasc Genet
October 2009
29. HPlus. Fred Hutchinson Cancer Research Center, Seattle, WA;
2003.
30. Li SS, Khalid N, Carlson C, Zhao LP. Estimating haplotype frequencies and standard errors for multiple single nucleotide polymorphisms. Biostatistics. 2003;4:513–522.
31. Lewontin RC. The Interaction of selection and linkage I. General considerations: heterotic models. Genetics. 1964;49:49 – 67.
32. Pritchard JK, Przeworski M. Linkage disequilibrium in humans: models
and data. Am J Hum Genet. 2001;69:1–14.
33. QUANTO 1.1: A computer program for power and sample size calculations for genetic-epidemiology studies. 2007.
34. Lewington S, Clarke R, Qizilbash N, Peto R, Collins R. Age-specific
relevance of usual blood pressure to vascular mortality: a meta-analysis of
individual data for one million adults in 61 prospective studies. Lancet.
2002;360:1903–1913.
35. Nordestgaard BG, Benn M, Schnohr P, Tybjaerg-Hansen A. Nonfasting
triglycerides and risk of myocardial infarction, ischemic heart disease,
and death in men and women. JAMA. 2007;298:299 –308.
36. Tolstrup JS, Grønbæk M, Tybjaerg-Hansen A, Nordestgaard BG. Alcohol
intake, alcohol dehydrogenase genotypes and liver and pancreas damage
and disease in the general population. Am J Gastroenterol. 2009; [Epub
ahead of print].
CLINICAL PERSPECTIVE
Alcohol intake in moderation is associated with a lower risk of myocardial infarction. It has been suggested that this
cardioprotective effect of alcohol depends on alcohol’s effect on other cardiovascular risk factors and possibly on the
individual’s genetic capacity for degrading alcohol. In this large population-based study, we found that alcohol intake was
associated with levels of the cardiovascular risk factors low-density lipoprotein and high-density lipoprotein cholesterol,
fibrinogen, blood pressure, and nonfasting triglycerides. However, the associations between alcohol intake and risk of
myocardial infarction and between alcohol intake and cardiovascular risk factors were not modified by genetic variation
in alcohol dehydrogenases. Our results suggest that the protective effect of alcohol on the development of myocardial
infarction does not depend on genetic variations affecting enzymes involved in alcohol metabolism.
Downloaded from circgenetics.ahajournals.org by on October 21, 2009
Alcohol Intake and Risk of Coronary Heart Disease in Younger, Middle-Aged,
and Older Adults
Ulla A. Hvidtfeldt, Janne S. Tolstrup, Marianne U. Jakobsen, Berit L. Heitmann,
Morten Grønbæk, Eilis O'Reilly, Katarina Bälter, Uri Goldbourt, Göran Hallmans,
Paul Knekt, Simin Liu, Mark Pereira, Pirjo Pietinen, Donna Spiegelman, June
Stevens, Jarmo Virtamo, Walter C. Willett, Eric B. Rimm and Alberto Ascherio
Circulation published online Mar 29, 2010;
DOI: 10.1161/CIRCULATIONAHA.109.887513
Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX
72514
Copyright © 2010 American Heart Association. All rights reserved. Print ISSN: 0009-7322. Online
ISSN: 1524-4539
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org
Subscriptions: Information about subscribing to Circulation is online at
http://circ.ahajournals.org/subscriptions/
Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters
Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax:
410-528-8550. E-mail:
[email protected]
Reprints: Information about reprints can be found online at
http://www.lww.com/reprints
Downloaded from circ.ahajournals.org by on April 9, 2010
Epidemiology and Prevention
Alcohol Intake and Risk of Coronary Heart Disease in
Younger, Middle-Aged, and Older Adults
Ulla A. Hvidtfeldt, MSc; Janne S. Tolstrup, MSc, PhD; Marianne U. Jakobsen, MSc, PhD;
Berit L. Heitmann, PhD; Morten Grønbæk, MD, PhD; Eilis O’Reilly, ScD; Katarina Bälter, PhD;
Uri Goldbourt, PhD; Göran Hallmans, MD, PhD; Paul Knekt, PhD; Simin Liu, MD, ScD;
Mark Pereira, PhD; Pirjo Pietinen, ScD; Donna Spiegelman, ScD; June Stevens, MSc, PhD;
Jarmo Virtamo, MD; Walter C. Willett, MD, DrPH; Eric B. Rimm, ScD; Alberto Ascherio, MD, DrPH
Background—Light to moderate alcohol consumption is associated with a reduced risk of coronary heart disease. This
protective effect of alcohol, however, may be confined to middle-aged or older individuals. Coronary heart disease
incidence is low in men ⬍40 years of age and in women ⬍50 years of age; for this reason, study cohorts rarely have
the power to investigate the effects of alcohol on coronary heart disease risk in younger adults. This study examined
whether the beneficial effect of alcohol on coronary heart disease depends on age.
Methods and Results—In this pooled analysis of 8 prospective studies from North America and Europe including 192 067
women and 74 919 men free of cardiovascular diseases, diabetes, and cancers at baseline, average daily alcohol intake
was assessed at baseline with a food frequency or diet history questionnaire. An inverse association between alcohol and
risk of coronary heart disease was observed in all age groups; hazard ratios among moderately drinking men (5.0 to 29.9
g/d) 39 to 50, 50 to 59, and ⱖ60 years of age were 0.58 (95% confidence interval [CI], 0.36 to 0.93), 0.72 (95% CI,
0.60 to 0.86), and 0.85 (95% CI, 0.75 to 0.97) compared with abstainers. However, the analyses indicated a smaller
incidence rate difference between abstainers and moderate consumers in younger adults (incidence rate difference, 45
per 100 000; 90% CI, 8 to 84) than in middle-aged (incidence rate difference, 64 per 100 000; 90% CI, 24 to 102) and
older (incidence rate difference, 89 per 100 000; 90% CI, 44 to 140) adults. Similar results were observed in women.
Conclusion—Alcohol is also associated with a decreased risk of coronary heart disease in younger adults; however, the
absolute risk was small compared with middle-aged and older adults. (Circulation. 2010;121:1589-1597.)
Key Words: age groups 䡲 alcohol consumption 䡲 coronary disease 䡲 epidemiology
T
he association between alcohol intake and coronary
heart disease (CHD) has been thoroughly investigated
over the past decades with regard to both the amount and
type consumed.1– 4 In recent years, the importance of
drinking pattern has also been considered.5–7 In general,
alcohol intake is consistently linked with a lower risk of
CHD. Age-specific incidence rates of CHD vary considerably, being very low in men ⬍40 years of age and in
women ⬍50 years of age.8 For this reason, the statistical
power to investigate the effects of alcohol on CHD in
younger adults is limited. Most results are obtained from
cohorts consisting of middle-aged and older adults, and
only a few studies have addressed the effects of alcohol on
CHD in younger adults.9,10 In principle, the cause of CHD
among younger adults may differ from that among older
individuals; for instance, relatively more cases of CHD
among younger adults may be attributable to genetic
causes.11,12 Hence, alcohol may not necessarily protect
against CHD in this age group. We pooled data from 8
studies to increase sample size and to enable the investi-
Received June 18, 2009; accepted February 12, 2010.
From the Centre for Alcohol Research, National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark (U.A.H., J.S.T.,
M.G.); Social Medicine Section, Department of Public Health, University of Copenhagen, Copenhagen, Denmark (U.A.H.); Research Unit for Dietary
Studies at the Institute of Preventive Medicine, Copenhagen University Hospital, Centre for Health and Society, Copenhagen, Denmark (B.L.H.);
Department of Clinical Epidemiology, Aarhus University Hospital, Aalborg, Denmark (M.U.J.); Departments of Nutrition (E.J.O., W.C.W., E.B.R.,
A.A.), Epidemiology (S.L., D.S., W.C.W., E.B.R., A.A.), and Biostatistics (D.S.), Harvard School of Public Health, Boston Mass; Channing Laboratory,
Department of Medicine, Brigham and Women’s Hospital/Harvard Medical School, Boston Mass (W.C.W., E.B.R., A.A.); Department of Medical
Epidemiology and Biostatistics, Karolinska Institute, Stockholm, Sweden (K.B.); Section of Epidemiology and Biostatistics, Henry N. Neufeld Cardiac
Research Institute, Department of Epidemiology and Preventive Medicine, Tel Aviv University, Tel Aviv, Israel (U.G.); Department of Public Health and
Clinical Medicine, Umeaa University, Umeaa, Sweden (G.H.); National Public Health Institute, Helsinki, Finland (P.K., P.P., J.V.); Division of
Epidemiology and Community Health, School of Public Health, University of Minnesota, Minneapolis (M.A.P.); and Departments of Nutrition and
Epidemiology, School of Public Health, University of North Carolina, Chapel Hill (J.S.).
Guest Editor for this article was Wendy S. Post, MD, MS.
Correspondence to Janne S. Tolstrup, PhD, Senior Researcher, Centre for Alcohol Research, National Institute of Public Health, Oster Farimagsgade
5A, 2nd floor, DK-1399 Copenhagen K. E-mail [email protected]
© 2010 American Heart Association, Inc.
Circulation is available at http://circ.ahajournals.org
DOI: 10.1161/CIRCULATIONAHA.109.887513
1589
Downloaded from circ.ahajournals.org
by on April 9, 2010
1590
Circulation
April 13, 2010
gation of associations between alcohol intake and CHD in
subsets of populations defined by age group.
Clinical Perspective on p 1597
Measurements
Methods
Study Population
The analyses were based on data from the Pooling Project of Diet
and Coronary Disease. The inclusion criteria were the following: a
prospective study with at least 150 incident coronary cases, assessment of usual dietary intake, and a validation study of the diet
assessment method. The following 12 studies met these criteria and
agreed to share data: Adventists Health Study (AHS),13 Atherosclerosis Risk in Communities Study (ARIC), 14 ␣-Tocopherol,
␤-Carotene Cancer Prevention Study (ATBC),15 Finnish Mobile
Clinic Health Examination (FMC),16 Glostrup Population Study
(GPS),17 Health Professionals Follow-Up Study (HPFS),18 Israeli
Ischemic Heart Disease Study (IIHD),19 Iowa Women’s Health
Study (IWHS),20 Nurses’ Health Study (NHS),21 Västerbotten Intervention Program (VIP),22 and Women’s Health Study (WHS).23 The
AHS included only nondrinkers, and the FMC and IIHD studies were
excluded from the present analysis because of missing information
on alcohol intake. In addition, IWHS was excluded from main
analyses because of self-reported information on CHD. The 8
remaining studies are presented in Table 1. The NHS was divided
into 2 segments, thereby taking advantage of repeated assessments
on dietary intake and the long follow-up period. The 2 segments are
referred to as NHSa (1980 to 1986) and NHSb (1986 to 1996). The
second segment comprised only women who remained free of CHD
Table 1.
after the first follow-up period, and cases were included in whichever
segment they occurred.
Average daily alcohol intake was assessed at baseline with a food
frequency or diet history questionnaire inquiring about typical intake
of alcoholic beverages. For each beverage, grams of daily alcohol
intake were calculated from information on the amount and frequency of consumption and the alcohol content of the beverage.
Study-specific conversion factors for the alcohol content were used.
A standard drink contains ⬇10 to 15 g pure alcohol. Total alcohol
intake was given by the sum of the beverage-specific intakes.
The outcome of interest was incident CHD events (both fatal and
nonfatal). All 8 included studies used validated methods to define
nonfatal and fatal CHD cases.24
Statistical Methods
Participants were excluded if they reported energy intakes beyond 3 SD
from the study-specific, log-transformed mean energy intake of the
baseline population (1% of the study population). Persons ⬍35 years of
age or with a history of cardiovascular disease, diabetes, or cancers
(other than nonmelanoma skin cancer) were also excluded. Participants
were followed up from baseline to date of CHD event, date of death, or
end of follow-up, whichever occurred first. Follow-up periods ⬍10
years (ARIC and GPS) were truncated to reduce heterogeneity. Individual studies were combined through the use of an aggregated pooled
analysis technique allowing for calculation of a single exposure-effect
estimate while adjusting for study origin.25
The hazard ratios of CHD were ascertained by the Cox
proportional-hazards regression model with age as underlying time
Baseline Characteristics of Included Studies
CHD Cases, n
Study and
Sex
Baseline
Cohort*
Year of
Questionnaire
Mean Age (90%
Limit), y
Follow-Up,
Person-y
CHD
Deaths
Alcohol Intake
Total CHD
Events
Median† (5th–95th
Percentile), g/d
Abstainers,
%
ARIC
Male
5217
1987–1989
54.6 (45–64)
45 652
51
267
12.1 (1.9–56.6)
45.8
Female
6462
1987–1989
53.9 (45–64)
58 019
18
122
6.8 (1.5–30.2)
69.8
21 141
1984–1988
57.3 (50–69)
121 813
534
1339
13.7 (0.8–62.3)
10.2
Male
1658
1974–1995
51.9 (35–80)
14 365
79
102
20.8 (2.7–72.6)
6.2
Female
1666
1974–1995
51.5 (35–80)
14 605
34
34
8.9 (1.3–35.1)
18.0
41 754
1986–1988
53.4 (39–77)
383 206
421
1273
9.7 (1.0–46.1)
23.0
81 415
1980–1982
47.1 (35–66)
513 915
97
397
5.6 (0.8–35.0)
31.4
61 706
1986–1988
52.6 (39–66)
607 049
208
696
4.9 (0.9–35.9)
34.2
9521
1992–1996
49.1 (39–70)
39 230
38
134
4.4 (0.2–15.8)
4.7
10 555
1992–1996
49.3 (39–70)
43 872
4
23
1.7 (0.1–7.3)
11.4
37 272
1992–1995
53.9 (38–89)
190 755
10
152
3.7 (0.9–28.4)
40.0
79 291
1974–1996
54.0 (35–80)
604 266
1123
3115
9.6 (0.9–50.4)
18.5
199 076
1974–1996
50.4 (35–89)
1 428 216
371
1424
4.7 (0.8–35.0)
33.9
ATBC
Male
GPS
HPFS
Male
NHSa
Female
NHSb
Female
VIP
Male
Female
WHS
Female
Total
Male
Female
*Sample size after exclusion of participants with baseline cardiovascular diseases, cancers, diabetes mellitus, and missing information on alcohol intake.
†Median values were calculated for drinkers only.
Downloaded from circ.ahajournals.org by on April 9, 2010
Hvidtfeldt et al
Table 2.
Alcohol and CHD in Age Strata
1591
Baseline Characteristics of Women in the Pooled Cohort According to Daily Alcohol Intake
Alcohol Intake, g/d
Total
n*
192 067
Nondrinkers
65 121
0.1– 4.9
5.0 –14.9
67 187
39 177
CHD events, n
1365
596
390
241
Age, mean (SD), y
50.4 (7.6)
51.0 (7.7)
49.8 (7.7)
50.1 (7.5)
Education, low, n (%)†
Smokers, n (%)
BMI, mean (SD), kg/m2
Physical inactivity, n (%)
8276 (4)
2186 (3)
4719 (7)
44 513 (23)
11 971 (18)
14 547 (22)
25.0 (10.1)
66 248 (25)
936 (2)
10 454 (27)
15.0 –29.9
12 258
65
58
50.5 (7.4)
ⱖ60.0
30.0 –59.9
7418
51.1 (7.3)
906
15
51.3 (7.2)
322 (3)
96 (1)
17 (2)
3628 (30)
3426 (46)
487 (54)
26.0 (5.3)
25.1 (4.5)
24.0 (3.8)
23.7 (3.6)
23.9 (3.8)
24.4 (4.2)
24 382 (37)
21 858 (33)
12 760 (33)
3859 (31)
2971 (40)
418 (46)
Diet, median (5th–95th
percentile)
Polyunsaturated fat,‡ g/d
5.4 (3.3–8.4)
5.5 (3.4–8.6)
5.4 (3.4–8.4)
5.4 (3.4–8.4)
5.3 (3.2–8.3)
4.8 (2.8–7.9)
4.0 (2.2–6.9)
Monounsaturated fat,‡ g/d
13.1 (8.1–20.5)
13.1 (7.9–20.7)
13.0 (8.3–20.5)
13.2 (8.4–20.5)
12.9 (8.2–19.7)
12.3 (7.7–19.0)
10.6 (6.4–16.8)
Saturated fat,‡ g/d
12.8 (7.7–20.0)
12.7 (7.5–20.1)
13.0 (8.0–20.2)
12.9 (7.9–20.1)
12.6 (7.6–19.7)
11.9 (7.2–18.6)
10.3 (5.7–16.8)
Fiber,‡ g/d
15.4 (8.4–25.7)
15.7 (8.5–26.7)
15.9 (8.9–26.1)
15.0 (8.3–24.4)
14.2 (7.8–23.1)
12.2 (6.7–20.4)
10.3 (5.0–18.9)
Cholesterol,‡ mg/d
254 (139–458)
252 (134–464)
249 (137–451)
264 (148–460)
263 (151–464)
252 (143–443)
222 (121–394)
Total energy, kcal/d
1603 (900–2627)
1579 (871–2637)
1581 (893–2589)
1614 (915–2608)
1678 (977–2689)
1758 (1070–2749)
2008 (1283–3050)
Hypertension, n (%)
35 637 (19)
13 566 (21)
11 695 (17)
6271 (16)
2137 (17)
1695 (23)
273 (30)
Dyslipidemia, n
20 850 (11)
8389 (13)
6599 (10)
3787 (10)
1217 (10)
731 (10)
127 (14)
Differences in distribution of covariates across levels of alcohol consumption were tested by ANOVA (age, BMI), Kruskal-Wallis (dietary factors), and ␹2 tests
(education, smoking, physical activity, hypertension, and dyslipidemia). All tests showed statistically significant differences (P⬍0.0001).
*After exclusion of participants with missing information on any of the relevant covariates.
†Defined as less than high school.
‡Energy adjusted.
scale, allowing for delayed entry (left censoring).26 Absolute risks
(incidence rates) describing the scale of CHD according to sex, age,
and level of alcohol intake were estimated by means of Poisson
regression.27 Absolute risk differences were calculated, and 90%
confidence limits were derived by bootstrap estimation (5000 replications) with the 5th and 95th percentiles of the distribution as the
lower and upper limits.
We performed primary analyses considering the risk of CHD in
categories of alcohol consumption (0, 0.1 to 4.9, 5.0 to 14.9, 15.0 to
29.9, 30.0 to 59.9, and ⱖ60.0 g/d in women; 0, 0.1 to 4.9, 5.0 to 14.9,
15.0 to 29.9, 30.0 to 59.9, 60.0 to 89.9, and ⱖ90.0 g/d in men) both
for each individual study and for the pooled cohort. The study
population was analyzed separately in the following 3 age groups: 39
to 49.9, 50 to 59.9, and ⱖ60 years. Age was updated during
follow-up, and participants were assigned to the appropriate age
category; thus, each person could contribute person-time at risk to
⬎1 age category.
Additional analyses exploring the risk of CHD per alcohol
increment (1 g/d) were performed. Alcohol was modeled continuously using second-degree fractional polynomials, thus allowing for
a single turning point (the nadir) in the risk function. Following the
results of Corrao and others,1 a model describing the dose-response
relationship of alcohol on CHD including both a linear and rootsquared term of alcohol was applied.
The P value for the test for trend was obtained by assigning the
median value within categories of alcohol intake and using this
variable as a continuous variable. SAS statistical package version 9.1
was used for all analyses.28
We harmonized the variables of the different studies, and the
following set of potential confounders was identified on the basis of
the method of causal diagrams as suggested by Greenland and
others29: educational level (less than high school, high school, more
than high school), smoking (never smokers, ex-smokers, and current
smokers of 1 to 4, 5 to 14, 15 to 24, or ⱖ25 cigarettes per day), body
mass index (BMI; ⬍18.5, 18.5 to 24.9, 25.0 to 29.9, and ⱖ30
kg/m2), total energy intake (kcal/d), and energy-adjusted quintiles of
cholesterol, dietary fiber, saturated fat, monounsaturated fat, and
polyunsaturated fat intake. Physical activity measures varied across
the cohorts, measured either according to an energy expenditure
score of weekly time spent on various activities during the past year
(ARIC, HPFS, NHS, VIP, and WHS) or according to the intensity of
the average weekly physical activity during the past 12 months
(ATBC, GPS). These measures were harmonized to a 5-level
variable from 1 (least active) to 5 (most active).30 –32 In addition,
models were stratified by study origin and baseline year to account
for differences in follow-up procedures or questionnaire design and
period effects. Information on postmenopausal hormone therapy use
was unavailable in VIP; therefore, the main analyses for women did
not include adjustment for this factor. Sensitivity analyses included
measures of self-reported history of physician-diagnosed elevated
cholesterol (dyslipidemia) and hypertension (yes/no).
Results
Baseline Characteristics
Baseline characteristics of participants from the 8 included
studies are shown in Table 1. In total, the pooled study
population comprised 199 076 women and 79 291 men who
experienced 1424 and 3115 coronary events during 1 428 216
and 604 266 person-years of follow-up, respectively. Baseline age of participants varied from 35 to 89 years in women
and from 35 to 80 years in men with a mean age of 50.4 and
54.0 years, respectively. The proportion of nondrinkers varied
substantially between studies from 11.4% to 53.0% in women
and from 4.7% to 45.8% in men. Median alcohol intakes
varied from 4.4 to 20.8 g/d in men and from 1.7 to 8.9 g/d in
women.
Tables 2 and 3 show characteristics of participants included in the pooled cohort according to alcohol consumption. Heavier alcohol intake was associated with higher
proportions of smokers, physical inactivity, and hypertension
and lower median intakes of fat and fiber; a moderate alcohol
intake was associated with the highest median cholesterol
intake compared with abstainers and heavy drinkers. In men,
Downloaded from circ.ahajournals.org by on April 9, 2010
1592
Circulation
April 13, 2010
Table 3.
Baseline Characteristics of Men in the Pooled Cohort According to Daily Alcohol Intake
Alcohol Intake, g/d
Total
n*
74 720
Nondrinkers
13 904
0.1– 4.9
19 030
5.0 –14.9
20 556
751
30.0 –59.9
7789
449
ⱖ90.0
60.0 – 89.9
1546
311
520 (1)
CHD events, n
2961
623
Age, mean (SD), y
54.0 (8.4)
54.4 (8.6)
53.3 (8.7)
53.6 (8.5)
54.6 (7.6)
55.2 (7.6)
55.2 (7.6)
55.8 (6.2)
Education, low, n (%)†
20 773 (28)
2513 (18)
6125 (32)
5580 (27)
3590 (32)
2283 (29)
450 (29)
232 (45)
Smokers, n (%)
28 144 (38)
3338 (24)
6086 (32)
7409 (36)
5688 (50)
4254 (55)
952 (62)
417 (80)
BMI, mean (SD),
kg/m2
25.8 (8.2)
26.0 (3.8)
25.8 (3.4)
25.7 (3.4)
25.8 (3.4)
25.9 (3.5)
26.1 (3.9)
26.5 (4.1)
18 850 (25)
3740 (27)
4411 (23)
4489 (22)
2956 (26)
2421 (31)
572 (37)
261 (80)
Physical inactivity,
n (%)
737
15.0 –29.9
11 375
59
31
Diet, median
(5th–95th
percentile)
Polyunsaturated
fat,‡ g/d
5.4 (3.3–8.9)
5.5 (3.3–9.0)
5.3 (3.4–8.8)
5.4 (3.4–8.9)
5.4 (3.2–9.0)
5.2 (3.0–8.8)
4.6 (2.6–8.2)
4.1 (2.2–7.8)
Monounsaturated
fat,‡ g/d
12.8 (8.5–16.8)
12.9 (8.0–17.3)
12.8 (8.6–16.8)
12.9 (8.8–16.8)
12.9 (9.0–16.6)
12.4 (8.3–16.3)
11.3 (7.4–15.4)
10.5 (6.8–15.5)
Saturated fat,‡ g/d
13.2 (7.5–23.8)
12.3 (6.9–22.8)
13.6 (7.7–24.0)
13.3 (7.7–24.0)
13.9 (7.8–24.5)
13.2 (7.4–23.5)
12.7 (6.5–21.6)
13.3 (6.5–21.0)
Fiber,‡ g/d
19.8 (11.5–32.3)
20.7 (11.6–35.7)
21.0 (12.7–33.5)
20.1 (12.3–31.8)
19.1 (11.5–30.1)
17.0 (10.1–27.4)
14.6 (8.3–24.5)
12.9 (6.6–21.2)
Cholesterol,‡ mg/d
322 (174–573)
315 (161–568)
303 (166–536)
322 (177–566)
349 (193–602)
343 (190–610)
323 (176–606)
324 (167–544)
Total energy, kcal/d
2135 (1149–3671)
1929 (1044–3438)
2047 (1120–3537)
2111 (1173–3598)
2311 (1274–3823)
2387 (1344–3911)
2652 (1518–4138)
3054 (1570–4662)
14 388 (19)
2759 (20)
3292 (17)
3634 (18)
2246 (20)
1902 (24)
421 (27)
134 (26)
5076 (9)
1145 (10)
1114 (8)
1312 (9)
720 (11)
609 (13)
147 (17)
29 (16)
Hypertension, n (%)
Dyslipidemia,§ n
Differences in distribution of covariates across levels of alcohol consumption were tested by ANOVA (age, BMI), Kruskal-Wallis (dietary factors), and ␹2 tests
(education, smoking, physical activity, hypertension, and dyslipidemia). All tests showed statistically significant differences (P⬍0.0001).
*After exclusion of participants with missing information on any of the relevant covariates.
†Defined as less than high school.
‡Energy-adjusted.
§This information was not available for ATBC.
a higher frequency low educational level was observed
among participants in the highest alcohol group. Similar
distributions of covariates according to alcohol intake were
observed across cohorts.
Alcohol Consumption and Risk of CHD
Tables 4 and 5 show the relative risks (hazard ratios) of
CHD by categories of alcohol intake for each of the studies
Table 4.
and pooled estimates for women and men. In women, an
inverse relation between alcohol intake and CHD was
found in each individual study except for VIP. The
confidence bounds around risk estimates for this particular
study were very broad because the reference category
included only 2 cases. In men, an inverse relation was
observed in all studies. In the pooled analysis, we observed
a significantly lower risk of CHD among women with an
Study-Specific and Pooled Hazard Ratios of CHD for Categories of Daily Alcohol Intake for Women
Hazard Ratio (95% CI) by Daily Alcohol Intake, g/d
n
Hazard Ratio,
Nondrinkers
(CHD⫽606)
0.1– 4.9
(CHD⫽397)
5.0 –14.9
(CHD⫽242)
15.0 –29.9
(CHD⫽65)
30.0 –59.9
(CHD⫽60)
ⱖ60.0
(CHD⫽15)
P for
Trend
Study specific
ARIC
6406
1.00 (Reference)
0.84 (0.44–1.59)
0.58 (0.29–1.17)
0.78 (0.30–2.00)
NA§
NA§
0.0738
GPS
1509
1.00 (Reference)
0.75 (0.27–2.06)
0.89 (0.31–2.55)
1.15 (0.31–4.23)
0.52 (0.05–5.21)
NA§
0.8418
NHSa
79 479
1.00 (Reference)
0.73 (0.57–0.94)
0.72 (0.55–0.95)
0.59 (0.38–0.94)
0.49 (0.29–0.82)
1.64 (0.77–3.49)
0.1548
NHSb
60 083
1.00 (Reference)
0.78 (0.65–0.94)
0.67 (0.54–0.84)
0.47 (0.32–0.69)
0.57 (0.39–0.82)
0.60 (0.26–1.38)
0.0003
9758
1.00 (Reference)
2.26 (0.28–18.47)
2.17 (0.12–38.67)
NA§
NA§
NA§
0.8654
34 832
1.00 (Reference)
1.02 (0.69–1.50)
0.81 (0.49–1.35)
0.37 (0.11–1.19)
0.58 (0.18–1.90)
1.32 (0.18–9.95)
0.1768
Age adjusted*
192 067
1.00 (Reference)
0.78 (0.69–0.89)
0.73 (0.62–0.84)
0.57 (0.44–0.74)
0.80 (0.61–1.05)
1.71 (1.02–2.86)
0.1112
Smoking adjusted†
192 067
1.00 (Reference)
0.75 (0.65–0.85)
0.62 (0.53–0.72)
0.47 (0.36–0.61)
0.52 (0.39–0.68)
1.02 (0.61–1.70)
⬍0.0001
Multivariable‡
192 067
1.00 (Reference)
0.78 (0.69–0.90)
0.68 (0.59–0.80)
0.52 (0.40–0.67)
0.53 (0.39–0.70)
0.93 (0.55–1.58)
⬍0.0001
VIP
WHS
Pooled
*Also adjusted for year of baseline questionnaire.
†Also adjusted for age and year of baseline questionnaire.
‡Multivariable hazard ratios were adjusted for age, year of baseline questionnaire, smoking, BMI, education, physical activity, energy intake, polyunsaturated fat,
monounsaturated fat, saturated fat, fiber, cholesterol intake, and study origin.
§Not applicable because of a limited number of cases.
Downloaded from circ.ahajournals.org by on April 9, 2010
Hvidtfeldt et al
Table 5.
Alcohol and CHD in Age Strata
1593
Study–Specific and Pooled Hazard Ratios of CHD for Categories of Daily Alcohol Intake for Men
Alcohol Intake, g/d
n
Hazard Ratio,
Nondrinkers
(CHD⫽623)
0.1– 4.9
(CHD⫽737)
5.0 –14.9
(CHD⫽751)
15.0 –29.9
(CHD⫽449)
30.0 –59.9
(CHD⫽311)
60.0 – 89.9
(CHD⫽59)
ⱖ90.0
(CHD⫽31)
P for
Trend
Study specific
ARIC
5166
1.00 (Reference)
1.16 (0.78–1.71)
1.29 (0.93–1.79)
0.87 (0.57–1.32)
0.73 (0.44–1.22)
0.74 (0.27–2.04)
1.40 (0.56–3.54)
ATBC
21 119
1.00 (Reference)
0.86 (0.71–1.03)
0.86 (0.72–1.03)
0.70 (0.58–0.85)
0.61 (0.49–0.77)
0.48 (0.32–0.74)
0.80 (0.51–1.27)
0.0001
1294
1.00 (Reference)
0.84 (0.30–2.32)
0.54 (0.21–1.38)
0.60 (0.24–1.50)
0.44 (0.16–1.18)
0.37 (0.09–1.52)
NA§
0.0304
GPS
0.4499
38 654
1.00 (Reference)
1.00 (0.85–1.17)
0.75 (0.63–0.88)
0.69 (0.56–0.86)
0.66 (0.52–0.83)
0.65 (0.41–1.01)
0.17 (0.02–1.23)
⬍0.0001
8486
1.00 (Reference)
0.50 (0.28–0.93)
0.24 (0.11–0.48)
0.25 (0.07–0.91)
0.73 (0.09–5.85)
NA§
NA§
0.009
Age-adjusted*
74 719
1.00 (Reference)
0.95 (0.85–1.06)
0.84 (0.75–0.93)
0.75 (0.66–0.85)
0.74 (0.64–0.85)
0.70 (0.53–0.91)
0.96 (0.67–1.39)
⬍0.0001
Smoking-adjusted†
74 719
1.00 (Reference)
0.94 (0.84–1.05)
0.81 (0.72–0.90)
0.71 (0.63–0.81)
0.66 (0.58–0.76)
0.60 (0.46–0.79)
0.81 (0.57–1.18)
⬍0.0001
Multivariable‡
74 719
1.00 (Reference)
0.96 (0.86–1.08)
0.83 (0.74–0.92)
0.72 (0.64–0.82)
0.66 (0.57–0.76)
0.58 (0.44–0.77)
0.77 (0.53–1.13)
⬍0.0001
HPFS
VIP
Pooled
*Also adjusted for year of baseline questionnaire.
†Also adjusted for age and year of baseline questionnaire.
‡Multivariable hazard ratios were adjusted for age, year of baseline questionnaire, smoking, BMI, education, physical activity, energy intake, polyunsaturated fat,
monounsaturated fat, saturated fat, fiber, cholesterol intake, and study origin.
§Not applicable because of a limited number of cases.
alcohol intake of up to 60 g/d and among men with an
alcohol intake of up to 90 g/d.
We also performed analyses describing the risk of CHD
according to alcohol consumption modeled as a continuous
variable (Figure 1). In both men and women, a reduction in
CHD risk was observed at low to moderate levels of alcohol.
The relative risk of CHD was 0.58 (95% confidence interval
[CI], 0.49 to 0.68) in women and 0.69 (95% CI, 0.62 to 0.76)
in men with a daily intake of 30 g/d, corresponding to ⬇2 to
3 drinks. Higher levels of alcohol consumption were not
associated with any discernible additional protection in
women and with only modest protection in men.
Alcohol Consumption and Risk of CHD in
Age Strata
We estimated the risks of CHD according to alcohol intake
separately for 3 age groups (ⱕ50, 50 to 59, and ⱖ60 years;
Figure 2). In all age groups and for both men and women, a
decreased risk of CHD according to alcohol intake was
observed but with broader confidence bounds for the youngest age group. The test for interaction between alcohol and
age was not statistically significant in either women (P⫽0.34)
or men (P⫽0.25). We also modeled the risk continuously for
the different age groups and observed similar shapes of the
curves (data not shown).
Incidence rates of CHD in women and men according to
alcohol intake and age are displayed in Figure 3. As expected,
the incidence of CHD was much lower in the younger
compared with older participants. The incidence rates of
CHD among female abstainers in the 3 age groups were 11
(95% CI, 1 to 109), 41 (95% CI, 1 to 400), and 103 (95% CI,
9 to 1018) per 100 000, respectively. In male abstainers, the
incidence rates were 114 (95% CI, 77 to 171), 262 (95% CI,
201 to 343), and 454 (95% CI, 354 to 553) per 100 000 for the
3 age groups, respectively. In all age groups and in both men
and women, the incidence rate was lower among participants
with a low to moderate alcohol intake compared with abstainers. In women, the incidence rate differences between drinking 0 g/d and 5.0 to 29.9 g/d were 3 (90% CI, ⫺1 to 25), 16
(90% CI, 0 to 111), and 35 (90% CI, 0 to 250). For men, the
corresponding incidence rate differences were 45 (90% CI, 8
to 84), 64 (90% CI, 24 to 102), and 89 (90% CI, 44 to 140)
per 100 000.
Sensitivity Analyses
Heterogeneity between study-specific effects was assessed by
the inclusion of an interaction term between alcohol and
study origin under the null hypothesis of no between-study
differences in the relative risk of CHD by alcohol intake,1,25
with no sign of heterogeneity detected (P⫽0.95 in women,
P⫽0.12 in men). In addition, comparing pooled risk estimates
Figure 1. Relative risk functions (95%
CI) describing the dose-response relation between alcohol intake and risk of
CHD. Analyses were adjusted for year of
baseline questionnaire, education, smoking, BMI, physical activity, total energy
intake, polyunsaturated fat, monounsaturated fat, saturated fat, fiber, and cholesterol intake. The fitted model (SE) is
reported. The 99th percentile of cases
was 64 g/d in women and 90 g/d in
men. The highest alcohol intake in
observed cases was 90 g/d in women
and 215 g/d in men.
Downloaded from circ.ahajournals.org by on April 9, 2010
1594
Circulation
April 13, 2010
Figure 2. Sex- and age-specific pooled
hazard ratios of CHD for categories of
daily alcohol intake. Multivariable hazard
ratios were adjusted for year of baseline
questionnaire, education, smoking, BMI,
physical activity, total energy intake,
polyunsaturated fat, monounsaturated
fat, saturated fat, fiber, and cholesterol
intake. n indicates number cases for
each sex and age group.
after systematically excluding each study at a time confirmed
that no single study strongly influenced the pooled estimates.33 Hence, the pooled hazard ratios are considered
appropriate summaries of the study-specific data. Performing
a test for interaction between age (time scale) and alcohol
consumption did not yield violations of the proportionalhazards assumption for either women (P⫽0.10) or men
(P⫽0.22).
Separate analyses were performed for fatal and nonfatal
events to examine whether the effect of alcohol on CHD
differed according to the severity of the outcome. The results
of this analysis did not reveal obvious differences between
the 2 measures of outcome, although a tendency toward an
elevated risk of fatal CHD was observed for the highest
alcohol category in both women and men (data not shown).
To examine the possibility that latent baseline symptoms of
CHD might reduce the alcohol intake, thereby biasing the
results, we performed analyses in which the first 2 or 4 years
after baseline were excluded. This did not attenuate the
estimates (data not shown).
Additional analyses for women were performed to examine
whether adjustment for postmenopausal hormone replacement therapy had an impact on results. This involved excluding participants with unavailable information on this particular covariate (n⫽9799, 5% of the study population). In the
remaining cohorts, postmenopausal hormone replacement
therapy did not appreciably affect the association between
alcohol and risk of CHD. In addition, the inclusion of history
of hypertension and dyslipidemia as covariates in the model
did not affect the risk estimates of CHD according to alcohol
consumption. Furthermore, an analysis was performed including the IWHS (n⫽29 801). This inclusion also did not
change the hazard ratios appreciably (data not shown).
A test for interaction between alcohol and smoking was
performed to examine whether the estimates of the effect
of alcohol on CHD risk differed according to smoking
status. Smoking did not modify the association between
alcohol and CHD in either men (P⫽0.79) or women
(P⫽0.14). Additional analyses including nonsmokers only
were performed separately for women (n⫽100 144) and
Downloaded from circ.ahajournals.org by on April 9, 2010
Hvidtfeldt et al
Figure 3. Incidence rates of CHD according to age and alcohol
intake. Analyses were adjusted for year of baseline questionnaire, education, smoking, BMI, physical activity, total energy
intake, polyunsaturated fat, monounsaturated fat, saturated fat,
fiber, and cholesterol intake.
men (n⫽46 576) for alcohol levels of 0, 0.1 to 14.9, 15.0
to 29.9, and ⱖ30.0 g/d and showed similar associations
(data not shown).
Discussion
In this pooled cohort of 8 prospective studies, we observed a
lower risk of CHD among men and women with a light to
moderate alcohol intake compared with nondrinkers. This
finding was consistent across different age groups without
significant variations in dose response.
The current data on the effect of alcohol on CHD in
younger adults are sparse. In a study based on the Honolulu
Heart Program, the authors compared CHD risk according to
conventional risk factors in middle-aged and older men (age,
45 to 93 years). Compared with nondrinkers, they observed a
lower risk of CHD among drinkers in middle-aged but not
among older participants (ⱖ75 years) and concluded accordingly that the relation between alcohol and CHD weakened
with age. The study, however, was limited by the simple
categorization of alcohol intake into drinkers or nondrinkers
and included men only.10
Several plausible explanations for the lowered risk of CHD
among moderate drinkers exist. Among those explanations,
the evidence is probably strongest for a mechanism involving
alcohol increasing high-density lipoprotein cholesterol and
reducing plasma fibrinogen levels, thereby reducing platelet
aggregability.7,34 The hypothesized J-shaped relation between
Alcohol and CHD in Age Strata
1595
alcohol intake and diabetes mellitus could also explain some
of the benefit from alcohol intake.35,36 In addition, alcohol has
an effect on plasminogen activator inhibitor-1 that would tend
to reduce thrombosis.37
Previous studies have suggested that the causes of CHD in
younger adults differ from the mechanisms involved with
CHD in older persons. Results from the Honolulu Heart
Program indicated that the effect of hypertension, BMI, and
cholesterol on CHD differed according to age. For instance,
the relative risk of CHD in hypertensive men declined from
3.7 in those 45 to 54 years of age to 1.7 in those ⱖ75 years
of age. Similarly, associations between BMI and total cholesterol weakened with advancing age.10 The Coronary Artery
Risk in Young Adults (CARDIA) study of men and women
33 to 45 years of age found that alcohol intake was associated
with expected dose response between alcohol and highdensity lipoprotein cholesterol levels and an inverse relation
between alcohol and fibrinogen levels.9 They also observed
an increased risk of coronary calcification with greater
alcohol consumption. Because coronary calcification is a
marker of atherosclerosis, this result in young adults is not
consistent with the results of the present study.9
Another aspect of alcohol consumption related to age is
drinking patterns. Younger adults may tend to binge drink
more often than older persons, which may increase their risk
of CHD5,7,9; however, findings from the CARDIA study did
not indicate a protective effect of alcohol intake on coronary
calcification in younger adults even after the exclusion of
binge drinkers.9
Our findings suggest a J-shaped curve in women, but in
men, the risk did not increase significantly at high amounts of
alcohol. Biomarkers that mediate the association between
alcohol and decreased risk of CHD such as high-density
lipoprotein and fibrinogen are found to explain a larger
proportion of the association among men than among women,
which may indicate that alcohol has specific effects on such
mediators according to sex.7 Other biological explanations for
sex-specific associations include differences in alcohol pharmacokinetics (ie, processing and elimination of alcohol in the
body), which depend largely on body composition.38,39 However, because the risk curves of this study were modeled
separately for the 2 sexes, comparisons between them are not
straightforward.
The present study is one of only a few existing studies
focusing on the effects of alcohol on CHD according to age.
Our work was based on a large body of data with thorough
measurements of alcohol intake and relevant covariates. A
strength was the availability of diet data in the Pooling
Project of Diet and Coronary Disease that enabled adjustment
for potential dietary confounders. The size of the study
population allowed us to perform subset analyses exploring
the association between alcohol and CHD in strata of younger
men and women, aspects that even large individual cohorts do
not have the power to address. The findings of the present
study were strengthened by the prospective design, which
provided information on the sequence of events allowing for
conclusions on causality, assuming proper confounding control. Potential confounders of the association between alcohol
and CHD were carefully selected on the basis of directed
Downloaded from circ.ahajournals.org by on April 9, 2010
1596
Circulation
April 13, 2010
acyclic graphs, ensuring a minimally sufficient set of covariates. Furthermore, the inclusion criteria of the Pooling Project
of Diet and Coronary Disease enabled adjustment for relevant
dietary factors, which most previous studies did not control
for. The pooled analyses included both cohorts and intervention studies from North America and Europe, and similar
effects were observed across the studies. Finally, an advantage of the Pooling Project is the inclusion of previously
unpublished results, thereby reducing the risk of publication
bias.
However, several limitations of the study should also be
considered. We focused on the importance of the amount of
alcohol consumed; however, other aspects of patterns of
alcohol intake may be equally important and were not
addressed. Our study included only information on current
alcohol consumption and confounders at baseline. For this
reason, the reference category of abstainers may contain
former drinkers who quit because of existing illness, which
could cause a moderate alcohol intake to appear more
protective than it is. Although several studies that included
only lifelong or long-term abstainers in this category have
confirmed a protective effect of alcohol even among healthy
individuals,6,40 the “sick-quitter” hypothesis is relevant in the
present context because older age groups may include more
abstainers who stopped drinking because of illness (eg,
hypertension).
As mentioned, patterns of alcohol consumption (eg, choice
of alcohol type and frequency of consumption) may differ
considerably with age, which we did not account for. Our
findings of protective effects of alcohol on CHD in all
examined age groups may indicate that neither the type of
alcohol nor the frequency of consumption modifies the
influence of alcohol on CHD considerably. However, future
research based on observational studies should emphasize
other measures of alcohol intake such as frequency of alcohol
consumed. Furthermore, cohort studies with information on
lifetime alcohol intake or repeated measurements of alcohol
intake and potential confounders could contribute valuable
insight because changes in alcohol intake over a given period
may be of great importance. Additional experimental studies
are also needed to expand the knowledge of biological
mechanisms.
Furthermore, this study focused only on CHD events.
Overall effects of alcohol on all-cause morbidity and mortality must be considered if we are to optimize alcohol guidelines for different age groups of the population. The lower
risk of all-cause mortality is expected to be caused mainly by
the effects of alcohol on CHD. In this study, a lower risk of
CHD was observed in all examined age groups in moderate
alcohol consumers compared with abstainers; however, the
absolute risk of CHD was rather low in the youngest age
group. Thus, considering the increasing contribution of CHD
to all-cause mortality with age, it is reasonable to assume that
the protective effect of alcohol on all-cause mortality is
mostly pronounced in older age groups. This issue has been
addressed in a few previous studies indicating that the
protective effect of alcohol consumption on mortality in
general is confined to middle-aged and older individuals.4,41,42 Unfortunately, information on all-cause mortality
was not collected in the database of the Pooling Project of
Diet and Coronary Disease.
Conclusions
This study supports current knowledge that alcohol in moderate amounts protects against CHD in both men and women.
Our findings further suggest that this effect is also present in
younger age groups. However, younger adults are at low risk
for CHD, and the beneficial effects obtained by a moderate
alcohol intake may be negligible compared with the increased
risk of, for instance, traffic accidents and cancer. Recommendations on alcohol intake among younger adults should
consider all-cause mortality and morbidity.
Sources of Funding
This research was supported by research grant R01 HL58904 from
the National Heart, Lung, and Blood Institute of the National
Institutes of Health. The Unit for Dietary Studies was funded by the
Female Researchers in Joint Action program of the Danish Medical
Research Council.
Disclosures
None.
References
1. Corrao G, Rubbiati L, Bagnardi V, Zambon A, Poikolainen K. Alcohol
and coronary heart disease: a meta-analysis. Addiction. 2000;95:
1505–1523.
2. Grønbæk M, Becker U, Johansen D, Gottschau A, Schnohr P, Hein HO,
Jensen G, Sørensen TI. Type of alcohol consumed and mortality from all
causes, coronary heart disease, and cancer. Ann Intern Med. 2000;133:
411– 419.
3. Rimm EB, Giovannucci EL, Willett WC, Colditz GA, Ascherio A,
Rosner B, Stampfer MJ. Prospective study of alcohol consumption and
risk of coronary disease in men. Lancet. 1991;338:464 – 468.
4. Fuchs CS, Stampfer MJ, Colditz GA, Giovannucci EL, Manson JE,
Kawachi I, Hunter DJ, Hankinson SE, Hennekens CH, Rosner Bl.
Alcohol consumption and mortality among women. N Engl J Med. 1995;
332:1245–1250.
5. Tolstrup J, Jensen MK, Tjønneland A, Overvad K, Mukamal KJ,
Grønbæk M. Prospective study of alcohol drinking patterns and coronary
heart disease in women and men. BMJ. 2006;332:1244 –1248.
6. Mukamal KJ, Conigrave KM, Mittleman MA, Camargo CA Jr, Stampfer
MJ, Willett WC, Rimm EB. Roles of drinking pattern and type of alcohol
consumed in coronary heart disease in men. N Engl J Med. 2003;348:
109 –118.
7. Mukamal KJ, Jensen MK, Grønbæk M, Stampfer MJ, Manson JE,
Pischon T, Rimm EB. Drinking frequency, mediating biomarkers, and
risk of myocardial infarction in women and men. Circulation. 2005;112:
1406 –1413.
8. Stamler J. Established major risk factors: historical overview. In: Marmot
M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology
to Public Health. New York, NY: Oxford University Press; 2005:18 –31.
9. Pletcher MJ, Varosy P, Kiefe CI, Lewis CE, Sidney S, Hulley SB.
Alcohol consumption, binge drinking, and early coronary calcification:
findings from the Coronary Artery Risk Development in Young Adults
(CARDIA) Study. Am J Epidemiol. 2005;161:423– 433.
10. Abbott RD, Curb JD, Rodriguez BL, Masaki KH, Yano K, Schatz IJ,
Ross GW, Petrovitch H. Age-related changes in risk factor effects on the
incidence of coronary heart disease. Ann Epidemiol. 2002;12:173–181.
11. Marenberg ME, Risch N, Berkman LF, Floderus B, de Faire U. Genetic
susceptibility to death from coronary heart disease in a study of twins.
N Engl J Med. 1994;330:1041–1046.
12. Zdravkovic S, Wienke A, Pedersen NL, Marenberg ME, Yashin AI, de
Faire U. Heritability of death from coronary heart disease: a 36-year
follow-up of 20 966 Swedish twins. J Intern Med. 2002;252:247–254.
13. Fraser GE, Sabate J, Beeson WL, Strahan TM. A possible protective
effect of nut consumption on risk of coronary heart disease: the Adventist
Health Study. Arch Intern Med. 1992;152:1416 –1424.
Downloaded from circ.ahajournals.org by on April 9, 2010
Hvidtfeldt et al
14. ARIC Investigators. The Atherosclerosis Risk in Communities (ARIC)
Study: design and objectives. Am J Epidemiol. 1989;129:687–702.
15. ATBC Cancer Prevention Study Group. The Alpha-Tocopherol, BetaCarotene Lung Cancer Prevention Study: design, methods, participant
characteristics, and compliance. Ann Epidemiol. 1994;4:1–10.
16. Knekt P, Reunanen A, Jarvinen R, Seppanen R, Heliovaara M, Aromaa
A. Antioxidant vitamin intake and coronary mortality in a longitudinal
population study. Am J Epidemiol. 1994;139:1180 –1189.
17. Hoyer AP, Engholm G. Serum lipids and breast cancer risk: a cohort
study of 5,207 Danish women. Cancer Causes Control. 1992;3:403– 408.
18. Rimm EB, Giovannucci EL, Stampfer MJ, Colditz GA, Litin LB, Willett
WC. Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. Am J Epidemiol. 1992;135:1114 –1126.
19. Balogh M, Medalie JH, Smith H, Groen JJ. The development of a dietary
questionnaire for an ischemic heart disease survey. Isr J Med Sci. 1968;
4:195–203.
20. Munger RG, Folsom AR, Kushi LH, Kaye SA, Sellers TA. Dietary
assessment of older Iowa women with a food frequency questionnaire:
nutrient intake, reproducibility, and comparison with 24-hour dietary
recall interviews. Am J Epidemiol. 1992;136:192–200.
21. Willett WC, Sampson L, Stampfer MJ, Rosner B, Bain C, Witschi J,
Hennekens CH, Speizer FE. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol. 1985;122:51– 65.
22. Hallmans G, Agren A, Johansson G, Johansson A, Stegmayr B, Jansson
JH, Lindahl B, Rolandsson O, Söderberg S, Nilsson M, Johansson I,
Weinehall L. Cardiovascular disease and diabetes in the Northern Sweden
Health and Disease Study Cohort: evaluation of risk factors and their
interactions. Scand J Public Health Suppl. 2003;61:18 –24.
23. Rexrode KM, Lee IM, Cook NR, Hennekens CH, Buring JE. Baseline
characteristics of participants in the Women’s Health Study. J Womens
Health Gend Based Med. 2000;9:19 –27.
24. Pereira MA, O’Reilly E, Augustsson K, Fraser GE, Goldbourt U,
Heitmann BL, Hallmans G, Knekt P, Liu S, Pietinen P, Spiegelman D,
Stevens J, Virtamo J, Willett WC, Ascherio A. Dietary fiber and risk of
coronary heart disease: a pooled analysis of cohort studies. Arch Intern
Med. 2004;164:370 –376.
25. Smith-Warner SA, Spiegelman D, Ritz J, Albanes D, Beeson WL,
Bernstein L, Berrino F, van den Brandt PA, Buring JE, Cho E, Colditz
GA, Folsom AR, Freudenheim JL, Giovannucci E, Goldbohm RA,
Graham S, Harnack L, Horn-Ross PL, Krogh V, Leitzmann MF,
McCullough ML, Miller AB, Rodriguez C, Rohan TE, Schatzkin A,
Shore R, Virtanen M, Willett WC, Wolk A, Zeleniuch-Jacquotte A,
Zhang SM, Hunter DJ. Methods for pooling results of epidemiologic
studies: the Pooling Project of Prospective Studies of Diet and Cancer.
Am J Epidemiol. 2006;163:1053–1064.
26. Clayton D, Hills M. Cox’s Method for Follow-Up Studies: Statistical
Models in Epidemiology. New York, NY: Oxford; 1993:298 –306.
Alcohol and CHD in Age Strata
1597
27. Woodward M. Modelling follow-up data. In: Woodward M, ed. Epidemiology: Study Design and Data Analysis. Boca Raton, Fla: Chapman &
Hall; 2005:601– 671.
28. SAS/STAT Software [computer program]. Version 9.1 for Windows. Cary,
NC: SAS Institute Inc; 2002.
29. Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic
research. Epidemiology. 1999;10:37– 48.
30. Devore EE, Kang JH, Okereke O, Grodstein F. Physical activity levels
and cognition in women with type 2 diabetes. Am J Epidemiol. 2009;
170:1040 –1047.
31. Tanasescu M, Leitzmann MF, Rimm EB, Hu FB. Physical activity in
relation to cardiovascular disease and total mortality among men with
type 2 diabetes. Circulation. 2003;107:2435–2439.
32. Hemilä H, Kaprio J, Albanes D, Virtamo J. Physical activity and the risk
of pneumonia in male smokers administered vitamin E and beta-carotene.
Int J Sports Med. 2006;27:336 –341.
33. Morton LM, Hartge P, Holford TR, Holly EA, Chiu BC, Vineis P,
Stagnaro E, Willett EV, Franceschi S, La Vecchia C, Hughes AM, Cozen
W, Davis S, Severson RK, Bernstein L, Mayne ST, Dee FR, Cerhan JR,
Zheng T. Cigarette smoking and risk of non-Hodgkin lymphoma: a
pooled analysis from the International Lymphoma Epidemiology Consortium (InterLymph). Cancer Epidemiol Biomarkers Prev. 2005;14:
925–933.
34. Rimm EB, Williams P, Fosher K, Criqui M, Stampfer MJ. Moderate
alcohol intake and lower risk of coronary heart disease: meta-analysis of
effects on lipids and haemostatic factors. BMJ. 1999;319:1523–1528.
35. Klatsky AL. Alcohol, cardiovascular diseases and diabetes mellitus.
Pharmacol Res. 2007;55:237–247.
36. Koppes LL, Dekker JM, Hendriks HF, Bouter LM, Heine RJ. Moderate
alcohol consumption lowers the risk of type 2 diabetes: a meta-analysis of
prospective observational studies. Diabetes Care. 2005;28:719 –725.
37. Lee KW, Lip GY. Effects of lifestyle on hemostasis, fibrinolysis, and
platelet reactivity: a systematic review. Arch Intern Med. 2003;163:
2368 –2392.
38. Makelä P, Paljärvi T, Poikolainen K. Heavy and nonheavy drinking
occasions, all-cause and cardiovascular mortality and hospitalizations: a
follow-up study in a population with a low consumption level. J Stud
Alcohol. 2005;66:722–728.
39. Mumenthaler MS, Taylor JL, O’Hara R, Yesavage JA. Gender differences in moderate drinking effects. Alcohol Res Health. 1999;23:
55– 64.
40. Klatsky AL, Armstrong MA, Friedman GD. Risk of cardiovascular mortality in alcohol drinkers, ex-drinkers and nondrinkers. Am J Cardiol.
1990;66:1237–1242.
41. White IR, Altmann DR, Nanchahal K. Alcohol consumption and mortality: modelling risks for men and women at different ages. BMJ. 2002;
325:191.
42. Gmel G, Gutjahr E, Rehm J. How stable is the risk curve between alcohol
and all-cause mortality and what factors influence the shape? A precisionweighted hierarchical meta-analysis. Eur J Epidemiol. 2003;18:631– 642.
CLINICAL PERSPECTIVE
The association between alcohol consumption and decreased risk of coronary heart disease is well established, but possible
age differences are plausible because of potential pathogenic differences in coronary heart disease events occurring in
younger compared with middle-aged or older individuals. We studied these relations in a large population of men and
women 35 to 89 years of age at baseline. We observed a lower risk of coronary heart disease among men and women with
a light to moderate alcohol intake compared with nondrinkers, and this finding was consistent and of similar size in all age
groups. However, the absolute risk of CHD was small in the youngest age group, and risk differences between abstainers
and light to moderate alcohol consumers were of negligible size. Therefore, our results provide strong evidence for a lower
risk of coronary heart disease among moderate consumers relative to nondrinkers in younger, middle-aged, and older
adults; however, considering absolute risks across age groups, younger adults are not likely to benefit from an overall
recommendation of moderate alcohol intake.
Downloaded from circ.ahajournals.org by on April 9, 2010
The Pharmacogenomics Journal (2007), 1–8
& 2007 Nature Publishing Group All rights reserved 1470-269X/07 $30.00
www.nature.com/tpj
ORIGINAL ARTICLE
Alcoholism and alcohol drinking habits predicted
from alcohol dehydrogenase genes
Janne Schurmann Tolstrup1,2,
Børge Grønne Nordestgaard2,3,
Søren Rasmussen1,
Anne Tybjærg-Hansen4 and
Morten Grønbæk1,3
1
Center for Alcohol Research, National Institute of
Public Health, Copenhagen, Denmark;
2
Department of Clinical Biochemistry, Herlev
University Hospital, Herlev, Denmark; 3The
Copenhagen City Heart Study, Bispebjerg
University Hospital, Copenhagen, Denmark and
4
Department of Clinical Biochemistry,
Copenhagen University Hospital, Copenhagen,
Denmark
Correspondence:
Dr JS Tolstrup, Center for Alcohol Research,
National Institute for Public Health, Oster
Farimagsgade 5a, Copenhagen, 1399
Denmark.
E-mail: [email protected]
Alcohol drinking habits and alcoholism are partly genetically determined.
Alcohol is degraded primarily by alcohol dehydrogenase (ADH) wherein
genetic variation that affects the rate of alcohol degradation is found in
ADH1B and ADH1C. It is biologically plausible that these variations may be
associated with alcohol drinking habits and alcoholism. By genotyping 9080
white men and women from the general population, we found that men and
women with ADH1B slow vs fast alcohol degradation drank more alcohol and
had a higher risk of everyday drinking, heavy drinking, excessive drinking and
of alcoholism. For example, the weekly alcohol intake was 9.8 drinks (95%
confidence interval (CI): 9.1–11) among men with the ADH1B 1/1 genotype
compared to 7.5 drinks (95% CI: 6.4–8.7) among men with the ADH1B 1/2
genotype, and the odds ratio (OR) for heavy drinking was 3.1 (95% CI: 1.7–
5.7) among men with the ADH1B 1/1 genotype compared to men with the
ADH1B 1/2 genotype. Furthermore, individuals with ADH1C slow vs fast
alcohol degradation had a higher risk of heavy and excessive drinking. For
example, the OR for heavy drinking was 1.4 (95% CI: 1.1–1.8) among men
with the ADH1C 1/2 genotype and 1.4 (95% CI: 1.0–1.9) among men with
the ADH1B 2/2 genotype, compared with men with the ADH1C 1/1
genotype. Results for ADH1B and ADH1C genotypes among men and
women were similar. Finally, because slow ADH1B alcohol degradation is
found in more than 90% of the white population compared to less than 10%
of East Asians, the population attributable risk of heavy drinking and
alcoholism by ADH1B 1/1 genotype was 67 and 62% among the white
population compared with 9 and 24% among the East Asian population.
The Pharmacogenomics Journal advance online publication, 9 October 2007;
doi:10.1038/sj.tpj.6500471
Keywords: genetic association study; alcoholism; population-based study; alcohol; alcohol
dehydrogenase; acetaldehyde
Introduction
Alcoholism and alcohol drinking in general represent huge public health
problems in most countries worldwide, preventing many individuals from
successfully holding a job or looking after a family. In addition, excessive alcohol
use leads to diseases such as liver cirrhosis, chronic pancreatitis, upper
gastrointestinal cancers, cardiomyopathy, polyneuropathy and dementia. It has
been shown in twin studies that heritability explains approximately 50% of
alcoholism and problem drinking in the white population.1,2
The region surrounding the alcohol dehydrogenase (ADH) gene cluster is
Received 18 March 2007; revised 4 June 2007;
known to be associated with alcoholism from whole-genome scans.3 Well-known
accepted 13 June 2007
ADH-genotypes, alcohol drinking habits and alcoholism
JS Tolstrup et al
2
functional polymorphisms of ADH1B and ADH1C may
explain this finding because the ADH1B 2 vs the ADH1B 1
allele confer a 38-fold increase in in vitro alcohol degradation
rate (that is, the conversion of ethanol to acetaldehyde) and
the ADH1C 1 vs the ADH1C 2 allele confer a 2.5-fold
increase in in vitro alcohol degradation rate.4
During alcohol degradation, acetaldehyde is only found
in low concentrations. If concentrations become high, for
example, during treatment with disulfiram (used in some
countries to prevent alcohol intake among alcoholics) or in
individuals with a defective acetaldehyde dehydrogenase
(found among Asians), individuals experience severe nausea
and flushing and automatically abstain from drinking
alcohol. It is possible that similar but less pronounced
responses to alcohol are more likely to be produced among
individuals carrying the fast alcohol degradation ADH1B 2
and ADH1C 1 alleles compared to individuals carrying the
slow alcohol degradation ADH1B 1 and ADH1C 2 alleles. If
so, individuals with slow alcohol degradation may be able to
drink larger quantities of alcohol without experiencing
discomfort due to elevated acetaldehyde levels, and consequently are more likely to use alcohol excessively and to
develop alcoholism. This issue has been addressed in
different populations in case–control settings where allele
frequencies of ADH1B and ADH1C are compared between
alcoholics and nonalcoholics.5–28 To our knowledge, this
has not been studied in a prospective setting in the general
white population and it is unknown if the ADH1B and
ADH1C polymorphisms are associated with alcohol drinking
habits, such as amount of usual intake.
In the present study, we have genotyped a sample of 9080
men and women from the general white population to test
the hypotheses that slow alcohol degradation ADH1B 1 and
ADH1C 2 alleles are associated with alcohol drinking habits
and with increased risk of alcoholism.
Results
The frequencies of the ADH1B and ADH1C alleles coding for
slow alcohol degradation was 0.98 (ADH1B 1) and 0.42
(ADH1C 2) (Table 1). Genotypes were in Hardy–Weinberg
equilibrium (P ¼ 0.8 for ADH1B genotypes and P ¼ 0.7 for
ADH1C genotypes by w2-test). ADH1B 2 were associated
with ADH1C 1 (linkage disequilibrium coefficients D0 ¼ 0.90
and r2 ¼ 0.012).
Table 1
For ADH1B, we found that men and women who were
homozygous for the slow alcohol degrading ADH1B 1 allele
had a higher alcohol intake than men and women who were
fast alcohol degrading ADH1B 2 heterozygotes or homozygotes. For example, men with the ADH1B 1/1 genotype
drank on average 9.8 drinks per week (95% confidence
interval (CI): 9.1–11) and men with the ADH1B 1/2
genotype drank on average 7.5 drinks per week (95% CI:
6.4–8.7) (Table 2). Furthermore, odds for any, daily, heavy
and excessive alcohol drinking were 2–4 times higher
among men and women who were ADH1B 1 homozygotes
than among men and women who were ADH1B 2 heteroor homozygotes.
Using the brief Michigan Alcoholism Screening Test (brief
MAST) we found that men with the slow alcohol degradation ADH1B 1/1 genotype had a two- to fourfold risk of
alcoholism compared to men with the fast alcohol degradation ADH1B 2/1 or ADH1B 2/2 genotypes (Table 2). The
hazard ratio of hospitalization for alcoholism in men and
women with the slow alcohol degradation ADH1B 1/1
genotype was 3.9 (95% CI: 1.0–16) and 2.7 (95% CI: 0.4–20).
For ADH1C, odds for heavy and excessive alcohol drinking
were 40–70% higher among men who were hetero- or
homozygous for the slow alcohol degrading ADH1C 2 allele
than among men who were homozygous for the fast alcohol
degrading ADH1C 1 allele (Table 3). Similar results were
found in women; however, effect sizes were slightly smaller
and only statistically significant for heavy drinking. ADH1C
genotype was not associated with alcoholism (Table 3).
Analyses on daily, heavy and excessive drinking and on
alcoholism (MAST score and hospitalizations) were repeated
excluding consistent nondrinkers, that is, participants who
reported no alcohol intake at every examination in which
they participated (7.5% of the total study population). This
restriction did not affect any of our results (data not shown).
Because of linkage disequilibrium between the ADH1B 2
and ADH1C 1 alleles, and the relatively large effect on
enzyme activity of the ADH1B polymorphism, our results for
ADH1C could be influenced by ADH1B genotype. Therefore,
ADH1C analyses were repeated solely on individuals who
were ADH1B 1 homozygotes (95% of the study cohort): we
found similar results, indicating that the effect of ADH1C
genotype was independent of ADH1B genotype (data not
shown).
We also performed analyses on genotype combinations,
ranking genotypes in order of expected total enzyme
Distribution of ADH1B and ADH1C genotypes
ADH1B
ADH1C
Men (n ¼ 4039)
Women (n ¼ 5041)
1/1 (fast)
1/2 (intermediate)
2/2 (slow)
1/1 (fast)
1/2 (intermediate)
2/2 (slow)
1272
110
2
1882
70
1
697
5
0
1570
130
3
2347
81
0
907
3
0
1/1 (slow)
1/2 (intermediate)
2/2 (fast)
The Pharmacogenomics Journal
ADH-genotypes, alcohol drinking habits and alcoholism
JS Tolstrup et al
3
Table 2
Associations between ADH1B genotype and alcohol drinking habits and alcoholism
Men (n ¼ 4039)
ADH1B alcohol degradation
Women (n ¼ 5041)
1/2+2/2 (n ¼ 188) (fast)1/1 (n ¼ 3851) (slow)1/2+2/2 (n ¼ 217) (fast)1/1 (n ¼ 4824) (slow)
Alcohol drinking habits
Weekly alcohol intake (n ¼ 3784 M, 4052 W)a
Any alcohol intake (n ¼ 3784 M, 4052 W)b
Daily drinking (n ¼ 2015 M, 1259 W)b
Heavy drinking (n ¼ 1262 M, 814 W)b
Excessive drinking (n ¼ 507 M, 353 W)b
7.5
1.0
1.0
1.0
1.0
Alcoholism
bMAST score X5 (n ¼ 351 M, 105 W)b
bMAST score X10 (n ¼ 217 M, 55 W)b
Hospitalization (n ¼ 167 M, 74 W)c
1.0 (reference)
1.0 (reference)
1.0 (reference)
(6.4–8.7)
(reference)
(reference)
(reference)
(reference)
9.8
2.1
2.5
3.1
2.7
(9.1–11)
(1.0–4.5)
(1.5–4.1)
(1.7–5.7)
(1.1–6.5)
2.9 (1.3–6.6)
2.6 (1.0–7.1)
3.9 (1.0–16)
3.0
1.0
1.0
1.0
1.0
(2.6–3.5)
(reference)
(reference)
(reference)
(reference)
1.0 (reference)
1.0 (reference)
1.0 (reference)
4.0
3.1
1.9
3.0
4.2
(3.7–4.3)
(1.8–5.3)
(1.1–3.5)
(1.4–6.4)
(1.3–13)
1.7 (0.5–5.2)
2.8 (0.4–20)
2.7 (0.4–20)
Abbreviations: bMAST, the brief Michigan Alcoholism Screening Test; M, men; W, women.
Heavy drinking was defined as drinking more than 21 drinks per week for men and 14 drinks per week for women, and excessive drinking as drinking more than 35
drinks per week for men and 21 drinks per week for women. n indicates the number of men and women who are cases in respective analyses. Adjustment was made for
ADH1C genotype, age, years of school education and examination year.
a
Shown numbers are mean number of drinks per week (95% CI), which is estimated among those who report any alcohol intake.
b
Shown numbers are OR (95% CI).
c
Shown numbers are hazard ratios (95% CI).
Table 3
Associations between ADH1C genotype and alcohol drinking habits and alcoholism
Men (n ¼ 4039)
ADH1C alcohol
degradation
Alcohol drinking habits
Weekly alcohol
intake
(n ¼ 3784 M, 4052 W)a
Any alcohol intake
(n ¼ 3784 M, 4052 W)b
Daily drinking
(n ¼ 2015 M, 1259 W)b
Heavy drinking
(n ¼ 1262 M, 814 W)b
Excessive drinking
(n ¼ 507 M, 353 W)b
Women (n ¼ 5041)
1/1 (n ¼ 1384)
(fast)
1/2 (n ¼ 1953)
(intermediate)
2/2 (n ¼ 702) 1/1 (n ¼ 1703)
(slow)
(fast)
9.8 (9.1–11)
10.4 (9.4–12)
10.5 (9.4–12)
4.0 (3.7–4.3)
1/2 (n ¼ 2428)
(intermediate)
2/2 (n ¼ 910)
(slow)
4.1 (3.7–4.5)
4.1 (3.7–4.6)
1.0 (reference)
1.3 (0.9–1.8)
0.7 (0.5–1.2) 1.0 (reference)
1.1 (0.9–1.4)
1.0 (0.7–1.4)
1.0 (reference)
1.1 (0.9–1.4)
1.1 (0.8–1.5) 1.0 (reference)
1.2 (0.9–1.5)
1.2 (0.9–1.7)
1.0 (reference)
1.4 (1.1–1.8)
1.4 (1.0–1.9) 1.0 (reference)
1.3 (1.0–1.7)
1.2 (0.8–1.7)
1.0 (reference)
1.6 (1.1–2.3)
1.7 (1.1–2.6) 1.0 (reference)
1.4 (0.9–2.1)
1.0 (0.6–1.6)
1.0 (0.8–1.3)
1.0 (0.7–1.4) 1.0 (reference)
0.8 (0.5–1.2)
0.8 (0.5–1.5)
1.0 (0.7–1.3)
1.0 (0.7–1.6) 1.0 (reference)
0.6 (0.3–1.0)
0.7 (0.3–1.5)
1.2 (0.9–1.7)
0.9 (0.6–1.5) 1.0 (reference)
1.4 (0.8–2.6)
2.2 (1.2–4.2)
Alcoholism
1.0 (reference)
bMAST score X5
(n ¼ 1076 M, 597 W)b
bMAST score X10
1.0 (reference)
(n ¼ 290 M, 65 W)b
Hospitalization
1.0 (reference)
(n ¼ 209 M, 90 W)c
Abbreviations: bMAST, the brief Michigan Alcoholism Screening Test; M, men; W, women.
Heavy drinking was defined as drinking more than 21 drinks per week for men and 14 drinks per week for women, and excessive drinking as drinking more than 35
drinks per week for men and 21 drinks per week for women. n indicates the number of men and women who are cases in the respective analyses. Adjustment was made
for ADH1B genotype, age, years of school education and examination year.
a
Shown numbers are mean number of drinks per week (95% CI), which is estimated among those who report any alcohol intake.
b
Shown numbers are OR (95% CI).
c
Shown numbers are hazard ratios (95% CI).
The Pharmacogenomics Journal
ADH-genotypes, alcohol drinking habits and alcoholism
JS Tolstrup et al
4
Alcohol intake
P for trend <0.001
5
Odds ratio
Relative intake
1.5
1.0
Any alcohol intake
P for trend <0.04
4
3
2
1
0
5
Odds ratio
4
3
2
3
2
1
0
0
Excessive drinking
P for trend <0.001
3
4
3
2
Heavy drinking
P for trend <0.001
4
1
5
Odds ratio
Daily drinking
P for trend <0.001
Odds ratio
Odds ratio
5
Alcoholism: bMAST≥5
P for trend =0.09
2
1
1
0
Alcoholism: bMAST≥10
P for trend =0.24
3
Hazard ratio
Odds ratio
3
0
2
1
2
1
0
0
ADH1B
Enzyme activity
ADH1C
Enzyme activity
N
Alcoholism: hospitalisations
P for trend =0.01
2/1+2/2
Fast
1/1
Fast
245
2/1+2/2
Fast
2/1+2/2
Int/slow
160
1/1
Slow
1/1
Fast
2842
1/1
Slow
1/2
Int
4229
1/1
Slow
2/2
Slow
1604
ADH1B
Enzyme activity
ADH1C
Enzyme activity
N
2/1+2/2
Fast
1/1
Fast
245
2/1+2/2
Fast
2/1+2/2
Int/slow
160
1/1
Slow
1/1
Fast
2842
1/1
Slow
1/2
Int
4229
1/1
Slow
2/2
Slow
1604
Figure 1 Association between the combined ADH1B and ADH1C genotypes and alcohol drinking habits and alcoholism. The combined ADH1B and
ADH1C genotypes are ranked according to expected enzyme activity ((ADH1B 2/1 þ ADH1B 2/2, ADH1C 1/1) 4(ADH1B 2/1 þ ADH1B 2/2,
ADH1C 2/1 þ ADH1C 2/2) 4(ADH1B 1/1, ADH1C 1/1) 4(ADH1B 1/1, ADH1C 1/2) 4(ADH1B 1/1, ADH1C 2/2)). ADH1C 1/2 and ADH1C 2/2
were combined because of few subjects in these categories among the ADH1B 2 hetero- and homozygotes. Analyses are for men and women
combined and P-values are for linear trend tests.
activity, and tested for linear trend in each of the variables
for alcohol drinking habits and alcoholism (Figure 1). For
most end points, there was a statistically significant trend
test in the expected direction: individuals with slow vs fast
alcohol degradation drank more alcohol and more often,
and had a higher risk of alcoholism.
The Pharmacogenomics Journal
ADH1B genotypes are very differently distributed among
whites and East Asian populations (Figure 2). Among the
white population, more than 90% carry the ADH1B 1/1
genotype, coding for slow alcohol degradation, whereas
among the East Asian population less than 10% carry this
genotype. Hence, the population attributable risk of heavy
ADH-genotypes, alcohol drinking habits and alcoholism
JS Tolstrup et al
5
Whites
East Asians
Genotype
frequency
1.0
0.8
0.5
0.3
0.0
OR heavy
drinking
5.0
4.0
3.0
2.0
1.0
Ref.
Ref.
Ref.
Ref.
0.0
PAR heavy
drinking/%
80
60
40
20
0
OR alcoholism
5.0
4.0
3.0
2.0
1.0
0.0
PAR
alcoholism/%
80
60
40
20
0
ADH1B 2/2+2/1 1/1
Enzyme activity Fast Slow
2/2+2/1 1/1
Fast
Slow
Figure 2 Genotype frequencies, OR and PARs of heavy drinking and
alcoholism according to ADH1B genotype among white and East Asian
populations. Slow alcohol degradation ADH1B 1/1 genotype is predominant among Caucasians, but rare among East Asian population.
Risks of heavy drinking and alcoholism according to ADH1B 1/1 among
white and East Asian populations are comparable, but because of the
different genotype distributions in the two populations, PARs are much
higher among white population. For East Asian population, genotype
frequencies, OR and PARs are calculated from previous studies.29,30 Ref,
reference group; OR, odds ratio; PAR, population attributable risk.
drinking and alcoholism according to the ADH1B 1/1
genotype is 67 and 62% among white population compared
with 9 and 24% among the East Asian population.
Discussion
Our results suggest that ADH1B and ADH1C genotypes are
associated with alcoholism and alcohol drinking habits.
Men and women with ADH1B slow vs fast alcohol degradation drank more alcohol, were more often daily, heavy and
excessive drinkers and had higher risks of alcoholism.
Men and women with ADH1C intermediate and slow vs
fast alcohol degradation were more often heavy and
excessive drinkers. As expected, effect sizes were smaller
for ADH1C than for ADH1B, but given the high frequency of
the ADH1C 2 allele, it is nevertheless a very interesting
finding. Moreover, the impact of the ADH1C 2 allele on
cumulative lifetime alcohol intake may be significant. Our
results also suggest that ADH1B and ADH1C genotypes may
partly explain why white people generally drink more
alcohol than East Asian. The population risk of heavy
drinking and alcoholism attributable to the ADH1B 1/1
genotype was 67 and 62% among white population and
only 9 and 24% among the East Asian population.
Among men, we found relative estimates for alcoholism
from two to three among ADH1B 1 homozygotes, which is
comparable to results from a recent meta-analysis consisting
predominantly of East Asian studies.29 Also, the odds ratio
(OR) for heavy drinking for men was 3.1 which agrees with
what is previously found among Asian men.30 Separate
estimates for women were not available for any of the end
points in previous studies. Our results were remarkably
similar for men and women indicating that the investigated
associations are not sex-specific.
The various end points were probably subject to some
misclassification, that is, sensitivity and/or specificity less
than 100%. Measures of heavy and excessive drinking will
be misclassified if amount of alcohol intake is under- or
overreported, which it in many instances probably is. The
MAST score is not a perfect screenings tool for alcoholism
either. However, these errors occur most likely independently of genotype, and because end points are binary, will
lead to bias toward the null.31 Hence, we do not consider
misclassification of end points to have caused our results.
For alcoholism defined by hospital registry information,
sensitivity is likely considerably less than 100% because
many alcoholics are untreated or treated at private clinics
that are not included in the national registers. However,
specificity could be close to perfect; few nonalcoholics are
presumably diagnosed as alcoholics. In this scenario,
nondifferential misclassification is not affecting the hazard
ratio.32
Never-drinkers have not been exposed to alcohol and
hence, their drinking status cannot have been affected by
ADH1B and ADH1C genotypes. It was not possible to
separate never-drinkers from nondrinkers in this study, so
performing analyses without never-drinkers was not an
option. Potentially, this could have caused bias, but since
excluding consistent nondrinkers from analyses had virtually no impact on our results, we do not consider this a
major limitation.
We modeled the amount of alcohol intake in five different
ways and alcoholism in three different ways. Hence, several
statistical tests have been performed which in some
instances call for caution. However, outcomes in this study
were not independent but merely represent similar
The Pharmacogenomics Journal
ADH-genotypes, alcohol drinking habits and alcoholism
JS Tolstrup et al
6
outcomes modeled differently and we chose not to adjust
for multiple comparisons.
A likely explanation for our findings is that differences in
enzyme activity from the ADH1B and ADH1C polymorphisms
result in intraindividual differences in alcohol degradation
rate and that, for a given level of alcohol intake, individuals
with fast alcohol degradation have higher levels of acetaldehyde and thus more unpleasant symptoms compared
with individuals with slow alcohol degradation. However,
an effect of the polymorphisms on alcohol degradation rate
in vivo has been difficult to demonstrate,33 which may have
been due to insufficiently sensitive laboratory methods. In a
more recent study which applied a more refined method for
measuring the rate of alcohol degradation, results showed a
significant difference in degradation rate according to the
ADH1B genotype.34 In further support of an in vivo effect of
the ADH1B genotype is that individuals with the most active
enzymes are consistently reported to experience more
unpleasant symptoms such as flushing when drinking
alcohol compared to individuals with the less active
enzymatic forms.35–38
Our study had several strengths. First of all, sample size
was large and provided adequate power to study associations
between the relatively rare ADH1B 1/2 genotype and several
end points, and to detect the small effect sizes associated
with the ADH1C genotypes. Furthermore, participants were
men and women from the general population, all of Danish
descent. Hence, population stratification is unlikely to have
affected our results. Alcohol drinking habits were described
in several dimensions and information on alcoholism was
obtained from two independent sources (questionnaire and
hospital registry information). All end points were assessed
independently from genotyping and participants were
unaware of the purpose of this study when enrolled.
In conclusion, our data suggest that alcoholism and
alcohol drinking habits are partly predictable from ADH1B
and ADH1C genotypes. Results for men and women were
comparable and, as expected, effects of ADH1B were larger
than effects of ADH1C.
Methods
Study population
Our data originate from The Copenhagen City Heart Study,
which is a series of studies conducted in the Danish general
population. Examinations consisted of interview and physical
examination, and more especially, blood was given for DNA
purification at the examination that was performed during
1991–1994. All participants gave written consent and the
ethics committee for Copenhagen and Frederiksberg approved the study (no. 100.2039/91). Enrolment and examination procedures have been described in more detail
elsewhere.39,40 Of the 17 180 individuals who were invited
to the 1991–1994 examination, 10 135 participated, 9259
gave blood and 9222 were successfully genotyped. Eligibility
criterion for participation was Danish citizenship and
therefore, the Copenhagen City Heart Study does not reflect
The Pharmacogenomics Journal
the ethnic admixture of Copenhagen (the proportion of
inhabitants with foreign citizenship was 8% in 1994).
However, even a few participants of foreign ethnicity could
potentially confound our results since the fast alcohol
degradation ADH1B 2 allele is rare among white population
and quite frequent in other populations. Information on
ethnicity was not assessed at the examinations, and hence
information on birthplace was obtained from the Civil
Registration System. Participants born in Asia, Africa, the
Middle East, South America or Greenland were excluded
from further study (n ¼ 211). In all, 9080 individuals were
eligible for analyses, some of whom also participated in the
examinations during 1981–1983 (n ¼ 6615) and during
2001–2003 (n ¼ 4684).
Genotyping procedures
The ADH1B 2 allele (rs1229984, Arg47His in exon 3) and
ADH1C 2 allele (rs698, Ile3490Val in exon 8) were identified by means of duplex polymerase chain reaction followed
by Nanogen microelectronic chip technology (Nanogen
NMW 1000 Nanochip Molecular Biology Workstation)41
using standard conditions (details available from authors).
In a validation study, the accuracy of the Nanogen method
was found to be comparable to restriction fragment length
polymorphism.42
End points
Questions on drinking habits were included in the questionnaire at the examinations during 1981–1983, 1991–
1994 and 2001–2003. Amount of alcohol intake was
reported as usual intake of weekly beers, wine and spirits.
Assuming one drink to be equal to 12 g of pure alcohol, a
measure of total weekly intake was calculated. We defined
heavy drinking as drinking more than 21 drinks per week
for men and 14 drinks per week for women, and excessive
drinking as drinking more than 35 drinks per week for men
and 21 drinks per week for women.43 Participants were
defined as daily drinkers if they reported to drink alcohol
every day.
We defined alcoholism from questionnaire as well as from
hospital discharge information. The former definition was
taken from the 1991–1994 questionnaire, which included a
screening test for alcoholism (10 question version of the
brief MAST44). The test included questions such as ‘Do you
feel you are a normal drinker?’ and ‘Have you ever gone to
anyone for help about your drinking?’ Test scores of X5 and
of X10 were used as dichotomized end points. Information
on hospitalizations for alcoholism was obtained from the
Danish Hospital Discharge Register where all hospitalizations in Denmark, classified according to the World Health
Organization’s International Classification of Diseases (ICD)
are registered.45 The following diagnoses indicative of
hospitalization due to alcoholism were obtained: ICD-8
codes 303.09–303.99 and ICD-10 codes F10.1–F10.4.
Statistical analyses
All statistical models included ADH1B and ADH1C genotypes, age and years of school education using the SAS/Stat
ADH-genotypes, alcohol drinking habits and alcoholism
JS Tolstrup et al
7
software (version 8.02). ADH1B 2 heterozygotes were
combined with ADH1B 2 homozygotes (n ¼ 6). Estimated
haplotype frequencies were calculated by Hplus.46,47 Linkage disequilibrium was expressed as r2 and D0 .48,49
To study the association between ADH1B and ADH1C
genotypes and amount of alcohol intake, the correlated
mixed distribution model was applied (Mixcorr macro50).
This model handles data with clumping at zero and a
lognormal distribution for nonzero values, and contains
components to model the probability of a nonzero value
and the mean of nonzero values, allowing for repeated
measurements using random effects.50 This means that if a
variable affects the mean amount by affecting both the
probability of occurrence of a nonzero value and also the
mean of a nonzero value, these effects can be separated and
quantified. Hence, two estimates are produced from this
model: the OR for having a nonzero alcohol intake (that is,
for not being a non-drinker) and the mean amount of
alcohol intake among those with a nonzero intake.
To study the association between ADH1B and ADH1C
genotypes and daily, heavy and excessive drinking, we
applied logistic regression allowing for repeated measurements using random effects (proc nlmixed). We applied
unconditional logistic regression to study associations
between ADH1B and ADH1C genotypes and dichotomized
brief MAST score (proc genmod).
Risk estimates for alcoholism defined from hospitalizations were computed by means of Cox proportional hazard
regression (proc phreg). Age was used as the time axis and
analyses were corrected for delayed entry. Vital status of the
participants was obtained from the National Central Person
Register. The observation time for each participant was the
period from participation in the Copenhagen City Heart
Study, until date of alcoholism, death, emigration outside
Denmark or January 1, 2004, whichever came first. We had
100% follow-up.
Population attributable risk was calculated as (proportion
of exposed in the population (OR1))/(proportion of
exposed in the population (OR1) þ 1).51
Abbreviations
ADH
bMAST
OR
PAR
Ref
alcohol dehydrogenase
brief Michigan Alcoholism Screening Test
odds ratio
population attributable risk
Reference group
Acknowledgments
This work was supported by grants from the Danish Graduate
School of Public Health, the Health Insurance Foundation, the
Ministry of the Interior and Health and the Danish National Board
of Health and The Danish Heart Foundation, The Danish Medical
Research Council, The Copenhagen County Research Foundation
and Chief Physician Johan Boserup and Lise Boserups Foundation.
We thank the participants of the Copenhagen City Heart Studies.
Duality of interest
None declared.
References
1 Han C, McGue MK, Iacono WG. Lifetime tobacco, alcohol and other
substance use in adolescent Minnesota twins: univariate and multivariate behavioral genetic analyses. Addiction 1999; 94: 981–993.
2 Kendler KS, Neale MC, Heath AC, Kessler RC, Eaves LJ. A twin-family
study of alcoholism in women. Am J Psychiatry 1994; 151: 707–715.
3 Saccone NL, Kwon JM, Corbett J, Goate A, Rochberg N, Edenberg HJ et
al. A genome screen of maximum number of drinks as an alcoholism
phenotype. Am J Med Genet 2000; 96: 632–637.
4 Bosron WF, Li TK. Genetic polymorphism of human liver alcohol and
aldehyde dehydrogenases, and their relationship to alcohol metabolism
and alcoholism. Hepatology 1986; 6: 502–510.
5 Vidal F, Lorenzo A, Auguet T, Olona M, Broch M, Gutierrez C et al.
Genetic polymorphisms of ADH2, ADH3, CYP4502E1 Dra-I and Pst-I,
and ALDH2 in Spanish men: lack of association with alcoholism and
alcoholic liver disease. J Hepatol 2004; 41: 744–750.
6 Konishi T, Luo HR, Calvillo M, Mayo MS, Lin KM, Wan YJ. ADH1B*1,
ADH1C*2, DRD2 (141C Ins), and 5-HTTLPR are associated with
alcoholism in Mexican American men living in Los Angeles. Alcohol Clin
Exp Res 2004; 28: 1145–1152.
7 Wall TL, Carr LG, Ehlers CL. Protective association of genetic variation in
alcohol dehydrogenase with alcohol dependence in native American
Mission Indians. Am J Psychiatry 2003; 160: 41–46.
8 Osaka R, Nanakorn S, Sakata R, Nishiyori A, Shibata A, Nakamura J et al.
Alcohol dehydrogenase-2 and aldehyde dehydrogenase-2 genotypes
and male alcohol use disorders in Khon Kaen, north-east Thailand.
Psychiatry Clin Neurosci 2003; 57: 37–45.
9 Chao YC, Wang SJ, Chu HC, Chang WK, Hsieh TY. Investigation of
alcohol metabolizing enzyme genes in Chinese alcoholics with avascular
necrosis of hip joint, pancreatitis and cirrhosis of the liver. Alcohol Alcohol
2003; 38: 431–436.
10 Frenzer A, Butler WJ, Norton ID, Wilson JS, Apte MV, Pirola RC et al.
Polymorphism in alcohol-metabolizing enzymes, glutathione S-transferases and apolipoprotein E and susceptibility to alcohol-induced
cirrhosis and chronic pancreatitis. J Gastroenterol Hepatol 2002; 17:
177–182.
11 Chambers GK, Marshall SJ, Robinson GM, Maguire S, Newton-Howes J,
Chong NL. The genetics of alcoholism in Polynesians: alcohol and
aldehyde dehydrogenase genotypes in young men. Alcohol Clin Exp Res
2002; 26: 949–955.
12 Ogurtsov PP, Garmash IV, Miandina GI, Guschin AE, Itkes AV, Moiseev
VS. Alcohol dehydrogenase ADH2-1 and ADH2-2 allelic isoforms in the
Russian population correlate with type of alcoholic disease. Addict Biol
2001; 6: 377–383.
13 Lee HC, Lee HS, Jung SH, Yi SY, Jung HK, Yoon JH et al. Association
between polymorphisms of ethanol-metabolizing enzymes and susceptibility to alcoholic cirrhosis in a Korean male population. J Korean
Med Sci 2001; 16: 745–750.
14 Borras E, Coutelle C, Rosell A, Fernandez-Muixi F, Broch M, Crosas B et
al. Genetic polymorphism of alcohol dehydrogenase in Europeans: the
ADH2*2 allele decreases the risk for alcoholism and is associated with
ADH3*1. Hepatology 2000; 31: 984–989.
15 Amadéo S, Noble EP, Fourcade-Amadéo ML, Tetaria C, Brugiroux MF,
Nicolas L et al. Association of D2 dopamine receptor and alcohol
dehydrogenase 2 genes with polynesian alcoholics. Eur Psychiatry 2000;
15: 97–102.
16 Chao YC, Wang LS, Hsieh TY, Chu CW, Chang FY, Chu HC. Chinese
alcoholic patients with esophageal cancer are genetically different from
alcoholics with acute pancreatitis and liver cirrhosis. Am J Gastroenterol
2000; 95: 2958–2964.
17 Rodrigo L, Alvarez V, Rodriguez M, Perez R, Alvarez R, Coto E.
N-acetyltransferase-2, glutathione S-transferase M1, alcohol dehydrogenase, and cytochrome P450IIE1 genotypes in alcoholic liver cirrhosis:
a case–control study. Scand J Gastroenterol 1999; 34: 303–307.
The Pharmacogenomics Journal
ADH-genotypes, alcohol drinking habits and alcoholism
JS Tolstrup et al
8
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Osier M, Pakstis AJ, Kidd JR, Lee JF, Yin SJ, Ko HC et al. Linkage
disequilibrium at the ADH2 and ADH3 loci and risk of alcoholism. Am J
Hum Genet 1999; 64: 1147–1157.
Chen CC, Lu RB, Chen YC, Wang MF, Chang YC, Li TK et al. Interaction
between the functional polymorphisms of the alcohol-metabolism genes
in protection against alcoholism. Am J Hum Genet 1999; 65: 795–807.
Whitfield JB, Nightingale BN, Bucholz KK, Madden PA, Heath AC,
Martin NG. ADH genotypes and alcohol use and dependence in
Europeans. Alcohol Clin Exp Res 1998; 22: 1463–1469.
Shen YC, Fan JH, Edenberg HJ, Li TK, Cui YH, Wang YF et al.
Polymorphism of ADH and ALDH genes among four ethnic groups in
China and effects upon the risk for alcoholism. Alcohol Clin Exp Res 1997;
21: 1272–1277.
Espinos C, Sanchez F, Ramirez C, Juan F, Najera C. Polymorphism of
alcohol dehydrogenase genes in alcoholic and nonalcoholic individuals
from Valencia (Spain). Hereditas 1997; 126: 247–253.
Chen WJ, Loh EW, Hsu YP, Cheng AT. Alcohol dehydrogenase and
aldehyde dehydrogenase genotypes and alcoholism among Taiwanese
aborigines. Biol Psychiatry 1997; 41: 703–709.
Nakamura K, Iwahashi K, Matsuo Y, Miyatake R, Ichikawa Y, Suwaki H.
Characteristics of Japanese alcoholics with the atypical aldehyde
dehydrogenase 2*2. I. A comparison of the genotypes of ALDH2,
ADH2, ADH3, and cytochrome P-4502E1 between alcoholics and
nonalcoholics. Alcohol Clin Exp Res 1996; 20: 52–55.
Tanaka F, Shiratori Y, Yokosuka O, Imazeki F, Tsukada Y, Omata M. High
incidence of ADH2*1/ALDH2*1 genes among Japanese alcohol dependents and patients with alcoholic liver disease. Hepatology 1996; 23:
234–239.
Higuchi S, Muramatsu T, Matsushita S, Murayama M, Hayashida M.
Polymorphisms of ethanol-oxidizing enzymes in alcoholics with inactive
ALDH2. Hum Genet 1996; 97: 431–434.
Chen WJ, Loh EW, Hsu YP, Chen CC, Yu JM, Cheng AT. Alcoholmetabolising genes and alcoholism among Taiwanese Han men:
independent effect of ADH2, ADH3 and ALDH2. Br J Psychiatry 1996;
168: 762–767.
Maezawa Y, Yamauchi M, Toda G, Suzuki H, Sakurai S. Alcoholmetabolizing enzyme polymorphisms and alcoholism in Japan. Alcohol
Clin Exp Res 1995; 19: 951–954.
Zintzaras E, Stefanidis I, Santos M, Vidal F. Do alcohol-metabolizing
enzyme gene polymorphisms increase the risk of alcoholism and
alcoholic liver disease? Hepatology 2006; 43: 352–361.
Sun F, Tsuritani I, Yamada Y. Contribution of genetic polymorphisms in
ethanol-metabolizing enzymes to problem drinking behavior in middleaged Japanese men. Behav Genet 2002; 32: 229–236.
Rothman K. Biases in Study Design. Epidemiology—An Introduction.
Oxford University Press: New York, 2002: 94–112.
Rothman K, Greenland S. Precision and Validity in Epidemiological Studies
Modern Epidemiology.2nd edn, Lippincott, Williams and Wilkins,
Philadelphia, USA, 1998, pp 115–134.
Mizoi Y, Yamamoto K, Ueno Y, Fukunaga T, Harada S. Involvement of
genetic polymorphism of alcohol and aldehyde dehydrogenases in
individual variation of alcohol metabolism. Alcohol Alcohol 1994; 29:
707–710.
The Pharmacogenomics Journal
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Neumark YD, Friedlander Y, Durst R, Leitersdorf E, Jaffe D, Ramchandani
VA et al. Alcohol dehydrogenase polymorphisms influence alcoholelimination rates in a male Jewish population. Alcohol Clin Exp Res 2004;
28: 10–14.
Yokoyama T, Yokoyama A, Kato H, Tsujinaka T, Muto M, Omori T et al.
Alcohol flushing, alcohol and aldehyde dehydrogenase genotypes, and
risk for esophageal squamous cell carcinoma in Japanese men. Cancer
Epidemiol Biomarkers Prev 2003; 12: 1227–1233.
Carr LG, Foroud T, Stewart T, Castelluccio P, Edenberg HJ, Li TK.
Influence of ADH1B polymorphism on alcohol use and its subjective
effects in a Jewish population. Am J Med Genet 2002; 112: 138–143.
Takeshita T, Yang X, Morimoto K. Association of the ADH2 genotypes
with skin responses after ethanol exposure in Japanese male university
students. Alcohol Clin Exp Res 2001; 25: 1264–1269.
Takeshita T, Mao XQ, Morimoto K. The contribution of polymorphism
in the alcohol dehydrogenase beta subunit to alcohol sensitivity in a
Japanese population. Hum Genet 1996; 97: 409–413.
The Copenhagen City Heart Study. Osterbroundersogelsen. A book of
tables with data from the first examination (1976–78) and a five year
follow-up (1981–83). Scand J Soc Med (Suppl) 1989; 41: 1–160.
Schnohr P, Jensen G, Lange P, Scharling H, Appleyard M. The
Copenhagen City Heart Study—Østerbroundersøgelsen. Tables with
data from the third examination 1991–94. Eur Heart J (Suppl) 2001; 3:
H1–H83.
Sosnowski RG, Tu E, Butler WF, O’Connell JP, Heller MJ. Rapid
determination of single base mismatch mutations in DNA hybrids by
direct electric field control. Proc Natl Acad Sci USA 1997; 94: 1119–
1123.
Sethi AA, Tybjaerg-Hansen A, Andersen RV, Nordestgaard BG. Nanogen
microelectronic chip for large-scale genotyping. Clin Chem 2004; 50:
443–446.
Jensen MK, Sorensen TI, Andersen AT, Thorsen T, Tolstrup JS,
Godtfredsen NS et al. A prospective study of the association between
smoking and later alcohol drinking in the general population. Addiction
2003; 98: 355–363.
Pokorny AD, Miller BA, Kaplan HB. The brief MAST: a shortened version
of the Michigan Alcoholism Screening Test. Am J Psychiatry 1972; 129:
342–345.
Jürgensen HJ, Frølund C, Gustafsen J, Mosbech H, Guldhammer B,
Mosbech J. Registration of diagnoses in the Danish National Registry of
patients. Methods Inf Med 1986; 25: 158–164.
Hplus. Fred Hutchinson Cancer Research Center: Seattle, WA, 2003.
Li SS, Khalid N, Carlson C, Zhao LP. Estimating haplotype frequencies
and standard errors for multiple single nucleotide polymorphisms.
Biostatistics 2003; 4: 513–522.
Lewontin RC. The interaction of selection and linkage. I. General
considerations: heterotic models. Genetics 1964; 49: 49–67.
Pritchard JK, Przeworski M. Linkage disequilibrium in humans: models
and data. Am J Hum Genet 2001; 69: 1–14.
Tooze JA, Grunwald GK, Jones RH. Analysis of repeated measures data
with clumping at zero. Stat Methods Med Res 2002; 11: 341–355.
Walter SD. Calculation of attributable risks from epidemiological data.
Int J Epidemiol 1978; 7: 175–182.
Scandinavian Journal of Public Health, 2010; 38: 489–494
ORIGINAL ARTICLE
Alcohol drinking habits, alcohol dehydrogenase genotypes
and risk of acute coronary syndrome
JANNE S. TOLSTRUP1, JANE L. HANSEN1, MORTEN GRØNBÆK1, ULLA VOGEL2,
ANNE TJØNNELAND3, ALBERT MARNI JOENSEN4 & KIM OVERVAD4,5
1
Center for Alcohol Research, National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark,
National Food Institute, Technical University of Denmark and Institute of Science, Systems and Models, Roskilde University,
Denmark, 3Institute of Cancer Epidemiology, Danish Cancer Society, Copenhagen, Denmark, 4Department of Cardiology,
Aalborg Hospital, Aarhus University Hospital, Aalborg, Denmark, and 5Department of Clinical Epidemiology, Aarhus
University Hospital, Aalborg, Denmark
2
Abstract
Aims: The risk of myocardial infarction is lower among light-to-moderate drinkers compared with abstainers. Results from
some previous studies, but not all, suggest that this association is modified by variations in genes coding for alcohol
dehydrogenase (ADH). We aimed to test this hypothesis, including alcohol as both the amount of alcohol and the frequency
of drinking. Methods: we conducted a nested case-cohort study within the Danish Diet, Cancer and Health study, including
1,645 men (770 incident cases of acute coronary syndrome from 1993–1997 through 2004 and 875 randomly selected
controls). Results: Higher alcohol intake (measured as amount or drinking frequency) was associated with lower risk of acute
coronary syndrome; however, there was no evidence that these finding were modified by ADH1B or ADH1C genotypes.
Conclusions: The importance of functional variation in alcohol dehydrogenase for the association between
alcohol drinking habits and the risk of developing acute coronary syndrome, if any, is very limited.
Key Words: Acute coronary syndrome, alcohol, cohort study, genetic epidemiology
Background
Moderate alcohol intake of approximately 1–2 drinks
per day for women and 2–3 drinks per day for men is
associated with a lower risk of coronary heart disease
[1–4]. It has been suggested that this association is
modified by variations in genes coding for alcoholdegrading enzymes: individuals carrying genotypes
that code for enzymes giving a slow alcohol degradation may have lower risks for a given level of
alcohol intake because they have alcohol in the blood
for a longer time period compared with individuals
with faster alcohol degradation.
Alcohol degradation is mainly catalyzed by alcohol
dehydrogenase (ADH). Different ADHs exist, but
the ADH1C gene is especially interesting because
of a commonly occurring polymorphism among
individuals of Caucasian origin (frequencies of the
two alleles are 58% and 42%). In vitro, there is a
2.5-fold difference in activity between the enzymes
that are coded at ADH1C1 and ADH1C2 [5,6].
Another alcohol dehydrogenase gene, ADH1B, has
two alleles that produce enzymes with a 38-fold
difference in alcohol degradation rate [5]. In human
studies, however, differences in alcohol degradation
rate are much more modest and not even consistently
observable [7–12]. In Caucasians, the frequency of
the most active allele (ADH1B2) is around 2% [6].
Although results from the first study on the subject
indicated that ADH1C slow alcohol degradation was
associated with a lower risk of myocardial infarction
than ADH1C fast alcohol degradation [13], later
results have been less convincing [14–18]. An explanation for this discrepancy could be that the initial
Correspondence: Janne S. Tolstrup, National Institute of Public Health, University of Southern Denmark, Oster Farimagsgade 5a, DK-1399 Copenhagen K,
Denmark. E-mail: [email protected]
(Accepted 7 April 2010)
ß 2010 the Nordic Societies of Public Health
DOI: 10.1177/1403494810371248
490
J. S. Tolstrup et al.
finding was due to chance, because the effect seemingly was confined to one category only, comprising
five cases and 37 controls. Another explanation could
be that the association between alcohol intake,
genotype and heart disease depends upon the drinking pattern: Among individuals who tend to drink
large quantities of alcohol at a time, the genotype may
be of less importance whereas among individuals
with a more frequent drinking pattern, the genotype
may be important for the physiological dose of
alcohol. A measure of drinking pattern has not been
included in any of the previous studies.
In this study, we aim at testing whether the
association between alcohol intake and risk of acute
coronary syndrome is modified by the ADH1B or
ADH1C genotypes while taking into account both the
amount and frequency of alcohol intake.
Material and methods
The Diet, Cancer and Health Study
During December 1993 to May 1997, 160,725
Danish men and women aged 50 to 65 years were
invited by mail to participate in the population-based
cohort study Diet, Cancer and Health [19]. Eligible
cohort members were born in Denmark, living in the
Copenhagen and Aarhus area, and had no previous
cancers. A visit at the study clinic was appointed by
telephone with people who agreed to participate
(27,178 men and 29,875 women (35%)) and a 192item food frequency questionnaire including questions on amount of alcohol intake was subsequently
sent by mail. The food frequency questionnaire was
scanned and interviewer-checked during the clinic
visit, where another questionnaire concerning lifestyle and background factors, such as smoking habits,
physical activity, and education was filled in. Height
and weight were also measured at baseline and a
blood sample was obtained. A description of the
development and validation of the food frequency
questionnaire has been published previously [20,21].
The protocol was approved by the Scientific Ethical
Committee (KF 01-116/96).
In the food frequency questionnaire, alcohol intake
was reported as average (over the preceding year)
intake of each beverage: Light, normal and strong
beer (in number of bottles); red, white and fortified
wine (in number of glasses); and spirits (in number of
drinks). The predefined responses were in 12 categories, ranging from ‘‘never’’ to ‘‘eight or more per
day’’. The alcohol content in the different beverage
types was defined as one bottle of light beer ¼ 8.9 g
ethanol; one bottle of regular beer or one glass
of wine ¼ 12.2 g ethanol; one bottle of strong
beer ¼ 17.5 g ethanol; one drink of fortified wine ¼
9.3 g ethanol; and one drink of spirits ¼ 9.9 g ethanol.
These categories were converted into number of
standard drinks, defined as containing 12 g ethanol,
and added to yield an average measure of amount of
alcohol intake.
Information on drinking frequency was obtained
from the background questionnaire, in which participants were asked about the frequency of alcohol
intake in seven possible response categories: Never
drink alcohol, less than once per month, one to three
times per month, once a week, two to four times per
week, five to six times per week, and daily.
Laboratory analysis
The ADH1B2 allele (rs1229984, Arg47His in exon
3) and ADH1C2 allele (rs1693482, Arg270Gln in
exon 8) were identified by means of Taqman allelic
discrimination (ABI 7300, Applied Biosystems).
DNA was isolated from frozen lymphocytes [22].
In short, 100 mg DNA were obtained from 107
lymphocytes and 20 ng of DNA were genotyped in
5 ml containing 1 Mastermix (Applied Biosystems,
Denmark), 100 nM probes, and 900 nM primers.
The case status of the samples was blinded during
laboratory analysis. Controls were included in each
run, and repeated genotyping of a random 10%
subset yielded 100% identical genotypes.
Endpoint and study design
Information on acute coronary syndrome (including
unstable angina, non-fatal and fatal myocardial
infarction) was obtained by linking the participants
(via the unique identification number assigned to all
Danish citizens) with central Danish registries of
hospital discharge diagnoses and causes of death
(ICD-8 codes 410-410.99, 427.27 and ICD-10
codes I20.0, I21.x, I46.x). A case-cohort study was
designed using incident validated cases as the endpoint. Cases were identified and verified between
baseline and 1 January 2004, the date of the last
available update from the Hospital Discharge
Register [23]. Consistent with the case-cohort
design [24], controls were selected from the entire
cohort at random. The present study includes men
only, because the number of incident cases among
women was not sufficiently high to perform analyses
considering alcohol intake and genotype simultaneously. In total, 770 cases of ACS and 875 controls
were successfully genotyped and thus constitute the
sample for statistical analysis.
Drinking habits, alcohol genotypes and acute coronary syndrome
Statistical approach
Hazard ratios (HRs) and 95% confidence intervals
(CIs) for the association between genotype and acute
coronary syndrome were estimated using Cox proportional hazard regression with Kalbfleisch and
Lawless weights and robust variance estimates suitable for case-cohort data [25]. Age was used as the
underlying time axis to ensure that the estimation
procedure was based on comparisons of individuals
of the same age and hence remove confounding by
age. We calculated age-adjusted hazard ratios and
hazard ratios adjusted for known risk factors for acute
coronary syndrome: Smoking (never, ex-smoker,
1–14, 15–24, or 24þ g tobacco per day), body mass
index (BMI, <25, 25–30, 30þ kg/m2), physical
activity in leisure time (hours per week), school
education (7, 8–10, or 10þ years in school), fruit
(quartiles), vegetables (linearly), fish (linearly), fat
(quartiles, % of total energy intake) and saturated fat
(% of total energy intake, linearly). Tests for linear
trend were performed by treating the median within
categories as a continuous variable.
We used the chi-square test to determine whether
the ADH1B and ADH1C genotypes were in HardyWeinberg equilibrium [26]. Tests for interaction were
performed by comparing a model including main
effects of alcohol and ADH1C genotypes with a
model also including the interaction terms by a log
likelihood test i.e. on a multiplicative scale. Analyses
were performed using Stata 10.1 (Stata Corp.,
College Station, TX).
Results
Men who experienced an event of acute coronary
syndrome tended to drink less alcohol, to smoke
more and to have shorter education (Table I).
ADH1B genotypes were in Hardy-Weinberg equilibrium in the comparison group (p ¼ 0.58) as well as
in the whole sample (p ¼ 0.46). The same applied
for the ADH1C genotype (p ¼ 0.56 and 0.29).
491
Allele frequencies were 0.98 (95% CI: 0.97–0.98)
(ADH1B1 slow), 0.02 (95% CI: 0.02–0.03) (ADH1B2
fast) and 0.61 (95% CI: 0.59–0.62) (ADH1C1, fast)
and 0.39 (95% CI: 0.38–0.41) (ADH1C2, slow).
Higher alcohol intake was associated with lower
risk of acute coronary syndrome: compared with men
who drank 1–6 drinks/week, the risk among men who
drank 7–20 and 21þ drinks/week was 0.78 (0.59–
1.03) and 0.70 (0.51–0.97) ((p for linear trend ¼
0.02). Within categories of total alcohol intake,
ADH1C genotypes were not consistently associated
with risk of acute coronary syndrome and in particular there was no general tendency that the risk
among men with the slow ADH1C2/2 genotype was
lower than among men with the faster ADH1C1/1 or
ADH1C1/2 genotypes (Table II). Consistent with
Table I. Characteristics by sex and acute coronary syndrome
status in 2,667 participants of the Diet, Cancer and Health Study.
Characteristic
Cases (n ¼ 770)
Controls
(n ¼ 875)
Age, yrsa
Alcohol, drinks/weeka
ADH1B genotype, %
1/1 (slow)
1/2 (fast)
ADH1B genotype, %
1/1 (fast)
1/2 (intermediate)
2/2 (slow)
Smokers, % current
BMI, kg/m2 a
Physical Activity,
hours/weeka
Education, % 7 yrs
Vegetables, grams/daya
Fruit, grams/daya
Saturated fat, % of total
energya intake
Fish, grams/daya
58 (51–65)
9.7 (0.3–42.2)
56 (51–64)
11.8 (0.8–48.9)
4.2
95.8
4.6
95.4
37
47
16
59
27 (22–34)
2 (0–11)
38
46
16
38
26 (22–32)
2 (0–10)
46
132 (38–331)
124 (17–465)
13 (6–17)
34
155 (49–373)
143 (22–518)
12 (8–16)
41 (12–95)
41 (11–104)
BMI, body mass index. aNumbers represent medians (5th and
95th percentiles).
Table II. Hazard ratios,a 95% confidence intervals of acute coronary syndrome and number of cases by alcohol intake and ADH1C
genotype.
ADH1C genotype (relative enzyme activity)
Alcohol intake (drinks/week)
<1
1–6
7–20
21þ
1/1 (Fast)
0.96 (0.47–1.93) (23)
1 (reference) (79)
0.88 (0.56–1.39) (99)
0.97 (0.59–1.59) (84)
1/2 (Intermediate)
1.86
1.38
0.97
0.73
(0.94–3.65)
(0.87–2.19)
(0.62–1.51)
(0.45–1.19)
(35)
(118)
(122)
(87)
2/2 (Slow)
1.45
1.10
0.91
0.84
(0.47–4.47)
(0.59–2.08)
(0.52–1.58)
(0.46–1.54)
(10)
(33)
(45)
(35)
a
Adjusted for smoking, BMI, physical activity, school education, intake of vegetables, fruit, saturated fat and fish. Test for interaction
between alcohol intake and ADH1C: 2 ¼ 0.35, p ¼ 0.95.
492
J. S. Tolstrup et al.
this observation, the p value for interaction between
alcohol and ADH1C genotype was 0.95. For the
ADH1B genotype, the statistical power to perform
similar analyses was limited, but results among men
drinking seven or more drinks/week did not indicate
that ADH1B slow metabolizers were at lower risk
than ADH1B fast metabolizers (data not shown).
The p value for interaction between alcohol and
ADH1B genotype was 0.75.
Higher drinking frequency was associated with a
lower risk of acute coronary syndrome among men
(p for linear trend ¼ 0.005) (Figure 1). Within categories of drinking frequency, the hazard ratios
according to the ADH1C genotype were comparable
(Figure 2). The p value for interaction between
drinking frequency and ADH1C genotype was 0.88.
For the ADH1B genotype, there seemed to be no
difference in risk of acute coronary syndrome according to drinking frequency either, even though the
statistical power was low because of the relatively low
4
Hazard ratio
2
p for trend=0.005
1
0.5
0.25
Never
<1
1
2–4
5–6
7
Alcohol drinking frequency (days/week)
Cases
Controls
30
16
103
84
86
86
267
295
88
152
185
232
Figure 1. Hazard ratio of acute coronary syndrome according to
drinking frequency. Vertical bars represent 95% CIs.
Hazard ratio
4
2
ADH1C·1/1 (fast)
ADH1C·2/1
(intermediate)
ADH1C·2/2 (slow)
1
0.5
0.25
<1
2–4
5+
Alcohol drinking frequency (days/week)
Cases
Controls
79 112 28
72 83 31
100 116 51
116 131 48
102 128 43
142 181 61
Figure 2. Hazard ratio of acute coronary syndrome according to
drinking frequency and ADH1C genotype. Vertical bars
represent 95% CIs.
frequency of the ADH1B2 allele (2%) (Table III).
The p value for interaction between drinking frequency and ADH1B genotypes was 0.87.
Discussion
In this study, we found no evidence that variation in
ADH1B and ADH1C genotypes are modifying associations between alcohol intake (amount or drinking
frequency) and acute coronary syndrome. A moderate alcohol intake of 1–2 drinks per day for women
and 2–3 drinks per day for men is consistently shown
to be associated with a decreased risk of coronary
heart disease; at higher intakes, there seem to be no
further risk reduction and some studies have even
reported an increase in risk. In this study cohort,
however, no increased risk has been observed even
among the most heavy drinkers (28 drinks/week for
women and 35 drinks/week for men) [27].
Limitations of the present study include that
information on alcohol intake was obtained only
once; hence it is possible that some participants have
changed their alcohol habits during follow-up, resulting in misclassification. However, since information
on alcohol intake (and other variables) was collected
prospectively, any misclassification is most likely
independent of disease status. Also, even though
the number of cases was large, we did not have
sufficient power to pick up very small effects of the
ADH1B and ADH1C genotypes.
Our study had several strengths. The follow-up
study design with almost complete follow-up made it
unlikely that selection bias affected the study. The
outcome information was obtained and validated
independently of alcohol intake. Information bias is
therefore of no concern. Furthermore, participants
were from the general population of Danish descent.
Hence, population stratification is unlikely to have
affected our results.
If the effect of alcohol on coronary heart disease is
modified by variations in ADH genes it really comes
down to the question of whether the risk is lower
Table III. Hazard ratios,a 95% confidence intervals of acute
coronary syndrome and number of cases by alcohol intake and
ADH1B genotype.
Drinking
frequency
(days/week)
1/2 (Fast)
<2
2þ
1 (reference) 10
0.56 (0.15–1.98) (22)
a
ADH1B genotype (relative enzyme activity)
1/1 (Slow)
1.08 (0.36–3.28) (209)
0.78 (0.26–2.33) (518)
Adjusted for smoking, BMI, physical activity, school education,
intake of vegetables, fruit, saturated fat and fish. Test for interaction between alcohol intake and ADH1B: 2 ¼ 0.03, p ¼ 0.87.
Drinking habits, alcohol genotypes and acute coronary syndrome
among slow metabolizers than among fast metabolizers for a given level of alcohol intake, because the
slow metabolizers may have alcohol in the blood for a
longer time period compared with the fast metabolizers. Initially, results in American physicians indicated that this is so [13]. However, this finding was
based on only five cases and 37 controls in the highest
alcohol drinking category; for a lower alcohol intake
there seemed to be no difference in risk of myocardial
infarction according to genotype. Another study also
found a significant interaction between alcohol and
the ADH1C genotype but only after post hoc
regrouping of alcohol [14]. In other studies, there
was no significant interaction between alcohol and
the ADH1C genotype [15–18]. For ADH1B alleles,
there is a much larger difference in alcohol degradation rate of the produced enzymes (340 mM/min
compared to 9 mM/min) than for enzymes produced
by ADH1C alleles (88 mM/min versus 35 mM/min)
[5]. We found no indication that the risk of acute
coronary syndrome was lower among men with
ADH1B slow compared with ADH1B fast alcohol
degradation, however, the statistical power to investigate these associations was limited.
In human studies, it has been difficult to demonstrate an effect of the ADH1B variations on the
alcohol degradation rate [7], which may be because
of insufficiently sensitive laboratory methods – in a
recent study where a more refined method for measuring the rate of alcohol degradation was applied, the
results showed a significant difference in degradation
rate according to the ADH1B genotype [8]. In further
support of an in vivo effect of the ADH1B genotype is
that individuals with the most active enzymes are
consistently reported to experience more unpleasant
symptoms such as flushing when drinking alcohol
compared with individuals with the less active enzymatic forms [9–12]. For ADH1C variations, similar
studies have not been performed, and because of the
more limited effect on alcohol degradation rate of
ADH1C compared to ADH1B, it is most likely more
difficult to observe in vivo effects.
In conclusion, we did not observe any associations
between ADH1B or ADH1C genotype and risk of
acute coronary syndrome and no interaction between
alcohol intake, genotypes and risk of acute coronary
syndrome. Hence, our findings do not support the
association between alcohol intake and ischaemic
heart disease being modified by genetic variation in
alcohol degrading enzymes.
Acknowledgements
We would like to thank the participants in the Diet,
Cancer and Health Study for their cooperation.
493
Also, we thank Katja Boll for preparing the data
file and Anne-Karin Jensen for excellent technical
support.
Grant/funding support: This research was supported by the Axel Muusfeldt Foundation and by the
Danish Cancer Society, grant DP00027, Svend
Andersen’s Fond and the grant IMAGE from the
Danish Ministry of Health, Research Centre for
Environmental Health’s Fund.
Financial disclosures: None declared.
References
[1] Stampfer MJ, Colditz GA, Willett WC, Speizer FE,
Hennekens CH. A prospective study of moderate alcohol
consumption and the risk of coronary disease and stroke in
women. N Engl J Med 1988;319(5):267–73.
[2] Corrao G, Rubbiati L, Bagnardi V, Zambon A,
Poikolainen K. Alcohol and coronary heart disease: a metaanalysis. Addiction 2000;95(10):1505–23.
[3] Gronbaek M, Becker U, Johansen D, Gottschau A,
Schnohr P, Hein HO, et al. Type of alcohol consumed and
mortality from all causes, coronary heart disease, and cancer.
Ann Int Med 2000;133(6):411–19.
[4] Flesch M, Rosenkranz S, Erdmann E, Bohm M. Alcohol and
the risk of myocardial infarction. Basic Res Cardiol
2001;96(2):128–35.
[5] Bosron WF, Li TK. Genetic polymorphism of human liver
alcohol and aldehyde dehydrogenases, and their relationship
to alcohol metabolism and alcoholism. Hepatology
1986;6(3):502–10.
[6] Tolstrup JS, Nordestgaard BG, Rasmussen S, TybjaergHansen A, Gronbaek M. Alcoholism and alcohol drinking
habits predicted from alcohol dehydrogenase genes.
Pharmacogenomics J 2007;8(3):220–7.
[7] Mizoi Y, Yamamoto K, Ueno Y, Fukunaga T, Harada S.
Involvement of genetic polymorphism of alcohol and aldehyde dehydrogenases in individual variation of alcohol
metabolism. Alcohol Alcohol 1994;29(6):707–10.
[8] Neumark YD, Friedlander Y, Durst R, Leitersdorf E,
Jaffe D, Ramchandani VA, et al. Alcohol dehydrogenase
polymorphisms influence alcohol-elimination rates in a
male Jewish population. Alcohol Clin Exp Res 2004;
28(1):10–14.
[9] Yokoyama T, Yokoyama A, Kato H, Tsujinaka T, Muto M,
Omori T, et al. Alcohol flushing, alcohol and aldehyde
dehydrogenase genotypes, and risk for esophageal squamous
cell carcinoma in Japanese men. Cancer Epidemiol
Biomarkers Prev 2003;12(11 pt 1):1227–33.
[10] Carr LG, Foroud T, Stewart T, Castelluccio P, Edenberg HJ,
Li TK. Influence of ADH1B polymorphism on alcohol use
and its subjective effects in a Jewish population. Am J Med
Genet 2002;112(2):138–43.
[11] Takeshita T, Yang X, Morimoto K. Association of the ADH2
genotypes with skin responses after ethanol exposure in
Japanese male university students. Alcohol Clin Exp Res
2001;25(9):1264–9.
[12] Takeshita T, Mao XQ, Morimoto K. The contribution of
polymorphism in the alcohol dehydrogenase beta subunit to
alcohol sensitivity in a Japanese population. Hum Genet
1996;97(4):409–13.
494
J. S. Tolstrup et al.
[13] Hines LM, Stampfer MJ, Ma J, Gaziano JM, Ridker PM,
Hankinson SE, et al. Genetic variation in alcohol dehydrogenase and the beneficial effect of moderate alcohol consumption on myocardial infarction. N Engl J Med 2001;
344(8):549–55.
[14] Younis J, Cooper JA, Miller GJ, Humphries SE, Talmud PJ.
Genetic variation in alcohol dehydrogenase 1C and the
beneficial effect of alcohol intake on coronary heart disease
risk in the Second Northwick Park Heart Study.
Atherosclerosis 2005;180(2):225–32.
[15] Djousse L, Levy D, Herbert AG, Wilson PW,
D’Agostino RB, Cupples LA, et al. Influence of alcohol
dehydrogenase 1C polymorphism on the alcohol-cardiovascular disease association (from the Framingham Offspring
Study). Am J Cardiol 2005;96(2):227–32.
[16] Heidrich J, Wellmann J, Doring A, Illig T, Keil U. Alcohol
consumption, alcohol dehydrogenase and risk of coronary
heart disease in the MONICA/KORA-Augsburg cohort
1994/1995–2002. Eur J Cardiovasc Prev Rehabil 2007;
14(6):769–74.
[17] Ebrahim S, Lawlor DA, Shlomo YB, Timpson N,
Harbord R, Christensen M, et al. Alcohol dehydrogenase
type 1C (ADH1C) variants, alcohol consumption traits,
HDL-cholesterol and risk of coronary heart disease in
women and men: British Women’s Heart and Health
Study
and
Caerphilly
cohorts.
Atherosclerosis
2008;196(2):871–8.
[18] Tolstrup JS, Gronbaek M, Nordestgaard BG. Alcohol intake,
myocardial infarction, biochemical risk factors, and alcohol
dehydrogenase genotypes. Circulation: Cardiovascular
Genetics 2009;2:507–14.
[19] Tjonneland A, Gronbaek M, Stripp C, Overvad K. Wine
intake and diet in a random sample of 48763 Danish men
and women. Am J Clin Nutr 1999;69(1):49–54.
[20] Overvad K, Tjonneland A, Haraldsdottir J, Ewertz M,
Jensen OM. Development of a semiquantitative food frequency questionnaire to assess food, energy and nutrient
intake in Denmark. Int J Epidemiol 1991;20(4):900–5.
[21] Tjonneland A, Overvad K, Haraldsdottir J, Bang S,
Ewertz M, Jensen OM. Validation of a semiquantitative
food frequency questionnaire developed in Denmark. Int J
Epidemiol 1991;20(4):906–12.
[22] Miller SA, Dykes DD, Polesky HF. A simple salting out
procedure for extracting DNA from human nucleated cells.
Nucleic Acids Res 1988;16(3):1215.
[23] Joensen AM, Jensen MK, Overvad K, Dethlefsen C,
Schmidt EB, Rasmussen LH, et al. Predictive values of
acute coronary syndrome discharge diagnoses differ in the
Danish National Patient Registry. J Clin Epidemiol
2009;62(2):188–94.
[24] Prentice RL. A case-cohort design for epidemiologic cohort
studies and disease prevention trials. Biometrika 1986;73:
1–12.
[25] Petersen L, Sorensen TI, Andersen PK. Comparison of casecohort estimators based on data on premature death of adult
adoptees. Stat Med 2003;22(24):3795–803.
[26] Cavalli-Sforza L, Bodmer W. Population structure. In: The
genetics of human populations. New York: Dover; 1999.
[27] Tolstrup J, Jensen MK, Tjonneland A, Overvad K,
Mukamal KJ, Gronbaek M. Prospective study of alcohol
drinking patterns and coronary heart disease in women and
men. BMJ 2006;332(7552):1244–8.
Blackwell Science, LtdOxford, UKADDAddiction0965-2140© 2004 Society for the Study of Addiction99Original ArticleJanne S. Tolstrup et al.Drinking pattern and mortality in middle-aged men and women
doi:10.1111/j.1360-0443.2004.00667.x
RESEARCH REPORT
Drinking pattern and mortality in middle-aged men and
women
Janne S. Tolstrup1, Majken K. Jensen1, Anne Tjønneland2, Kim Overvad3,4 & Morten Grønbæk1
Centre for Alcohol Research, National Institute of Public Health, Copenhagen, Denmark1, Institute of Cancer Epidemiology, the Danish Cancer Society,
Copenhagen, Denmark2 , Department of Clinical Epidemiology, Aalborg Hospital, Aalborg, Denmark3 and Department of Epidemiology and Social Medicine,
Aarhus University, Aarhus, Denmark4
Correspondence to:
Janne S. Tolstrup
Centre for Alcohol Research
National Institute of Public Health
Svanemøllevej 25
DK-2100
Copenhagen
Denmark
E-mail: [email protected]
Submitted 3 April 2003;
initial review completed 23 July 2003;
final version accepted 24 November 2003
RESEARCH REPORT
ABSTRACT
Aims To address the prospective association between alcohol drinking pattern
and all-cause mortality.
Design Population-based cohort study conducted between 1993 and 2003.
Setting Denmark.
Participants A total of 26 909 men and 29 626 women aged 55–65 years.
Measurements We obtained risk estimates for all-cause mortality for different
levels of quantity and frequency of alcohol intake adjusted for life-style factors,
including diet.
Findings During follow-up, 1528 men and 915 women died. For the same
average consumption of alcohol, a non-frequent intake implied a higher risk of
death than a frequent one.
Conclusions Drinking pattern and not just the total amount of alcohol consumed is important for the association between alcohol intake and mortality.
These results suggest that future public guidelines concerning sensible alcohol
drinking should include messages about drinking pattern together with quantity of alcohol.
KEYWORDS
mortality.
Alcohol drinking, drinking behaviour, follow-up studies,
INTRODUCTION
A large number of prospective studies have consistently
reported a J-shaped relation between an average measure
of alcohol intake and all-cause mortality [1–3]. This
characteristic form most probably reflects a beneficial
effect on the cardiovascular system of light alcohol
intake, and harmful implications, such as liver cirrhosis
and cancer, of high consumption. These associations
have been addressed mainly without taking drinking pattern into account, with the exception of some recent studies [4–10]. Although these studies differed with regard to
type and quality of measures of drinking patterns, results
implied consistently the importance of drinking pattern
in addition to the total quantity consumed. Most recently
it has been suggested that drinking frequency, and not
© 2004 Society for the Study of Addiction
the total amount of alcohol, is the primary determinant
of the inverse association between alcohol intake and coronary heart disease [8]. However, it seems unlikely that
the cardioprotective benefits would outweigh the detrimental effects of a high alcohol intake, regardless of the
drinking pattern. Hence, for public health purposes, a
more universal outcome such as mortality from allcauses is relevant, because it constitutes a scientific basis
for creating guidelines on sensible drinking.
The aim of the present study is to investigate the association between frequency of drinking episodes for a
given level of total alcohol consumption and all-cause
mortality. We use data from a large prospective cohort
study consisting of middle-aged men and women and
have the ability to adjust for related life-style factors such
as diet and physical activity.
Addiction, 99, 323–330
324
Janne S. Tolstrup et al.
METHODS
During December 1993 to May 1997, 160 725 Danish
men and women aged 50–65 years were invited by mail
to participate in the population-based study ‘Diet, Cancer
and Health’ [11]. Eligible subjects were born in Denmark
and had no previous cancers at the time of inclusion.
With the invitation, a detailed 192-item food frequency
questionnaire including questions concerning average
alcohol intake was enclosed. A first visit to the study
clinic was arranged by telephone with subjects who
agreed to participate [27 178 men and 29 875 women
(35%)]. The food frequency questionnaire was returned
during the clinic visit, where another questionnaire concerning life-style and background factors including information on frequency of alcohol intake was completed. A
description of the food frequency questionnaire has been
published previously [12]. The study was conducted in
accordance with the Helsinki Declaration II and was
approved by the Ethical Committees for the Copenhagen
and the Aarhus municipalities (KF 01–116/96).
Alcohol intake and drinking patterns
In the background questionnaire, subjects reported their
usual frequency of alcohol intake in seven possible
response categories: never drink alcohol, less than once
per month, one to three times per month, once a week,
two to four times per week, five to six times per week and
daily.
In the food frequency questionnaire, participants
were asked to state their average quantity (during the last
year) of alcohol consumption as the intake of specific
amounts of each beverage: light, normal and strong beer
(in number of bottles); red, white and fortified wine (in
number of glasses); and spirits (in number of drinks). The
possible response categories were no alcohol intake, less
than one per month, one per month, two to three per
month, one per week, two to four per week, five to six per
week, one per day, two to three per day, four to five per day,
six to seven per day and eight or more per day. Based on
ethanol content in the different beverage types, these categories were converted into number of standard drinks
(12 g alcohol) per week and added to yield an average
measure of total alcohol intake.
For a given quantity of total alcohol intake, two
groups of drinkers were formed to differentiate between
individuals drinking little alcohol frequently and individuals consuming a larger quantity of alcohol more rarely.
Frequent drinkers were defined as individuals who consumed alcohol at least 2 days per week and non-frequent
drinkers were defined as subjects who used to drink alcohol less often. Abstainers were defined as subjects who, in
both questionnaires, reported never to drink.
© 2004 Society for the Study of Addiction
For women, total alcohol intake was categorized into
five levels (none, less than one, one to six, seven to 13 and
more than 13 drinks per week) and for men, total alcohol
intake was categorized into six levels (none, less than one,
one to six, seven to 13, 14–20 and more than 20 drinks
per week).
Education
In the life-style questionnaire, education was estimated
from length of basic schooling as 7 years or less, 8–10
years or 11 years and longer.
Smoking habits
Subjects reported if they were never-smokers, ex-smokers
or current-smokers. Current smokers reported number of
daily cigarettes, cheroots, cigars and pipes. Assuming one
cigarette to be equivalent to 1 g, one cheroot or one pipe
to 3 g and one cigar to 5 g tobacco, total amount of smoking was calculated. Two variables were constructed, one
indicating smoking status (never, ex, current) and one
indicating amount of smoking (0 for never and ex-smokers, and 1–14 g per day, 15–24 g per day or more than 24
g per day for current smokers).
Body mass index
The participants’ height and weight were measured in
light clothes and without shoes. Body mass index (BMI)
was calculated as weight (kg) divided by squared height
(m) and modelled as linear splines after log-transformation with knots at 18.5, 25, and 30 kg/m2. These limits
were set in accordance with guidelines from the World
Health Organization [13].
Physical activity
Subjects reported if they were physically active during leisure time, including undertaking sports, housework, gardening, taking walks and bicycling. For each activity, a
dichotomized variable was computed with the cut-point
defined as performing or not performing the activity in
question.
Diet
Indicators of a healthy diet among the participants were
chosen from the food frequency questionnaire. For intake
of fish, cooked vegetables, salad and fruit, respectively, the
intake was dichotomized as high or low. The cut-points
were defined as close as possible to the 10th percentile of
the sex-specific distribution (fish, once a month or less;
vegetables, twice per month or less; salad, once a month
or less; and fruit, once a week or less). The participants
also indicated which type of fat used mainly for cooking
Addiction, 99, 323–330
Drinking pattern and mortality in middle-aged men and women
and two groups were formed: the participants in one
group who used mainly olive oil and those in another
group who used mainly other types of fat for cooking. Use
of fat spread on bread was used as a measure of saturated
fat intake because one-third of saturated fat intake in
Denmark is consumed as spread on bread. Two groups
were formed, users and non-users of fat spread on bread.
Diseases before baseline
Information on the participants’ health status when
entering the study cohort was obtained from the population-based Danish Patient Register, which keeps records
of all somatic hospitalizations in Denmark since 1977.
The diagnoses are classified according to the World
Health Organization’s International Classification of Diseases, 8th revision (ICD-8). By linking the study cohort to
this register, information on the participants’ health status from 1977 to baseline (1993–97) was obtained.
Dichotomized variables were constructed for stroke (ICD8 codes: 430–438), acute myocardial infarction (ICD-8
code: 410), angina pectoris (ICD-8 codes: 411 and 413),
other cardiovascular diseases (ICD-8 codes: 390–409,
412, 414–429 and 439–458) and other diagnoses
implying diseases with a chronic character. The latter
includes infectious, endocrinological, nervous system,
chronic lung, gastrointestinal, alcohol-related, urological
and muscular diseases (ICD-8 codes: 40–46, 79–83, 93–
95, 240–289, 340–358, 490–493, 530–537, 560–573,
577, 580–584 and 710–738).
Follow-up
Vital status of the study population sample was followed
until 20 February 2003 by using the unique person
identification number in the Civil Registration System.
The observation time for each participant was the period
from enrolment into the study (December 1993 to May
1997) until 20 February 2003, death (n = 2443), emigration (n = 255) or disappearance (n = 4), whichever
came first.
Statistical analysis
Subjects with incomplete information on alcohol intake
(n = 104) or on any of the potential confounders
(n = 240) were excluded from the analyses. A few subjects had reported conflicting answers between their
average total alcohol intake and the frequency of alcohol
intake, and as it was difficult to categorize such subjects
they were excluded from the analyses (n = 174). A total of
56 535 subjects were eligible for this study.
Pearson’s correlation coefficient was calculated to
examine the magnitude of correlation between drinking
frequency and amount of drinking.
© 2004 Society for the Study of Addiction
325
Risk estimates were computed by means of Cox proportional hazard regression models [14] (SAS/STAT program software). Age was used as the time axis to ensure
that the estimation procedure was based on comparisons
of individuals at the same age. The analyses were corrected for delayed entry, such that individuals were considered at risk only from the age at entry into the study
cohort. In one model (Fig. 1), the frequency of drinking
was categorized into two levels (subjects consuming alcohol at least 2 days per week and subjects consuming
alcohol less often) for each level of total alcohol intake. In
another model (Table 1a,b), the frequency of drinking
was categorized into four levels (once per week or less,
two to four times per week, five to six times per week and
daily drinking) for every level of total alcohol intake. For
each model, all combinations of frequency and level of
total alcohol intake was entered simultaneously. Having
had a diagnosis of a disease before baseline, school education, smoking, BMI, intake of fish, fruit, salad and vegetables, use of olive oil in cooking and of fat on bread were
included as covariates in the adjusted model. All analyses
were performed for each sex separately. The assumption
of proportional hazards in the Cox model was tested for
each covariate by evaluating the parallelism of the stratified survival curves graphically and by constructing
time-dependent variables for the covariates in question
and testing these for statistical significance. No violations
were detected. Analyses were repeated after exclusion of
subjects with a disease before baseline.
We used the Wald test to examine the joint hypothesis
of differences in the hazard ratio for mortality between
non-frequent and frequent drinkers for a weekly alcohol
intake of more than one drink per week.
RESULTS
Among men who reported to consume any alcohol,
21 083 did so at least twice per week while 4450 drank
alcohol less frequently (Table 2a). Among alcohol-consuming women, 16 659 were frequent drinkers and
8103 were non-frequent drinkers (Table 2b). Among
both men and women, the median alcohol consumption
was higher among frequent drinkers for each category of
total alcohol intake than among the corresponding nonfrequent drinkers. Overall, drinking frequency was correlated moderately to amount of drinking [Pearson’s correlation coefficient = 0.70 (women) and 0.63 (men)].
Among both men and women, non-frequent drinkers
generally had a lower educational level, were more often
smokers, more often obese and eating fewer vegetables
and fruit than frequent drinkers.
During a mean follow-up of 6.8 years, 1528 men and
915 women died. The adjusted hazard ratios for non-frequent and frequent drinkers according to total alcohol
Addiction, 99, 323–330
326
Janne S. Tolstrup et al.
3
Women
Hazard ratio of mortality
Hazard ratio of mortality
Men
Non-frequent‡
Frequent
2
1
3
Non-frequent‡
Frequent
*
2
*
1
*
*
**
*
0
0
0
<1
1-6
7-13
14-20
21+
0
<1
1-6
Alcohol intake (drinks per week)
7-13
14+
Alcohol intake (drinks per week)
Figure 1 Hazard ratios (age-adjusted estimates are represented by broken lines and fully adjusted† estimates by full lines) for all-cause mortality according to quantity and frequency of alcohol intake in men and women‡ (frequent = at least 2 drinking days per week, nonfrequent = less than 2 drinking days per week).
†Adjusted for education, smoking, BMI, physical activity, diet and diseases before baseline;
‡reference category is drinkers of less than one but more than zero drinks per week.
*p = value less than 0.05 compared to reference category
Table 1 Adjusted hazard ratios* of all-cause mortality (95% confidence limits) according to quantity and frequency of alcohol intake.
Frequency of alcohol intake
Alcohol intake,
drinks per week
(a) Men
0
Abstainers
1.31
(0.96–1.78)
Less than one
1–6
–
–
7–13
–
14–20
–
21 +
–
(b) Women
0
1.62
(1.19–2.19)
Less than one
–
1–6
7–13
14 +
*
Once per
week or less
2–4 times
per week
5–6 times
per week
Daily
–
1.00
(reference)
0.61
(0.47–0.79)
0.61
(0.42–0.90)
1.11
(0.62–2.00)
1.25
(0.70–2.24)
–
–
–
–
0.74
(0.56–0.97)
0.56
(0.43–0.73)
0.61
(0.43–0.87)
1.03
(0.76–1.41)
–
0.91
(0.50–1.66)
0.51
(0.36–0.73)
0.52
(0.35–0.76)
0.68
(0.51–0.92)
–
0.84
(0.39–1.82)
0.63
(0.44–0.89)
0.67
(0.49–0.93)
0.84
(0.66–1.06)
–
–
–
–
0.90
(0.70–1.17)
0.94
(0.72–1.24)
1.06
(0.69–1.62)
–
0.72
(0.32–1.64)
0.84
(0.56–1.27)
1.03
(0.72–1.46)
–
0.83
(0.31–2.24)
0.90
(0.61–1.33)
1.31
(1.02–1.68)
–
1.00
(reference)
1.05
(0.85–1.30)
1.30
(0.85–2.01)
2.19
(1.15–4.17)
Adjusted for education, smoking, BMI, physical activity, diet and diseases before baseline.
© 2004 Society for the Study of Addiction
Addiction, 99, 323–330
Drinking pattern and mortality in middle-aged men and women
327
Table 2 Baseline characteristics of frequent and non-frequent drinkers according to categories of total alcohol intake.
Alcohol intake, drinks per week
1–6
Nonfrequent
(a) Men, characteristics*
Subjects (n)
3508
Alcohol intake (median drinks per week)
3.1
Age (mean years)
56
Education (% lowest level)
43
BMI (% more than 30 kg/m 2)
21
Smoking (% current)
39
Vegetable intake (% in lowest consumption group) 11
Fruit intake (% in lowest consumption group)
6
7–13
(b) Women, characteristics†
Subjects (number)
7417
Alcohol intake (median drinks per week)
2.5
Age (mean years)
56
Education (% lowest level)
38
Body mass index (% more than 30 kg/m 2)
19
Smoking (% current)
31
Vegetable intake (% in lowest consumption group) 19
Fruit intake (% in lowest consumption group)
11
*
More than 20
Frequent
Nonfrequent
Frequent
Nonfrequent
Frequent
2669
4.7
56
37
17
33
10
6
741
8
56
45
22
42
10
7
6981
9.4
56
31
16
33
7
5
117
17.6
56
51
32
44
13
9
3464
17.9
56
31
15
37
9
8
1–6
Nonfrequent
14–20
7–13
Nonfrequent
84
27
56
50
25
60
23
19
Frequent
7969
30.5
56
32
20
47
10
12
More than 13
Frequent
Nonfrequent
Frequent
Nonfrequent
Frequent
5005
4.5
56
27
14
26
16
9
580
7.7
57
40
17
43
22
12
6326
8.7
56
25
11
29
14
10
106
18.6
57
42
18
55
23
25
5328
20.6
56
21
11
41
15
18
In addition, 1376 men drank less than one drink per week; †4864 women drank less than one drink per week.
intake compared with non-drinkers (drinkers of more
than zero but less than one drink per week) were estimated (Fig. 1). The hazard ratios of mortality were higher
among non-frequent drinkers than among frequent
drinkers for a weekly alcohol intake of more than one
drink per week [P = 0.03 (men) and P = 0.05 (women)
using the Wald test]. Among non-frequent drinking men,
the hazard ratios were 0.61 (95% CI, 0.47–0.79), 0.61
(95% CI, 0.42–0.90), 1.11 (95% CI, 0.62–2.00) and
1.25 (95% CI, 0.70–2.24) for subjects drinking one to
six, seven to 13, 14–20 and more than 20 weekly drinks,
respectively. Correspondingly, the hazard ratios among
frequent drinking men were 0.76 (95% CI, 0.58–0.99),
0.57 (95% CI, 0.44–0.72), 0.61 (95% CI, 0.47–0.80)
and 0.83 (95% CI, 0.66–1.04). Among non-frequent
drinking women, the hazard ratios were 1.05 (95% CI,
0.85–1.30), 1.31 (95% CI, 0.85–2.01) and 2.19 (95%
CI, 1.15–4.17) for subjects drinking one to six, seven to
13 and more than 13 weekly drinks, respectively. Correspondingly, the hazard ratios among frequent drinking
women were 0.89 (95% CI, 0.69–1.14), 0.91 (95% CI,
0.72–1.16) and 1.20 (95% CI, 0.96–1.51).
© 2004 Society for the Study of Addiction
The mortality rate ratios for different combinations of
quantity and frequency of alcohol intake were also estimated (Table 1a,b). For men, the lowest risk estimates
was for drinking seven to 13 drinks per week distributed
5–6 days per week (0.51 95% CI: 0.36–0.73) and for
drinking 14–21 drinks per week distributed on 5–6 days
per week (0.52 95% CI 0.35–0.76) (Table 2a). The highest hazard ratios were obtained among men drinking on
1 day per week or more rarely; for this category the hazard ratio was 1.11 (95% CI: 0.62–2.00) for drinking 14–
20 drinks per week and 1.25 (95% CI: 0.70–2.24) for
drinking more than 20 drinks per week. The hazard ratio
for drinking totally 21 or more drinks per week distributed on 7 days per week was 0.84 (95% CI: 0.66–1.06).
For women, the lowest risk estimate was for drinking one
to six drinks per week distributed on 5–6 days per week
(0.72 95% CI: 0.32–1.64) and for drinking seven to 13
drinks per week distributed on 5–6 days per week (0.84
95% CI: 0.56–1.27) (Table 2a). The highest hazard ratios
were obtained among women drinking on 1 day per week
or more rarely; for this category the hazard ratio was
1.30 (95% CI: 0.85–2.01) for drinking seven to 13 drinks
Addiction, 99, 323–330
328
Janne S. Tolstrup et al.
per week and 2.19 (95% CI: 1.15–4.17) for drinking
more than 13 drinks per week. The hazard ratio for drinking 14 or more drinks per week distributed on 7 days per
week was 1.31 (95% CI: 1.02–1.68). Risk estimates of
subjects without diseases before baseline did not differ
from the estimates for all subjects (data not shown).
The following covariates were associated independently and positively with mortality for men: having a
diagnose of acute myocardial infarction, angina, other
cardiovascular diseases or any chronic disease before
baseline, not performing any physical activity, smoking,
not eating fruit and salad, not using olive oil for cooking,
having a BMI < 18 kg/m2 and having school education
for less than 11 years. For women, having had a diagnosis of acute myocardial infarction or any chronic disease
before baseline, not performing any physical activity,
smoking, not eating salad and having a BMI < 18 kg/m2
were independently and positively associated with
mortality.
DISCUSSION
We found that drinking pattern influenced the relation
between alcohol intake and all-cause mortality. For the
same average consumption of alcohol, a non-frequent
intake implied a higher risk of death than a frequent one.
However, frequent heavy drinking (>20 drinks per week
for men and >13 drinks per week for women) also implied
an increased risk of death compared to light drinking.
Our study population consisted of middle-aged men and
women. This age group constitutes a high-risk population for heart diseases and it is therefore qualified for
investigating how the deleterious effects of alcohol are
balanced against the protecting effect, according to
drinking frequency and amount of intake.
The follow-up period in this study was 6.8 years,
which is a shorter period than that seen in most other epidemiological studies. This means that the information on
alcohol intake given by the participants at baseline probably describes more accurately the factual behaviour of
the subjects at follow-up in the present study. The combination of this relatively short follow-up period, the large
number of participants and a large variation in frequency
and amount of alcohol intake allows us to estimate the
hazard ratios for non-frequent and frequent drinkers
separately.
The finding that the association between alcohol
intake and mortality depends upon drinking pattern has
been suggested previously [4–7]. Although other measures of drinking patterns were used, results imply consistently a hazardous effect of drinking alcohol in large
amounts per occasion. Most of these studies did not assess
drinking pattern over the whole spectrum of total alcohol
intake and it was difficult to differentiate between the
© 2004 Society for the Study of Addiction
influence from total alcohol intake and drinking pattern.
We avoided the term ‘binge drinking’, which in most
studies is defined as drinking a minimum number of
drinks per occasion, such as six or 13 [6,7], because the
participants were not asked directly about occasional
heavy drinking and we can therefore not comment on
this with the present data. The drinking pattern in the
present study was constructed by combining information
on average quantity with usual drinking frequency, as
has been performed in some other studies [8,10].
In the present study, covariates were distributed
unequally in the two groups of drinkers for most factors
and the more ‘unhealthy’ pattern was observed consistently among the non-frequent drinkers (Table 1). Also,
Kesse et al. showed that dietary habits are unequally distributed on different categories of alcohol intake [15].
This underlines the importance of a thorough confounder control when addressing alcohol intake and
drinking pattern as independent variables. We held information on smoking habits, physical activity, BMI, diet
and school education, which provided the possibility to
adjust for these potential confounders. To adjust for diet,
five presumed indicators of a healthy diet were chosen:
intake of fruit, vegetables, saturated fat, plant oil and fish.
Adjusting for diet and physical activity reduced the difference in hazard ratios between frequent and non-frequent
drinkers and hence the importance of drinking pattern.
Possible confounders of our results are social factors, as
high volumes of alcohol per occasion have been shown to
be associated with negative social circumstances [16]. In
the present study, adjustment was made for education,
which is expected to correlate strongly with social status.
However, more detailed information on other social factors was not accounted for.
Among light drinkers there was little difference in hazard ratios between non-frequent and frequent drinkers.
Consequently, the reduced risk of death in light drinkers
compared with abstainers seems to depend less on drinking pattern than suggested previously [9]. The beneficial
effect of a light alcohol intake on cardiovascular disease
has several plausible biological mechanisms, including an
increase of serum high-density lipoprotein (HDL) [17],
inhibition of platelet production, activation and aggregation [18,19] and increased fibrinolysis [20,21]. The influence of drinking pattern on these mediators has been
studied in interventions with moderate to heavy drinkers,
where there were no differences in lipid profile or fibrinolysis between weekend and daily drinkers [22,23]. In
contrast, non-frequent drinkers had a higher degree of
coronary occlusion and a decreased HDL to low-density
lipoprotein ratio compared with drinkers with a more regular drinking pattern [24]. The question is whether the
latter finding applies to individuals with a light to moderate alcohol intake, especially as another study has shown
Addiction, 99, 323–330
Drinking pattern and mortality in middle-aged men and women
that most light drinkers rarely drink daily and that most
daily drinkers are not light drinkers [25].
We used information on mortality from the Civil Registration System, which is updated to 2003. In the future,
it will be interesting to include information on cause-specific deaths, but because the follow-up time in the registers containing this information is much shorter than for
the Civil Registration System, it is not yet possible.
The non-frequent drinking pattern compared to frequent drinking involves higher alcohol concentrations in
the gastrointestinal tract and in the blood as the non-frequent drinkers consume more alcohol per drinking occasion than do frequent drinkers. This could lead to an
enhancement of the harmful effects of alcohol, including
alcoholic liver disease and upper gastrointestinal cancers.
Wetterling et al. have investigated drinking patterns
among alcoholics and found that the occurrence of alcohol-related disorders were more common among subjects
with frequent inebriation compared with more continuous drinkers with similar life-time alcohol intake [26]. To
our knowledge, the association between drinking pattern
and neoplasms in the gastrointestinal tract has not been
investigated. Occasional drinking of high consumptions
of alcohol is probably also stronger when associated with
accidents and suicide, due to increased risk-taking
behaviours.
In conclusion, we found that frequency of drinking for
moderate and high consumption of alcohol influenced
the association between alcohol intake and mortality. At
these levels, mortality was higher among non-frequent
drinkers compared with frequent drinkers.
5.
6.
7.
8.
9.
10.
11.
12.
13.
ACKNOWLEDGEMENTS
This study was supported by grants from the Danish Cancer Society and the Danish National Board of Health.
These sponsors had no role in the design or conduct of the
study, in the collection, analysis or interpretation of the
data, nor in the preparation, review or approval of
the manuscript. The authors thank Ms Katja Boll for preparing the data file for statistical analysis.
14.
15.
16.
REFERENCES
17.
1. Poikolainen, K. (1995) Alcohol and mortality. Journal of
Clinical Epidemiology, 48, 455–465.
2. White, I. R. (1999) The level of alcohol consumption at
which all-cause mortality is least. Journal of Clinical Epidemiology, 52, 967–975.
3. Shaper, A. G. (1990) Alcohol and mortality: a review of
prospective studies. British Journal of Addiction, 85, 837–
847.
4. Rehm, J., Greenfield, T. K. & Rogers, J. D. (2001) Average
volume of alcohol consumption, patterns of drinking, and
© 2004 Society for the Study of Addiction
18.
19.
329
all-cause mortality: results from the US National Alcohol
Survey. American Journal of Epidemiology, 153, 64–71.
Trevisan, M., Schisterman, E., Mennotti, A., Farchi, G. &
Conti, S. (2001) Drinking pattern and mortality: the Italian
Risk Factor and Life Expectancy pooling project. Annals of
Epidemiology, 11, 312–319.
Malyutina, S., Bobak, M., Kurilovitch, S., Gafarov, V., Simonova, G., Nikitin, Y. & Marmot, M. (2002) Relation between
heavy and binge drinking and all-cause and cardiovascular
mortality in Novosibirsk, Russia: a prospective cohort
study. Lancet, 360, 1448–1454.
Laatikainen, T., Manninen, L., Poikolainen, K. &
Vartiainen, E. (2003) Increased mortality related to heavy
alcohol intake pattern. Journal of Epidemiology and Community Health, 57, 379–384.
Mukamal, K. J., Conigrave, K. M., Mittleman, M. A., Camargo, C. A. Jr, Stampfer, M. J., Willett, W. C. & Rimm, E. B.
(2003) Roles of drinking pattern and type of alcohol consumed in coronary heart disease in men. New England Journal of Medicine, 348, 109–118.
Murray, R. P., Connett, J. E., Tyas, S. L., Bond, R., Ekuma,
O., Silversides, C. K. & Barnes, G. E. (2002) Alcohol volume,
drinking pattern, and cardiovascular disease morbidity and
mortality: is there a U-shaped function? American Journal of
Epidemiology, 155, 242–248.
McElduff, P. & Dobson, A. J. (1997) How much alcohol and
how often? Population based case–control study of alcohol
consumption and risk of a major coronary event. British
Medical Journal, 314, 1159–1164.
Tjonneland, A., Gronbaek, M., Stripp, C. & Overvad, K.
(1999) Wine intake and diet in a random sample of 48 763
Danish men and women. American Journal of Clinical Nutrition, 69, 49–54.
Overvad, K., Tjonneland, A., Haraldsdottir, J., Ewertz, M. &
Jensen, O. M. (1991) Development of a semiquantitative
food frequency questionnaire to assess food, energy and
nutrient intake in Denmark. International Journal of Epidemiology, 20, 900–905.
World Health Organization (WHO) (2000) Obesity: Preventing and Managing the Global Epidemic. Report of a WHO consultation, no. 894. Geneva: WHO.
Cox, D. R. (1972) Regression models and time tables. Journal of the Royal Statistical Society, 34, 187–202.
Kesse, E., Clavel-Chapelon, F., Slimani, N. & van Liere, M.
(2001) Do eating habits differ according to alcohol consumption? Results of a study of the French cohort of the
European Prospective Investigation into Cancer and Nutrition (E3N-EPIC). American Journal of Clinical Nutrition, 74,
322–327.
Rehm, J. & Gmel, G. (1999) Patterns of alcohol
consumption and social consequences. Results from an
8-year follow-up study in Switzerland. Addiction, 94,
899–912.
Langer, R. D., Criqui, M. H. & Reed, D. M. (1992) Lipoproteins and blood pressure as biological pathways for effect of
moderate alcohol consumption on coronary heart disease.
Circulation, 85, 910–915.
Mikhailidis, D. P., Barradas, M. A. & Jeremy, J. Y. (1990)
The effect of ethanol on platelet function and vascular prostanoids. Alcohol, 7, 171–180.
Mikhailidis, D. P., Jenkins, W. J., Barradas, M. A., Jeremy, J.
Y. & Dandona, P. (1986) Platelet function defects in
chronic alcoholism. British Medical Journal (Clinical Research
Edition), 293, 715–718.
Addiction, 99, 323–330
330
Janne S. Tolstrup et al.
20. Krobot, K., Hense, H. W., Cremer, P., Eberle, E. & Keil, U.
(1992) Determinants of plasma fibrinogen: relation to body
weight, waist-to-hip ratio, smoking, alcohol, age, and sex.
Results from the second MONICA Augsburg survey 1989–
90. Arteriosclerosis and Thrombosis, 12, 780–788.
21. Brunner, E. J., Marmot, M. G., White, I. R., O’Brien, J. R.,
Etherington, M. D., Slavin, B. M., Kearney, E. M. & Smith,
G. D. (1993) Gender and employment grade differences in
blood cholesterol, apolipoproteins and haemostatic factors in the Whitehall II study. Atherosclerosis, 102, 195–
207.
22. Rakic, V., Puddey, I. B., Dimmitt, S. B., Burke, V. & Beilin, L.
J. (1998) A controlled trial of the effects of pattern of alcohol
intake on serum lipid levels in regular drinkers. Atherosclerosis, 137, 243–252.
© 2004 Society for the Study of Addiction
23. Dimmitt, S. B., Rakic, V., Puddey, I. B., Baker, R., Oostryck,
R., Adams, M. J., Chesterman, C. N., Burke, V. & Beilin, L. J.
(1998) The effects of alcohol on coagulation and fibrinolytic factors: a controlled trial. Blood Coagulation and Fibrinolysis, 9, 39–45.
24. Gruchow, H. W., Hoffmann, R. G., Anderson, A. J. & Barboriak, J. J. (1982) Effects of drinking patterns on the relationship between alcohol an coronary occlusion.
Atherosclerosis, 43, 393–404.
25. Knupfer, G. (1987) Drinking for health: the daily light
drinker fiction. British Journal of Addiction, 82, 547–
555.
26. Wetterling, T., Veltrup, C., Driessen, M. & John, U. (1999)
Drinking pattern and alcohol-related medical disorders.
Alcohol and Alcoholism, 34, 330–336.
Addiction, 99, 323–330
Alcohol Consumption and Risk of Atrial Fibrillation in Men and Women: The
Copenhagen City Heart Study
Kenneth J. Mukamal, Janne S. Tolstrup, Jens Friberg, Gorm Jensen and Morten
Grønbæk
Circulation 2005;112;1736-1742; originally published online Sep 12, 2005;
DOI: 10.1161/CIRCULATIONAHA.105.547844
Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX
72514
Copyright © 2005 American Heart Association. All rights reserved. Print ISSN: 0009-7322. Online
ISSN: 1524-4539
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org/cgi/content/full/112/12/1736
Subscriptions: Information about subscribing to Circulation is online at
http://circ.ahajournals.org/subscriptions/
Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters
Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax:
410-528-8550. E-mail:
[email protected]
Reprints: Information about reprints can be found online at
http://www.lww.com/reprints
Downloaded from circ.ahajournals.org by on December 17, 2007
Alcohol Consumption and Risk of Atrial Fibrillation in Men
and Women
The Copenhagen City Heart Study
Kenneth J. Mukamal, MD, MPH, MA; Janne S. Tolstrup, MS; Jens Friberg, MD;
Gorm Jensen, MD, DMSc; Morten Grønbæk, MD, PhD, DMSc
Background—The relationship of the full range of alcohol consumption with risk of incident atrial fibrillation has been
inconsistent in previous, mainly case-control studies.
Methods and Results—In a prospective cohort study, we studied the association between self-reported alcohol use and
incident atrial fibrillation among 16 415 women and men enrolled in the Copenhagen City Heart Study. We ascertained
use of beer, wine, and spirits individually at up to 3 study visits with a structured questionnaire. We identified cases of
atrial fibrillation by routine study ECGs and a validated nationwide registry of all hospitalizations. A total of 1071 cases
occurred during follow-up. Among both women and men, alcohol consumption throughout the moderate range was not
associated with risk of atrial fibrillation. However, consumption of 35 or more drinks per week among men was
associated with a hazard ratio of 1.45 (95% CI 1.02 to 2.04); few women consumed this amount of alcohol.
Approximately 5% of cases of atrial fibrillation among men were attributable to heavy alcohol use. Further adjustment
for blood pressure and incident coronary heart disease and congestive heart failure did not attenuate the association
(hazard ratio 1.63; 95% CI 1.15 to 2.31).
Conclusions—Heavy alcohol consumption is associated with a higher risk of atrial fibrillation, at least among men. This
relationship does not appear to be related to the adverse effects of heavy drinking on coronary heart disease or blood
pressure. (Circulation. 2005;112:1736-1742.)
Key Words: alcohol 䡲 fibrillation 䡲 arrhythmia 䡲 epidemiology
S
ubstantial evidence indicates that moderate alcohol consumption is associated with a lower risk of cardiovascular
disease than abstention or heavy drinking.1,2 Cohort studies
also suggest that moderate alcohol use is inversely associated
with risk of congestive heart failure.3,4 However, although
anecdotal evidence implicates episodic heavy drinking as a
trigger of atrial fibrillation (AF),5 the relationship of the full
range of alcohol use with risk of incident AF is less certain.
At least 4 case-control studies have found relatively similar
odds of AF among abstainers and moderate drinkers but
significantly higher odds of AF among heavier drinkers.6 –9 A
recent cohort study found a higher risk of AF even among
men who consumed ⬇2 drinks per day.10 However, 2
prospective cohort studies have not confirmed this finding,11,12 and Psaty and colleagues12 found that alcohol use
was inversely associated with risk of AF in the Cardiovascular Health Study. Moreover, experimental studies suggest that
alcohol administration decreases susceptibility to and duration of AF in canine models.13,14 Thus, substantial contro-
versy remains about the relation of alcohol use and risk of
AF.
To address the prospective association of alcohol use and
risk of AF more fully, we studied more than 16 000 participants of the Copenhagen City Heart Study (CCHS), a
population-based cohort study of residents of Copenhagen,
Denmark. As part of the study, participants have received
routine resting ECGs at 3 examinations, and a national
register records diagnoses from all hospitalizations, which
provides reliable and valid assessments of incident AF.
Methods
Study Population
The CCHS began in 1976 with an original population sample of
19 698 persons randomly drawn from the Copenhagen Population
Register, of whom 14 223 attended an initial examination. After the
first examination in 1976 to 1978, this sample (including individuals
who did not attend the first examination) was reinvited to participate
in 2 more examinations in 1981 to 1983 and in 1991 to 1994. At the
second examination in 1981 to 1983, an additional sample of 500
Received March 7, 2005; revision received July 5, 2005; accepted July 7, 2005.
From the Division of General Medicine and Primary Care (K.J.M.), Beth Israel Deaconess Medical Center, Boston, Mass; Centre for Alcohol Research
(J.S.T., M.G.), National Institute of Public Health, Copenhagen, Denmark; The Copenhagen City Heart Study (J.F.), Bispebjerg University Hospital,
Copenhagen, Denmark; and the Department of Cardiovascular Medicine (G.J.), Hvidovre University Hospital, Copenhagen, Denmark.
Correspondence to Kenneth J. Mukamal, MD, MPH, MA, Division of General Medicine and Primary Care, Beth Israel Deaconess Medical Center, 330
Brookline Ave, RO-114, Boston, MA 02215. E-mail [email protected]
© 2005 American Heart Association, Inc.
Circulation is available at http://www.circulationaha.org
DOI: 10.1161/CIRCULATIONAHA.105.547844
1736 by on December 17, 2007
Downloaded from circ.ahajournals.org
Mukamal et al
participants (aged 20 to 24 years) was invited, and at the third
examination in 1991 to 1994, a final 3000 subjects (aged 20 to 49
years) were invited. The number of new participants was 14 223 at
the first visit, 1563 at the second, and 2360 at the third, for a total of
18 146 individuals. At each of the 3 examinations, participants
completed a questionnaire concerning their medical history, socioeconomic status, exercise, smoking, and drinking habits. We excluded participants with prevalent AF on their baseline ECG
(n⫽118); those with self-reported coronary heart disease, stroke, use
of cardiac medication, or use of antihypertensives at baseline
(n⫽1431); and those with missing information on alcohol use
(n⫽180), which left us with 16 415 participants eligible for analysis.
The CCHS was approved by the Ethics Committee of Copenhagen
and Frederiksberg Municipality, Denmark. New and returning participants gave informed consent verbally at the first 2 examinations
(ie, up to 1983) and in writing at each subsequent examination.
Assessment of Alcohol Consumption
Participants reported their alcohol use in standardized interviews.
They individually reported their intake of beer (in bottles), wine (in
glasses), and spirits (in units), with response categories of “never/
hardly ever,” “monthly,” “weekly,” or “daily” and number of drinks
per day among daily drinkers. As previously described,15,16 lessthan-daily intake was estimated from these categories by regression
and added to daily intake (of those beverages consumed daily) as
needed. In a validation analysis, the age-adjusted correlation coefficients between self-reported alcohol intake and measured levels of
HDL cholesterol were 0.20 among men and 0.22 among women
(P⬍0.001 for both), similar to findings from other representative
cohorts.17,18
We classified participants according to their usual weekly intake
of alcohol as in previous studies19: ⬍1 serving, 1 to 6 servings, 7 to
13 servings, 14 to 20 servings, 21 to 27 servings, 28 to 34 servings,
and ⱖ35 servings. Because of the more limited range of intake
among women, their highest category of alcohol intake was ⱖ21
servings per week.
Because intake was queried separately for each alcoholic beverage, we could not determine overall drinking frequency in the entire
cohort with certainty. Using the response categories noted, we
estimated drinking frequency among the subset of participants who
predominately consumed a single beverage type (⬎90% of their
overall alcohol use).
Determination of AF
Participants underwent a resting 12-lead ECG at each study examination. Each ECG was coded according to the Minnesota coding
system by 2 independent coders; in cases of dispute, a third coder
settled the disagreement. A physician then reviewed and confirmed
all ECGs classified as AF. In a randomly chosen sample representing
10% (817/8170) of patients aged ⱖ65 years with ECGs initially
coded without AF, only 2 ECGs (0.2%) were recoded as AF on
review.20
In addition, the Danish National Hospital Discharge Register
records discharge diagnoses from all Danish hospital admissions. We
identified first hospitalizations with a diagnosis of AF using International Classification of Diseases, 8th Revision (ICD-8) codes
427.93 and 427.94 through 1993 and International Classification of
Diseases, 10th Revision (ICD-10) code I48.9 from 1994 forward.
In a previous validation of this registry, review of a subset of 116
medical records by a cardiologist confirmed the diagnosis in 112
cases.10 We further assessed the validity of our determinations of AF
by review of a random sample of 25 medical records coded with AF
and 25 medical records coded with coronary heart disease (but not
AF) by a single cardiologist (JF). Three records coded with AF had
no evidence of AF during hospitalization (specificity 88%); 3 records
coded without AF had ECG evidence of AF (sensitivity 88%).
We also examined the validity of AF diagnoses in 2 additional
ways. First, we found an incidence rate of AF after the third
examination for adults 65 years and older of 11.6 cases per 1000
person-years; the comparable incidence rate in an older population in
the United States during a similar time period was 19.2/1000.12
Alcohol and Atrial Fibrillation
1737
Second, we found that participants with AF diagnosed in this manner
had a multivariable-adjusted hazard ratio of stroke (using published
criteria21) after the diagnosis of AF of 2.62 (95% CI 2.14 to 3.21)
among men and 4.85 (95% CI 4.02 to 5.87) among women.
Other Covariates
Arterial blood pressure was measured in the left arm with the patient
in the sitting position after 5 minutes’ rest. Study staff measured
height and weight with subjects dressed in loose-fitting clothing
without shoes. We classified participants as having diabetes if they
reported a physician-made diagnosis or if they had a nonfasting
glucose of 11.1 mmol/L (200 mg/dL) at any examination.
Statistical Analysis
We describe baseline characteristics of CCHS participants with
median values for continuous variables and proportions for categorical variables. We tested for homogeneity in baseline characteristics
with Kruskal-Wallis tests for continuous variables and ␹2 tests for
categorical variables.
Participants accrued person-time from study entry until their first
hospitalization with AF, the first examination at which AF was
noted, death, or January 1, 2001. Because we did not have data on
hospitalizations before the first study examination (with which to
exclude participants in sinus rhythm with a history of AF), we
excluded the 2 participants in the original cohort with a diagnosis of
AF within 1 year of study entry. In a sensitivity analysis, we set the
censoring date for participants with AF diagnosed at each study
examination to the median censoring date of participants diagnosed
by hospitalization in the preceding time period; this did not change
our results.
In initial multivariable Cox proportional hazards regression analyses, we simultaneously adjusted for age, height, body mass index,
smoking, physical activity, education, income, family history of
cardiovascular disease, cohabitation, FEV1 (forced expiratory volume in 1 second), and history of diabetes. We assigned indicator
variables to participants with missing information on FEV1 (n⫽828),
income (n⫽493), and body mass index (n⫽451); analyses restricted
to participants without missing information for covariates yielded
very similar results that are not shown here. Age was used as the time
axis in Cox models to ensure comparison of individuals at the same
age and hence maximize adjustment for age.
To evaluate possible mediators, we performed analyses that
additionally adjusted for systolic blood pressure and treated hypertension, incident congestive heart failure, and incident coronary heart
disease. Heavy alcohol intake has been identified as a risk factor for
each of these variables,22 and all have been associated with risk of
AF,12 which implies that they could explain at least part of an
association between alcohol intake and risk of AF. For the latter 2
variables, we used time-varying covariates set to the date of first
hospitalization for each diagnosis.
We performed primary analyses using updated measures of
alcohol consumption and other covariates, in which we prospectively
assessed the risk of AF in between-examination increments, based on
determinations of alcohol consumption and other covariates derived
from the preceding questionnaire. We assessed the risk associated
with individual beverage types in similar fashion. In beveragespecific analyses, we examined the association of individual categories of intake of a given beverage relative to abstention from that
beverage while simultaneously adjusting for the standard covariates
that we incorporated into other models and for intake of each of the
other 2 beverage types.23 For tests of linear trend, we treated the
median value within categories of alcohol intake as a continuous
variable. To explore the dose-response relationship further, we also
performed regression analyses using linear splines24 with knots at
increments of 3 drinks/wk. We estimated the population-attributable
risk related to heavy alcohol intake using standard methods.25
Results
Baseline Characteristics
Table 1 shows the characteristics of participants categorized
at baseline according to usual alcohol consumption. Individ-
Downloaded from circ.ahajournals.org by on December 17, 2007
1738
Circulation
September 20, 2005
TABLE 1. Characteristics of 16 415 CCHS Participants Free of Clinical Cardiovascular Disease According to Usual
Alcohol Consumption
Weekly No. of Drinks
Women
Weekly intake (in drinks) of:
Beer
Wine
Liquor
Age, y*
Education ⬍8 y, %
Cohabitating, %
Ever smokers, %
BMI, kg/m2
Height, cm
Diabetes, %
Exercise intensity, %
None
Low
Moderate
High
FEV1, mean % predicted
Family Hx CVD, %
Income in lower tertile, %
Men
Weekly intake (in drinks) of:
Beer
Wine
Liquor
Age, y*
Education ⬍8 y, %
Cohabitating, %
Ever smokers, %
BMI, kg/m2
Height, cm
Diabetes, %
Exercise intensity, %
None
Low
Moderate
High
FEV1, mean % predicted
Family Hx CVD, %
Income in lower tertile, %
⬍1
1 to 6
7 to 13
14 to 20
ⱖ21 (Women)
21 to 27 (Men)
28 to 34
ⱖ35
n⫽2631
n⫽3969
n⫽1461
n⫽451
n⫽315
䡠䡠䡠
䡠䡠䡠
0
0
0
56 (33, 73)
61
63
67
24.3
161
3
0.9
1.2
0
50 (25, 68)
37
71
72
23.4
163
1
3.7
4.8
1.2
48 (26, 69)
26
70
77
23.0
164
1
4.7
5.7
1.3
49 (31, 69)
30
73
85
22.8
164
0
7.0
14.0
4.6
50 (33, 68)
29
65
85
23.6
164
1
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
26
54
19
1
14
60
25
1
13
57
28
2
20
54
25
2
23
49
26
2
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
⬍0.001
⬍0.001
⬍0.001
0.26
80.1
33
44
n⫽824
81.8
33
24
n⫽1703
82.0
35
22
n⫽1799
81.2
41
22
n⫽1269
81.0
32
23
n⫽609
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
n⫽435
䡠䡠䡠
䡠䡠䡠
䡠䡠䡠
n⫽949
⬍0.001
0.03
⬍0.001
0.1
0
0.1
54 (31, 75)
54
71
83
25.2
173
5
2.1
1.7
1.0
51 (25, 71)
40
77
81
24.8
175
3
7.0
1.7
1.7
50 (25, 70)
37
79
85
25.0
175
3
7.4
3.0
2.0
50 (26, 70)
36
77
89
25.1
175
3
16
2.1
2.0
52 (28, 70)
41
77
92
25.2
175
3
28
1.7
2.0
51 (33, 68)
44
74
91
25.9
174
5
35
1.1
2.0
50 (33, 68)
50
68
93
26.1
174
4
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
0.08
25
44
28
4
17
44
34
5
15
48
33
4
14
49
32
4
15
48
32
4
22
45
29
3
26
44
26
4
⬍0.001
0.01
⬍0.001
0.42
79.1
30
38
80.7
27
25
80.7
30
19
80.0
28
19
79.3
33
17
78.6
27
20
78.4
30
21
⬍0.001
0.09
⬍0.001
P
BMI indicates body mass index; Family Hx CVD, family history of cardiovascular disease.
Medians and probability values from Kruskal-Wallis tests are shown for continuous variables, and proportions and probability values from ␹2 tests are shown for
categorical variables.
*Median (5th percentile, 95th percentile).
uals in the lowest drinking category tended to be older and to
have lower income levels, whereas those in the higher
drinking categories were more likely to smoke. Women
consumed wine in the greatest amount, whereas men chiefly
consumed beer.
Average Alcohol Consumption and Risk of AF
We documented 1071 cases of incident AF. Of these 1071, 68
were diagnosed by a study ECG, 891 from hospitalization
records, and 112 from both sources. In both age- and
multivariable-adjusted analyses, the risk of AF was very
Downloaded from circ.ahajournals.org by on December 17, 2007
Mukamal et al
Alcohol and Atrial Fibrillation
1739
TABLE 2. Hazard Ratios and 95% CIs for Risk of AF According to Usual Alcohol Consumption Among CCHS Participants Free of
Cardiovascular Disease
Weekly No. of Drinks
⬍1
1 to 6
7 to 13
14 to 20
ⱖ21 (Women)
21 to 27 (Men)
28 to 34
ⱖ35
P*
Women
Cases
248
148
81
28
18
䡠䡠䡠
䡠䡠䡠
Person-years
59 466
63 380
27 249
9450
5970
䡠䡠䡠
䡠䡠䡠
Age-adjusted
1.00
0.91 (0.74–1.12) 1.04 (0.81–1.34) 1.03 (0.70–1.52) 1.26 (0.78–2.04)
䡠䡠䡠
䡠䡠䡠
0.35
Multivariable model† (95% CI)
1.00
0.89 (0.72–1.09) 0.99 (0.76–1.28) 0.96 (0.64–1.43) 1.04 (0.64–1.70)
䡠䡠䡠
䡠䡠䡠
0.87
With mediators‡
1.00
0.87 (0.70–1.07) 1.04 (0.80–1.35) 1.03 (0.69–1.53) 1.07 (0.65–1.75)
䡠䡠䡠
䡠䡠䡠
0.60
Men
Cases
Person-years
Age-adjusted (95% CI)
81
15 834
1.00
123
143
75
37
29
60
29 084
30 794
19 198
9345
7049
12 688
0.98 (0.74–1.30) 1.19 (0.91–1.57) 1.01 (0.74–1.39) 1.05 (0.71–1.55) 1.19 (0.78–1.82) 1.62 (1.16–2.27) 0.005
Multivariable model† (95% CI)
1.00
0.96 (0.73–1.28) 1.16 (0.88–1.53) 0.95 (0.69–1.31) 0.98 (0.66–1.45) 1.06 (0.69–1.64) 1.45 (1.02–2.04) 0.05
With mediators‡ (95% CI)
1.00
1.01 (0.76–1.34) 1.26 (0.95–1.66) 1.02 (0.74–1.41) 1.03 (0.69–1.54) 1.31 (0.85–2.02) 1.63 (1.15–2.31) 0.007
*Probability values derive from tests of trend.
†The multivariable model adjusted for age, smoking, education, cohabitation, family history of cardiovascular disease, diabetes, income, physical activity, BMI, FEV1,
and height.
‡Mediators included use of blood pressure medication, systolic blood pressure, and incident diagnoses of coronary heart disease and congestive heart failure
similar between abstainers and those who consumed up to 14
drinks per week (Table 2). Among men, we found a higher
risk among consumers of 35 or more drinks per week. We did
not find a similar risk among women who consumed 21 or
more drinks per week, but very few women consumed 28 or
more drinks per week (with a total of only 4 cases of AF
among such women). Given that 12.5% of men in the CCHS
consumed 35 or more drinks per week, we estimate that ⬇5%
of cases of AF among men were attributable to this level of
intake.
In stratified analyses (not shown), we found no evidence of
interaction among participants stratified by sex (P⫽0.96),
median age (for men, 55.8 years, P⫽0.22; for women, 56.8
years, P⫽0.32), or median body mass index (for men, 25.5
kg/m2, P⫽0.68; for women, 24.0 kg/m2, P⫽0.49). In all
cases, higher risk was generally restricted to men who drank
heavily. In exploratory analyses, using linear splines, of the
dose-response relationship among men, there appeared to a
threshold relationship similar to that suggested by analyses of
alcohol intake in categories. The risk of AF increased notably
at a threshold of ⬇35 drinks per week, with a relatively flat
relationship at lower levels of intake.
Potential Mediators of the Alcohol-AF Relation
We assessed a series of potential mediators of the association
of heavy alcohol intake with risk of AF, including blood
pressure, incident coronary heart disease during follow-up,
incident congestive heart failure during follow-up, and all 3.
As expected, all 3 were strongly and independently associated
with risk of AF. However, adjustment for these factors
individually or together had relatively little effect on our
results, consistent with the hypothesis that these factors do
not mediate the relation of alcohol use and AF. For example,
the relative risk of AF among men who consumed 35 or more
drinks per week was 1.45 in the basic model (as noted above)
and 1.63 (95% CI 1.15 to 2.31) with further adjustment for all
3 factors (Table 2).
Because the association of intake of 28 to 34 drinks per
week with risk of AF was stronger after adjustment for
mediators, we conducted a post hoc analysis that grouped all
men who consumed 28 or more drinks per week. In this
analysis, the hazard ratio for drinking 28 or more drinks per
week was 1.29 (95% CI 0.95 to 1.77) with multivariable
adjustment and 1.50 (95% CI 1.10 to 2.06) after the inclusion
of potential mediators. Given the latter hazard ratio, ⬇8% of
cases of AF among men were attributable to intake of 28 or
more drinks per week after adjustment for mediators.
Exploratory Analyses of Beverage Type and
Drinking Frequency
Table 3 shows results from exploratory beverage type analyses. There were no statistically significant associations
identified. It appeared that consumption of 21 or more drinks
per week of both beer and spirits tended to be related to
higher risk of AF than found with abstention from these
beverages among men.
We found no clear evidence that drinking frequency was
related to risk of AF in either women or men (data not
shown). This finding was consistent in analyses adjusted for
age, for all covariates, and for all covariates with additional
adjustment for overall quantity of alcohol use.
Discussion
In this prospective cohort study, alcohol intake of 35 or more
drinks per week was associated with a higher risk of AF, at
least among men. Previous studies of the relation of alcohol
use and risk of AF have not yielded consistent results. In a
case-control study from the UK General Practice Research
Database, Ruigomez and colleagues9 found that physicianreported alcohol use above 42 U per week was associated
Downloaded from circ.ahajournals.org by on December 17, 2007
1740
Circulation
September 20, 2005
TABLE 3. Adjusted Hazard Ratios for Risk of AF According to Updated Consumption of Individual
Alcoholic Beverages
Weekly No. of Drinks
⬍1
1 to 6
7 to 13
ⱖ14 (Women)
14 to 20 (Men)
ⱖ21
P
Women
Beer
Cases
Person-years
Hazard ratio (95% CI)
272
129
14
8
䡠䡠䡠
104 854
50 484
7067
3110
䡠䡠䡠
1.00
1.04 (0.83–1.30)
0.65 (0.38–1.12)
1.04 (0.51–2.13)
䡠䡠䡠
0.36
Wine
Cases
Person-years
Hazard ratio (95% CI)
316
160
30
17
䡠䡠䡠
78 983
71 646
9275
5612
䡠䡠䡠
1.00
0.95 (0.76–1.19)
1.09 (0.73–1.62)
1.13 (0.68–1.88)
䡠䡠䡠
0.70
Spirits
Cases
Person-years
Hazard ratio (95% CI)
395
92
29
7
䡠䡠䡠
116 140
42 229
5454
1692
䡠䡠䡠
1.00
0.85 (0.65–1.09)
1.11 (0.75–1.65)
1.19 (0.55–2.57)
䡠䡠䡠
123
212
92
52
69
0.58
Men
Beer
Cases
Person-years
Hazard ratio (95% CI)
24 274
45 758
24 182
12 454
17 323
1.00
1.12 (0.88–1.42)
1.01 (0.76–1.34)
1.06 (0.75–1.48)
1.27 (0.92–1.74)
288
203
37
13
7
0.26
Wine
Cases
Person-years
Hazard ratio (95% CI)
53 260
58 105
7862
3190
1575
1.00
0.85 (0.69–1.05)
0.97 (0.67–1.39)
0.81 (0.45–1.43)
0.99 (0.46–2.13)
269
220
49
25
14
0.65
Spirits
Cases
Person-years
Hazard ratio (95% CI)
58 027
53 766
7592
2766
1842
1.00
1.06 (0.86–1.31)
0.92 (0.67–1.28)
1.19 (0.76–1.87)
1.47 (0.85–2.56)
0.26
Probability values derive from tests of trend. Hazard ratios were adjusted for age, smoking, education, cohabitation, family history
of cardiovascular disease, diabetes, income, physical activity, body mass index, FEV1, height, and intake of the other 2 beverages.
with an OR of 2.4, with no evidence for an inverse association
at lower levels of consumption. An innovative case-control
analysis of the Framingham Study that incorporated persontime contributed to the study found an OR of 1.34 (95% CI
1.01 to 1.78) among consumers of more than 36 g (⬇3 drinks)
of alcohol daily but not at lower levels of intake.6 In a
case-control study from Helsinki, consumption of more than
30 g/d was acutely associated with onset of AF, whereas
lighter consumption was not associated with risk.8 An analysis of the Diet, Cancer, and Health cohort found 25% to 46%
higher risks of AF associated with average intake of ⬇20 g/d
or more among men but not among women.10 In sharp
contrast to these studies, Psaty and colleagues12 found alcohol
use to be inversely associated with risk of AF in a 3-year
follow-up study of the original Cardiovascular Health Study
cohort. To the best of our knowledge, ours is the first
prospective cohort study to support the findings of earlier
case-control studies that suggested a higher risk restricted to
heavy drinking.
The one previous study that assessed beverage type found
no differences in risk,10 despite some evidence that antioxi-
dants can modify AF risk. Mihm and colleagues26 found that
patients with AF have higher levels of 2 markers of oxidative
stress in right atrial myofibrillar isolates than do patients
without AF undergoing cardiac surgery. The same group
subsequently found that ascorbate, an antioxidant, alleviated
the pacing-induced shortening of the atrial effective refractory period in dogs and that supplemental ascorbate was
associated with an adjusted OR of 0.34 (95% CI 0.10 to 1.19)
for postoperative AF among patients undergoing CABG
surgery.27 In the present study, only the heaviest level of
overall alcohol intake was associated with higher risk, and no
single beverage was consumed in sufficient amounts to
compare their respective effects at that level of consumption
with confidence. Further studies, and perhaps meta-analyses
of existing studies, are needed to clarify this issue.
Heavier drinkers could sustain a higher risk of AF in a few
related ways. First, chronic heavy alcohol use itself could
affect atrial structure and size as a direct cardiotoxin, an effect
suggested in rat models.28 Second, chronic heavy drinking
could have direct proarrhythmic effects. Third, heavier drinkers are likely to have repeated exposure to episodic heavy
Downloaded from circ.ahajournals.org by on December 17, 2007
Mukamal et al
drinking (ie, binge drinking), which could increase the risk of
triggering a single episode of AF. To support this, porcine
models show that acute alcohol administration increases the
inducibility of AF, but mainly at very high blood alcohol
concentrations.29 Fourth, alcohol consumption could cause
brief and otherwise asymptomatic episodes of AF to become
persistent. A case-crossover study of patients with acute AF
could best assess this distinction30 but would not change the
importance of avoiding heavier drinking.
Specific limitations of the present study warrant discussion. As in any observational study, our results could be
influenced by differences between participants in factors
other than alcohol consumption for which we did not control.
For example, we did not have data on dietary factors other
than ethanol, although little is known about how such factors
affect risk of AF; this limitation may particularly affect
beverage type analyses.31 To have influenced our primary
results, any uncontrolled confounder would need to be
associated with both alcohol consumption and risk of AF and
generally be unrelated to the other covariates in our multivariable models.
Although we relied on self-reported alcohol consumption
in this study, the measures used to estimate alcohol consumption have been found to be valid when compared with a
formal dietary interview in Danish populations.32 Furthermore, we found a correlation of alcohol use and HDL
cholesterol of the expected magnitude.33
We relied on a hospital registry and periodic study-related
ECGs to document cases of AF. Because study examinations
occurred every 5 to 10 years, it is likely that we missed some
cases of AF among participants who were not hospitalized
during the follow-up period.
Although the CCHS is a population-based cohort study
with a high participation rate, participants are nearly all
white, native-born Danish adults. As a result, our results must
be generalized to other populations with an appropriate
degree of caution.
Although we found that risk of AF appeared to be higher
only among consumers of 35 drinks per week, it is difficult to
determine the exact nature of the dose-response relationship,
especially at high levels of intake that were relatively uncommon, although we explored this with spline analyses that
provided similar findings. This was especially true for analyses of individual beverage types, because consumption of
any individual beverage was limited in many cases. We also
lack power to exclude a modest increase in risk associated
with moderate drinking, although our data suggest that there
is not a substantially increased risk associated with moderate
intake.
In summary, alcohol consumption of 35 or more drinks per
week was associated with an increased risk of AF among men
in a population-based cohort of more than 16 000 adults. This
finding supports the widely held, but rarely examined, clinical observation that heavy alcohol use can lead to AF and
confirms the cardiotoxic effects of heavy drinking. Studies on
the effects of alcohol use on patients with established AF are
still needed.
Alcohol and Atrial Fibrillation
1741
Acknowledgments
This study was supported by the National Board of Health, the
Ministry for the Interior and Health, and The Health Insurance
Foundation, Denmark.
References
1. Ellison RC. Balancing the risks and benefits of moderate drinking. Ann
N Y Acad Sci. 2002;957:1– 6.
2. Klatsky AL. Moderate drinking and reduced risk of heart disease. Alcohol
Res Health. 1999;23:15–23.
3. Walsh CR, Larson MG, Evans JC, Djousse L, Ellison RC, Vasan RS,
Levy D. Alcohol consumption and risk for congestive heart failure in the
Framingham Heart Study. Ann Intern Med. 2002;136:181–191.
4. Abramson JL, Williams SA, Krumholz HM, Vaccarino V. Moderate
alcohol consumption and risk of heart failure among older persons.
JAMA. 2001;285:1971–1977.
5. Lowenstein SR, Gabow PA, Cramer J, Oliva PB, Ratner K. The role of
alcohol in new-onset atrial fibrillation. Arch Intern Med. 1983;143:
1882–1885.
6. Djousse L, Levy D, Benjamin EJ, Blease SJ, Russ A, Larson MG,
Massaro JM, D’Agostino RB, Wolf PA, Ellison RC. Long-term alcohol
consumption and the risk of atrial fibrillation in the Framingham Study.
Am J Cardiol. 2004;93:710 –713.
7. Rich EC, Siebold C, Campion B. Alcohol-related acute atrial fibrillation:
a case-control study and review of 40 patients. Arch Intern Med. 1985;
145:830 – 833.
8. Koskinen P, Kupari M, Leinonen H, Luomanmaki K. Alcohol and new
onset atrial fibrillation: a case-control study of a current series. Br
Heart J. 1987;57:468 – 473.
9. Ruigomez A, Johansson S, Wallander MA, Rodriguez LA. Incidence of
chronic atrial fibrillation in general practice and its treatment pattern.
J Clin Epidemiol. 2002;55:358 –363.
10. Frost L, Vestergaard P. Alcohol and risk of atrial fibrillation or flutter: a
cohort study. Arch Intern Med. 2004;164:1993–1998.
11. Koskinen P. A 4-year prospective follow-up study of the role of alcohol
in recurrences of atrial fibrillation. J Intern Med. 1991;230:423– 426.
12. Psaty BM, Manolio TA, Kuller LH, Kronmal RA, Cushman M, Fried
LP, White R, Furberg CD, Rautaharju PM. Incidence of and risk
factors for atrial fibrillation in older adults. Circulation. 1997;96:
2455–2461.
13. Kostis JB, Goodkind MJ, Skvaza H, Gerber NH Jr, Kuo PT. Effect of
alcohol on the atrial fibrillation threshold in dogs. Angiology. 1977;28:
583–587.
14. Nguyen TN, Friedman HS, Mokraoui AM. Effects of alcohol on experimental atrial fibrillation. Alcohol Clin Exp Res. 1987;11:474 – 476.
15. Becker U, Deis A, Sorensen TI, Gronbaek M, Borch-Johnsen K, Muller
CF, Schnohr P, Jensen G. Prediction of risk of liver disease by alcohol
intake, sex, and age: a prospective population study. Hepatology. 1996;
23:1025–1029.
16. Becker U, Deis A, Sorensen TIA, Gronbaek M, Florvall Muller C,
Schnohr P, Borch-Johnsen K, Jensen G. Alcohol intake in a population
study: assessment and characterization. Alcologia. 1995;7:35– 42.
17. Ellison RC, Myers RH, Zhang Y, Djousse L, Knox S, Williams RR,
Province MA. Effects of similarities in lifestyle habits on familial
aggregation of high density lipoprotein and low density lipoprotein
cholesterol: the NHLBI Family Heart Study. Am J Epidemiol. 1999;
150:910 –918.
18. Linn S, Carroll M, Johnson C, Fulwood R, Kalsbeek W, Briefel R.
High-density lipoprotein cholesterol and alcohol consumption in US
white and black adults: data from NHANES II. Am J Public Health.
1993;83:811– 816.
19. Vadstrup ES, Petersen L, Sorensen TI, Gronbaek M. Waist circumference
in relation to history of amount and type of alcohol: results from the
Copenhagen City Heart Study. Int J Obes Relat Metab Disord. 2003;27:
238 –246.
20. Friberg J, Scharling H, Gadsboll N, Jensen GB. Sex-specific increase in
the prevalence of atrial fibrillation (The Copenhagen City Heart Study).
Am J Cardiol. 2003;92:1419 –1423.
21. Truelsen T, Nielsen N, Boysen G, Gronbaek M. Self-reported stress and
risk of stroke: the Copenhagen City Heart Study. Stroke. 2003;34:
856 – 862.
Downloaded from circ.ahajournals.org by on December 17, 2007
1742
Circulation
September 20, 2005
22. Beilin LJ, Puddey IB. Alcohol, hypertension and cardiovascular disease:
implications for management. Clin Exp Hypertens. 1993;15:1157–1170.
23. Rimm EB, Stampfer MJ. Wine, beer, and spirits: are they really horses of
a different color? Circulation. 2002;105:2806 –2807.
24. Greenland S. Dose-response and trend analysis in epidemiology: alternatives to categorical analysis. Epidemiology. 1995;6:356 –365.
25. Wacholder S, Benichou J, Heineman EF, Hartge P, Hoover RN. Attributable risk: advantages of a broad definition of exposure. Am J Epidemiol.
1994;140:303–309.
26. Mihm MJ, Yu F, Carnes CA, Reiser PJ, McCarthy PM, Van Wagoner
DR, Bauer JA. Impaired myofibrillar energetics and oxidative injury
during human atrial fibrillation. Circulation. 2001;104:174 –180.
27. Carnes CA, Chung MK, Nakayama T, Nakayama H, Baliga RS, Piao S,
Kanderian A, Pavia S, Hamlin RL, McCarthy PM, Bauer JA, Van
Wagoner DR. Ascorbate attenuates atrial pacing-induced peroxynitrite
formation and electrical remodeling and decreases the incidence of postoperative atrial fibrillation. Circ Res. 2001;89:E32–E38.
28. Piano MR, Rosenblum C, Solaro RJ, Schwertz D. Calcium sensitivity and
the effect of the calcium sensitizing drug pimobendan in the alcoholic
isolated rat atrium. J Cardiovasc Pharmacol. 1999;33:237–242.
29. Anadon MJ, Almendral J, Gonzalez P, Zaballos M, Delcan JL, De
Guevara JL. Alcohol concentration determines the type of atrial arrhythmia induced in a porcine model of acute alcoholic intoxication. Pacing
Clin Electrophysiol. 1996;19:1962–1967.
30. Maclure M, Mittleman MA. Should we use a case-crossover design?
Annu Rev Public Health. 2000;21:193–221.
31. Tjonneland A, Gronbaek M, Stripp C, Overvad K. Wine intake and diet
in a random sample of 48763 Danish men and women. Am J Clin Nutr.
1999;69:49 –54.
32. Gronbaek M, Heitmann BL. Validity of self-reported intakes of wine,
beer and spirits in population studies. Eur J Clin Nutr. 1996;50:487– 490.
33. Giovannucci E, Colditz G, Stampfer MJ, Rimm EB, Litin L, Sampson L,
Willett WC. The assessment of alcohol consumption by a simple selfadministered questionnaire. Am J Epidemiol. 1991;133:810 – 817.
Downloaded from circ.ahajournals.org by on December 17, 2007
nature publishing group
ORIGINAL CONTRIBUTIONS
1
Alcohol Intake, Alcohol Dehydrogenase Genotypes,
and Liver Damage and Disease in the Danish
General Population
Janne S. Tolstrup, PhD1, Morten Grønbæk, MD, PhD, DMSc1,2, Anne Tybjærg-Hansen, MD, DMSc2,3 and
Børge G. Nordestgaard, MD, DMSc2,4
OBJECTIVES:
We tested the hypothesis that alcohol, alone and in combination with alcohol dehydrogenase
(ADH ) 1B and ADH1C genotypes, affects liver damage and disease in the general population.
METHODS:
Information on alcohol intake and on liver disease was obtained from 9,080 men and women
from the Copenhagen City Heart Study. Biochemical tests for the detection of liver damage were
specific for alanine aminotransferase (ALT), aspartate aminotransferase (AST)-to-ALT ratio (AST/
ALT), -glutamyl transpeptidase (-GT), albumin, bilirubin, alkaline phosphatase, coagulation
factors, and erythrocyte volume.
RESULTS:
Increasing alcohol intake was associated with increasing erythrocyte volume, AST/ALT, and
levels of ALT, -GT, albumin, bilirubin, coagulation factors, and with decreasing levels of alkaline
phosphatase. Multifactorially adjusted hazard ratios for alcoholic liver disease overall were 0.9
(95% confidence interval (CI), 0.6 –1.4), 1.4 (0.8 –2.5), 1.8 (0.9–3.5), and 4.1 (2.5 –7.0) for
an alcohol intake of 1–13, 14–20, 21–27, and ≥28 drinks per week, respectively, compared
with drinking < 1 drink per week (P for trend < 0.0001); the corresponding hazard ratios for
alcoholic liver cirrhosis were 1.7 (0.6 – 4.7), 2.0 (0.8 –7.1), 6.5 (2.0 –21), and 13 (4.6 –37) (P
for trend < 0.0001). ADH1B and ADH1C genotypes were not associated with and did not modify
the effect of alcohol on biochemical tests or risk of liver disease.
CONCLUSIONS: Increasing alcohol intake from none to low (1–6 drinks per week) through to moderate (7–20 drinks
per week) and excessive intake (≥21 drinks per week) leads to stepwise increases in signs of liver
damage with no threshold effect, and to an increased risk of liver disease. The minor changes in
biochemical tests for low alcohol intake may not account for subclinical liver disease.
Am J Gastroenterol advance online publication, 23 June 2009; doi:10.1038/ajg.2009.370
INTRODUCTION
Excessive alcohol intake associates with biochemical signs
of liver damage and disease, which has mostly been shown
in case–control studies (1). Whether this is also the case for
low-to-moderate levels of alcohol intake in individuals in the
general population is unknown. Alcohol-related disease, such
as liver cirrhosis, may take several years to develop. Intermediate states from a pathologically healthy organ to full-blown
disease exist in all ranges, but are often difficult to study in the
general population because the disease has not yet manifested
itself into clinical symptoms. Biomarkers for liver damage
are appealing alternatives, as they are immediate and sensitive measures, objectively evaluating the current state of the
organ.
The production of toxic acetaldehyde within cells is believed
to be one of the causes of liver damage and disease in response
1
Center for Alcohol Research, National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark; 2The Copenhagen City Heart Study,
Faculty of Health Sciences, Bispebjerg Hospital, Copenhagen University Hospital, Copenhagen, Denmark; 3Department of Clinical Biochemistry, Faculty of Health
Sciences, Rigshospitalet, Copenhagen University Hospital, University of Copenhagen, Copenhagen, Denmark; 4Department of Clinical Biochemistry, Faculty
of Health Sciences, Herlev Hospital, Copenhagen University Hospital, Herlev, Denmark. Correspondence: Børge G. Nordestgaard, MD, DMSc, Department of
Clinical Biochemistry, Herlev Hospital, Herlev Ringvej 75, 2730 Herlev, Denmark. E-mail: [email protected]
Received 22 August 2008; accepted 5 February 2009
© 2009 by the American College of Gastroenterology
The American Journal of GASTROENTEROLOGY
LIVER AND BILIARY TRACT
see related editorial on page x
LIVER AND BILIARY TRACT
2
Tolstrup et al.
to alcohol intake (2–4). Alcohol is degraded in the liver to
acetaldehyde, which is then further degraded to acetate by
the successive actions of alcohol dehydrogenase (ADH) and
acetaldehyde dehydrogenase. Among Europeans, a common
genetic variation affecting alcohol degradation rates is found
in ADH1B and ADH1C genes (5). Therefore, ADH genotypes producing enzymes with different velocities of alcohol
conversion to acetaldehyde could lead to different intracellular acetaldehyde concentrations in response to alcohol
intake, and thus to differences in damage and disease of the
liver (6–10).
We tested the hypotheses that alcohol intake, alone and combined with ADH1B and ADH1C genotypes, affects biochemical
signs of liver damage and risk of liver disease in the general
population. For this purpose, we studied 9,080 individuals
from the Danish general population, all well characterized with
respect to alcohol intake (11).
cigars, and pipes). Assuming one cigarette to be equivalent
to 1 g of tobacco, one cheroot or one cigar to be equivalent to
3 g of tobacco, and one pipe to be equivalent to 5 g of tobacco,
the total amount of daily smoking was calculated for current
smokers. School education was reported as number of years of
basic schooling.
Genotyping procedures
The ADH1B2 allele (rs1229984, Arg47His in exon 3) and
ADH1C2 allele (rs698, Ile349Val in exon 8) were identified by
means of a duplex PCR, followed by Nanogen microelectronic
chip technology (Nanogen NMW 1000 Nanochip Molecular
Biology Workstation (14)) using standard conditions (details
available from authors). In a validation study, the accuracy of
the Nanogen method was found to be comparable with restriction fragment length polymorphism (15).
Outcome measures
METHODS
Study population
Our data originate from The Copenhagen City Heart Study,
which consists of four consecutive studies conducted in Denmark between 1976 and 1978, 1981 and 1983, 1991 and 1994,
and between 2001 and 2003. Participants who had been living in the Copenhagen area for more than 20 years were randomly chosen from the general population. Before visiting
the study clinic, participants completed a questionnaire that
included questions on alcohol intake and other lifestyle factors. At the clinic visit, physical examinations were carried out
and questionnaires were checked for missing information and
any uncertainties were clarified. In addition, blood samples
were obtained for measuring different biochemical and DNA
markers at the 1991–1994 examination. Hence, participants
who attended this examination were included in this study.
Of 10,135 individuals who participated in this examination
(response rate = 61%), 91% donated blood for biochemical and
DNA analyses. Participants of Asian or Black descent (n = 142),
or with missing questionnaire data (n = 14), were excluded
from further study. In all, 9,080 Whites of Danish ethnicity
were eligible for this study, some of whom also participated
in the examinations between 1976 and 1978 (n = 6,408), 1981
and 1983 (n = 6,615), and between 2001 and 2003 (n = 4,684).
All participants gave informed consent and the Ethics Committee for Copenhagen and Frederiksberg approved the study
(#100.2039/91). Enrollment and examination procedures have
been described in more detail elsewhere (12,13).
Variables
The amount of usual alcohol intake was reported as weekly
consumption of beer (in bottles), wine (in glasses), and spirits
(in units). Assuming one drink to be equal to 12 g of pure alcohol, a measure of total weekly alcohol intake was calculated.
Smoking was reported as the status (never, ex, or current) and
amount of smoking (in number of daily cigarettes, cheroots,
The American Journal of GASTROENTEROLOGY
Liver damage. Plasma biochemical tests indicating liver damage
(alanine aminotransferase (ALT), aspartate aminotransferase
(AST)-to-ALT ratio (AST/ALT), -glutamyl transpeptidase
(-GT), albumin, bilirubin, alkaline phosphatase, coagulation factors II, VII, and X, and erythrocyte volume) were used.
These were measured using standard hospital assays (Konelab
and Boehringer Mannheim) subjected to a daily internal quality control assessing assay precision and a monthly external
quality control assessing assay accuracy. For the 1991–1994
examination, AST and albumin were measured immediately
after blood was collected and ALT, -GT, bilirubin, and alkaline phosphatase were analyzed on plasma frozen at − 80 °C
for 12–15 years, whereas for the 2001–2003 examination, ALT,
-GT, albumin, bilirubin, alkaline phosphatase, coagulation
factors II, VII, and X, and erythrocyte volume were also analyzed immediately after blood was collected.
Liver disease. Information on diseases of the liver was obtained
from the Danish Patient Registry (16) and from the Danish
Causes of Death Registry (17), in which all somatic hospitalizations and causes of death are registered for the entire country. Diagnoses are classified according to the World Health
Organization’s International Classification of Diseases (ICD),
8th and 10th revision. Vital status of the participants was obtained from the Danish Civil Registration System, wherein information on address and vital status is registered for every
Danish citizen. The following diagnoses were obtained: alcohol-induced liver cirrhosis (ICD-8: 571.09 and ICD10: K703)
and any alcohol-induced liver disease (ICD-8: 155.09–155.89,
456.00–456.09, 570.00–570.99, 571.09, 571.10–571.99, 573.00–
573.09, 785.19–785.39 and ICD10: K70 I85 C22 R18).
Statistical analysis
Associations between alcohol intake, ADH1B and ADH1C
genotypes, and biochemical tests were investigated by general linear models. Biochemical tests were log transformed
to approximate normal distributions. We used the 2-test to
VOLUME 104 | XXX 2009 www.amjgastro.com
Alcohol Intake and Liver Damage and Disease
Table 1. Characteristics of participants of the Copenhagen
City Heart Study 1991–1994 examination
Variable
Men (n = 4,039)
Women (n = 5,041)
Age (years)
60 (35, 76)
62 (36, 76)
Alcohol intake (drinks per
week)a,b
10 (0, 32)
3.0 (0, 15)
1/1 (slow)
95.4
95.7
1/2 + 2/2 (fast)
4.6
4.3
1/1 (fast)
34
34
2/1 (intermediate)
48
48
2/2 (slow)
18
18
≤ 7 years
34
35
8–10 years
38
39
28
26
26 (22, 31)
24 (20, 31)
53
46
20 (7, 30)
15 (5, 21)
a
ADH1B genotype (%)
ADH1C genotype (%)
Education (%)
≥ 11 years
2 a
Body mass index (kg/m )
Current smokers (%)
c
Amount of smoking (g/day)
ADH, alcohol dehydrogenase.
a
Median (10th, 90th percentile). bOne drink corresponds to 12 g of pure alcohol.
c
Median (10th, 90th percentile), among current smokers only.
RESULTS
Baseline characteristics
Of the 9,080 participants, 4,039 (44%) were men (Table 1).
The mean age was 60 years (10th, 90th percentile = 35, 76) for
men and 62 years (10th, 90th percentile = 36, 76) for women.
The median alcohol intake was 10 drinks per week (10th,
90th percentile = 0, 32) among men and 3 (10th, 90th percentile = 0, 15) among women, and 53% of the men and 46% of
the women were current smokers (Table 1). The frequency
of ADH1B and ADH1C genotypes coding for the most active
enzymes was 4.5 (ADH1B·1/2 + 2/2) and 34% (ADH1C·1/1).
Both ADH1B and ADH1C genotypes were in the Hardy–
Weinberg equilibrium (P = 0.43 and P = 0.64, respectively).
Linkage disequilibrium coefficients Lewontin’s D⬘ was 0.90
and the correlation coefficient r2 was 0.01 between ADH1B2
and ADH1C1, both coding for the fast alcohol degradation
enzymatic forms.
Alcohol intake, ADH genotypes, and liver damage
Increasing amount of alcohol intake was associated with
increasing levels of ALT, AST/ALT, -GT, albumin (at
the 2001–2003 examination only), bilirubin, coagulation
factors II, VII, and X, and erythrocyte volume, and with
decreasing levels of alkaline phosphatase among both men
and women (Figures 1 and 2). ADH1B and ADH1C genotypes were not associated with any of these biochemical
© 2009 by the American College of Gastroenterology
signs of liver damage (data not shown). ADH1B and ADH1C
genotypes did not interact with alcohol intake on any of the
biochemical signs of liver damage in either gender (all
P values >0.05).
Alcohol intake, ADH genotypes, and liver disease
Increasing amount of alcohol intake was associated with increasing risk of alcoholic liver disease overall and with alcoholic
liver cirrhosis (Table 2). The multifactorially adjusted hazard
ratios for alcoholic liver disease overall were 0.9 (95% confidence interval (95% CI), 0.6–1.4) for a weekly alcohol intake of
1–13 drinks per week, 1.4 (95% CI, 0.8–2.5) for 14–20 drinks
per week, 1.8 (95% CI, 0.9–3.5) for 21–27 drinks per week,
and 4.1 (95% CI, 2.5–7.0) for 28 + drinks per week, compared
with drinking < 1 drink per week; the corresponding hazard
ratios for alcoholic liver cirrhosis were 1.7 (95% CI, 0.6–4.7),
2.0 (95% CI, 0.8–7.1), 6.5 (95% CI, 2.0–21), and 13 (95% CI,
4.6–37) (P for trend < 0.0001). ADH1B and ADH1C genotypes
were not associated with alcoholic liver disease. The association between alcohol intake and liver disease did not seem to
be modified by the ADH1C genotype, as hazard ratios for liver
disease were comparable within the strata of the ADH1C genotype (Figure 3); we did not have enough statistical power to
stratify similarly for the ADH1B genotype. Alcohol intake did
The American Journal of GASTROENTEROLOGY
LIVER AND BILIARY TRACT
determine whether ADH1B and ADH1C genotypes were in the
Hardy–Weinberg equilibrium (18). In all analyses, ADH1B·1/2
was combined with ADH1B·2/2 because of the low number in
the latter group (n = 5).
Risk estimates for liver disease during follow-up were computed by means of Cox proportional hazard regression models.
Age was used as the time axis to ensure that the estimation procedure was based on comparisons of individuals of the same
age, and hence confounding by age was eliminated. Analyses
were corrected for delayed entry. The observation time for each
participant was the period from the person’s first participation in the Copenhagen City Heart Study, until date of disease,
death from other causes, emigration outside Denmark, or 1
August 2007, whichever came first. Information on independent variables was updated for participants who participated in
more than one examination. Assumptions of the proportional
hazards models were tested and no violations were detected.
Analyses were adjusted for factors known to be associated with
alcohol intake, such as age, length of school education ( ⭐ 7,
8–10, ⭓ 11 years), and smoking (never smokers, ex-smokers,
current smokers of 1–14, 15–24, >24 g of tobacco per day).
The test for linear trend in the risk estimates for alcohol
intake was carried out by treating the median within categories as a continuous variable, and the test for linear trend in
ADH1C genotypes was carried out by coding the genotypes,
0, 1, and 2. Linkage disequilibrium was expressed as r2 and D⬘
(19,20).
3
Women
(n=4,629)
P<0.0001
P =0.001
AST/ALT
ALT (U/l)
20
15
10
P =0.004
P<0.0001
P=0.009
P =0.03
P<0.0001
2.0
1.5
14
Bilirubin (µM)
γ-GT (U/l)
P <0.0001
40
12
10
8
P =0.15
P = 0.70
500
<1 (30%)
1–6 (38%)
7–13 (19%)
14 –20 (7%)
21+ (6%)
<1 (13%)
1–6 (24 %)
7–13 (25%)
14 –20 (13%)
21– 27 (11%)
28+ (14%)
490
Drinks per week
Alkaline
phosphatase (U/l)
20
480
P =0.006
1.0
P<0.0001
60
510
Women
(n =4,629)
2.5
25
80
Men
(n=3,633)
<1 (30%)
1–6 (38%)
7–13 (19%)
14 –20 (7%)
21+ (6%)
Men
(n =3,633)
<1 (13%)
1–6 (24 %)
7–13 (25%)
14 –20 (13%)
21–27 (11%)
28+ (14%)
Tolstrup et al.
Albumin (µM)
LIVER AND BILIARY TRACT
4
100
90
80
70
Drinks per week
Figure 1. Biochemical signs of liver damage according to alcohol intake in the general population. Values are from the 1991–1994 examination of the
Copenhagen City Heart Study. Average values shown are estimated from generalized linear models and apply to a reference person (age 40 years,
never smoker, and education ⭐ 7 years). P values are for linear trend. AST, aspartate aminotransferase; ALT, alanine aminotransferase, -GT, -glutamyl
transpeptidase.
not interact with ADH1B and ADH1C genotypes on the risk of
liver disease (all P values >0.05).
DISCUSSION
In this European general population study, we found that
increasing alcohol intake from none to low (1–6 drinks per
week) through to moderate (7–20 drinks per week) and excessive intake (21 + drinks per week) leads to stepwise increases
in signs of liver damage with no threshold effect, and to an
increased risk of liver disease. For some of the biochemical tests,
only small differences in blood levels between low, moderate,
and excessive alcohol intake were observed, and such minor
changes may not be indicative of a subclinical damage of the
liver. Hence, these findings may only have minor clinical relevance; however, mechanistically, it is nevertheless an interesting
finding that levels of these biochemical markers were influenced at very low alcohol intakes with no apparent threshold
effect. ADH1B and ADH1C genotypes were not associated with
biochemical signs of liver damage, or with risk of liver disease.
This study and most other studies (6) have not been able to
show any effect of ADH1B and ADH1C genotypes on risk of
alcoholic liver disease. Furthermore, in support of this finding,
The American Journal of GASTROENTEROLOGY
we could not even show the influence of these genotypes on the
biochemical signs of liver damage in individuals in the general
population.
Increased levels of ALT, AST/ALT, γ-GT, bilirubin, and
alkaline phosphatase, and decreased levels of coagulation factors II + VII + X, are generally indicative of acute liver damage
(21,22). However, these biochemical tests have never previously been measured in such a large sample from the general
population with detailed information on usual alcohol intake.
Nevertheless, the increased levels of ALT, AST/ALT, γ-GT, and
bilirubin observed in this study in those with the highest alcohol intake is in accordance with that expected, i.e., increased
alcohol intake associates with increased biochemical signs of
liver cell damage.
Surprisingly, we observed that the amount of alcohol intake
was inversely associated with levels of alkaline phosphatase in
the general population, which is unlikely to be a chance finding
because results were highly significant, consistent in men and
women, and observed repeatedly at examinations conducted
between 1991 and 1993 and between 2001 and 2003. A possible explanation for this finding is that, although increased
alkaline phosphatase values are indicative of acute hepatobiliary tissue damage (21,22), a sustained high intake of alcohol,
VOLUME 104 | XXX 2009 www.amjgastro.com
Women
(n =2,704)
P <0.0001
P =0.0001
24
19
14
P<0.0001
P<0.0001
91
Erythrocyte
volume (10–15 l)
γ-GT (U/l)
80
60
40
20
P =0.04
P =0.0001
650
625
P <0.0001
P =0.0009
P =0.005
P =0.01
P <0.0001
87
11
9
7
120
110
<1 (16%)
1–6 (44%)
7–13 (25%)
14–20 (8%)
21+ (7%)
100
Drinks per week
90
80
70
60
<1 (6%)
1–6 (29%)
7–13 (28%)
14 –20 (13%)
21–27 (12%)
28+ (12%)
P<0.0001
Alkaline
phosphatase (U/l)
P =0.004
<1 (6%)
1–6 (29%)
7–13 (28%)
14–20 (13%)
21–27 (12%)
28+ (12%)
Coagulation factors
II, VII, and X (%)
P <0.0001
89
13
600
130
Women
(n=2,704)
85
Bilirubin (µM)
Albumin (µM)
675
Men
(n=1,967)
<1 (16%)
1–6 (44%)
7–13 (25%)
14 –20 (8%)
21+ (7%)
ALT (U/l)
29
Men
(n =1,967)
Drinks per week
Figure 2. Biochemical signs of liver damage according to alcohol intake in the general population. Values are from the 2001–2003 examination of the
Copenhagen City Heart Study. Average values shown are estimated from generalized linear models and apply for a reference person (age 40 years, never
smoker, and education ⭐ 7 years). P values are for linear trend. ALT, alanine aminotransferase, -GT, -glutamyl transpeptidase.
possibly leading to mild, clinically undetected chronic pancreatitis with secondary mild biliary obstruction and/or hepatocyte
deaths, could reduce the number of cells producing alkaline
phosphatase. Mild, clinically unrecognized biliary obstruction
could also be the reason for the observed increased bilirubin
levels in those in the general population with the highest
alcohol intake.
It was also somewhat surprising that increased alcohol
intake associated with increased levels of coagulation factors II + VII + X, equivalent to decreased prothrombin time.
Chronic damage of hepatocytes by alcohol usually leads to a
reduced capacity to produce proteins, such as these coagulation factors and albumin, as seen in alcoholic liver cirrhosis, whereas in less-severe forms of alcoholic liver disease,
prothrombin time and thus levels of coagulation factors
II + VII + X may not be affected (21,22). However, our obser© 2009 by the American College of Gastroenterology
vation agrees with previous epidemiological data that heavy
alcohol consumption leads to a more procoagulant state with
higher levels of coagulation factor VII (23). Alternatively,
because albumin also increased in response to increasing
alcohol intake, the mechanism behind the increased levels of
coagulation factors II + VII + X may be a different, but not yet
understood, one.
Limitations include the fact that we studied Whites only.
Therefore, our results may not necessarily apply to other
ethnic groups. Our study had several strengths. First of all,
the sample size is large and the wide range of alcohol intake
provided enough statistical power to study the effects of
low-to-moderate, as well as excessive, alcohol consumption.
Furthermore, all participants were men and women from
the general population of Danish descent. Hence, population stratification is unlikely to have affected our results.
The American Journal of GASTROENTEROLOGY
5
LIVER AND BILIARY TRACT
Alcohol Intake and Liver Damage and Disease
6
Tolstrup et al.
6
Alcoholic liver
cirrhosis (n = 64)
Alcoholic liver
disease (n = 171)
Sex and age adjusted
Alcohol (drinks per week)
≥28 (n =602)
16 (5.5 – 43)
Alcohol intake·ADH1C genotype interaction: P =0.96
5
Hazard ratio
LIVER AND BILIARY TRACT
Table 2. Risk of liver disease according to alcohol intake and
ADH1B and ADH1C genotype
4
3
2
ADH1C·1/1 (n =3,087)
4.7 (2.8 – 7.9)
1
21–27 (n =746)
7.4 (2.3 – 23)
1.9 (1.0 – 3.8)
14–20 (n =857)
2.2 (0.6 – 7.9)
1.5 (0.8 – 2.6)
1–13 (n =4,917)
1.9 (0.7 – 5.0)
0.9 (0.6 –1.4)
1.0
1.0
< 0.0001
< 0.0001
0.9 (0.3 – 3.0)
1.1 (0.5 – 2.4)
1.0
1.0
2/2 (n =1,612)
1.2 (0.6 – 2.5)
1.0 (0.6 –1.5)
2/1 (n =4,381)
1.1 (0.6 – 2.0)
1.1 (0.8 –1.5)
1/1 (n =3,087)
1.0
1.0
P value for trend
0.87
0.80
≥28 (n =602)
13 (4.6 – 37)
4.1 (2.5 – 7.0)
21–27 (n =746)
6.5 (2.0 – 21)
1.8 (0.9 – 3.5)
14–20 (n =857)
2.0 (0.6 – 7.1)
1.4 (0.8 – 2.5)
1–13 (n =4,917)
1.7 (0.6 – 4.7)
0.9 (0.6 –1.4)
1.0
1.0
< 0.0001
< 0.0001
ADH1C·1/2 (n =4,381)
ADH1C·2/2 (n =1,612)
0
<1
< 1 (n =1,958)
P value for trend
ADH1B genotype
1/1 (n =8,675)
1/2 + 2/2 (n =405)
ADH1C genotype
Multifactorial adjusted
Alcohol (drinks per week)a
< 1 (n =1,958)
P value for trend
ADH1B genotype
b
1/1 (n =8,653)
0.9 (0.3 – 2.8)
1.1 (0.5 – 2.3)
1.0
1.0
2/2 (n =1,607 )
1.2 (0.6 – 2.5)
1.0 (0.6 –1.4)
2/1 (n =4,371)
1.1 (0.6 – 2.0)
1.1 (0.8 –1.5)
1/1 (n =3,077)
1.0
1.0
P value for trend
0.86
0.82
1/2 + 2/2 (n =402)
ADH1C genotype b
ADH, alcohol dehydrogenase.
Values are hazard ratios (95% confidence intervals) from Cox regression.
P values for trend test examine whether increased levels of alcohol intake or
number of ADH1C·2 alleles associate with increased hazard ratios.
a
Adjusted for sex, age, smoking, and education. bAdjusted for sex, age, smoking,
education, and other ADH genotypes.
We had several different biochemical tests for liver damage,
which were obtained objectively from the study participants.
The American Journal of GASTROENTEROLOGY
1–13
14 –20
21–27
28+
Alcohol intake (drinks per week)
n = 1,958
4,917
857
746
602
Figure 3. Hazard ratio of alcoholic liver disease according to alcohol intake
and alcohol dehydrogenase (ADH)1C genotype by Cox regression.
Hence, it is unlikely that these measures are differentially
biased according to alcohol intake.
Information on the amount of alcohol was obtained by selfreport in a food frequency questionnaire. This method is found
to be valid when compared with a formal dietary interview in
Danish populations (24,25). Although this comparison does not
represent a true validation of frequency questionnaires, dietary
interviews are considered to convey more accurate information
than frequency questionnaires. Furthermore, the correlation
between alcohol and high-density lipoprotein in this cohort is
of the expected magnitude (26).
In conclusion, in the general population, increasing alcohol
intake leads to increasing signs of liver damage with no threshold effect, and to an increased risk of liver disease. The minor
changes in biochemical tests for low alcohol intake may not
account for subclinical liver disease.
CONFLICT OF INTEREST
Guarantor of the article: Janne S. Tolstrup, PhD.
Specific author contributions: Conception and study design,
data analysis and interpretation, and writing the manuscript:
Janne S. Tolstrup; conception and study design, data analysis
and interpretation, and critical revision of the paper: Morten
Grønbæk, Anne Tybjærg-Hansen, and Børge G. Nordestgaard.
All authors approved the final version of the paper.
Financial support: None.
Potential competing interests: None.
ACKNOWLEDGMENTS
The authors thank the participants in the Copenhagen City
Heart Study for their outstanding cooperation. This research
was supported (in part) by the Danish Graduate School of
Public Health, the Danish Heart Foundation, Chief Physician Johan Boserup and Lise Boserup Foundation, the Health
Insurance Foundation, the Ministry of the Interior and
Health, the Danish Cancer Society, and the Danish National
Board of Health.
VOLUME 104 | XXX 2009 www.amjgastro.com
Study Highlights
WHAT IS CURRENT KNOWLEDGE
Excessive alcohol intake is associated with risk of liver
damage and disease; however, whether this is also the
case for low-to-moderate levels of alcohol intake is
unknown.
The significance of alcohol dehydrogenase (ADH ) gene
variations for risk of liver damage and disease among
Caucasians is unknown, especially for the ADH1B variation, which is unevenly distributed in this population.
WHAT IS NEW HERE
Increasing alcohol intake from none through to moderate and excessive intake leads to stepwise increases in
signs of liver damage with no threshold effect, and to an
increased risk of liver disease.
Variations in ADH1B and ADH1C genes is not independently associated with risk of liver damage or with liver
disease, and do not seem to modify the effect of alcohol
intake on any of these conditions.
3
3
3
3
REFERENCES
1. Corrao G, Bagnardi V, Zambon A et al. A meta-analysis of alcohol consumption and the risk of 15 diseases. Prev Med 2004;38:613–9.
2. Lieber CS. Alcoholic fatty liver: its pathogenesis and mechanism of progression to inflammation and fibrosis. Alcohol 2004;34:9–19.
3. Adachi M, Brenner DA. Clinical syndromes of alcoholic liver disease. Dig
Dis 2005;23:255–63.
4. Witt H, Apte MV, Keim V et al. Chronic pancreatitis: challenges and advances in pathogenesis, genetics, diagnosis, and therapy. Gastroenterology
2007;132:1557–73.
5. Bosron WF, Li TK. Genetic polymorphism of human liver alcohol and
aldehyde dehydrogenases, and their relationship to alcohol metabolism and
alcoholism. Hepatology 1986;6:502–10.
6. Zintzaras E, Stefanidis I, Santos M et al. Do alcohol-metabolizing enzyme
gene polymorphisms increase the risk of alcoholism and alcoholic liver
disease? Hepatology 2006;43:352–61.
7. Sun L, Konig IR, Jacobs A et al. Mean corpuscular volume and ADH1C
genotype in white patients with alcohol-associated diseases. Alcohol Clin
Exp Res 2005;29:788–93.
8. Cichoz-Lach H, Partycka J, Nesina I et al. Genetic polymorphism of alcohol
dehydrogenase 3 in alcohol liver cirrhosis and in alcohol chronic pancreatitis. Alcohol Alcohol 2006;41:14–7.
© 2009 by the American College of Gastroenterology
9. Matsumoto M, Takahashi H, Maruyama K et al. Genotypes of
alcohol-metabolizing enzymes and the risk for alcoholic chronic
pancreatitis in Japanese alcoholics. Alcohol Clin Exp Res 1996;20:
289A–92A.
10. Verlaan M, te Morsche RH, Roelofs HM et al. Genetic polymorphisms in
alcohol-metabolizing enzymes and chronic pancreatitis. Alcohol Alcohol
2004;39:20–4.
11. Tolstrup JS, Nordestgaard BG, Rasmussen S et al. Alcoholism and alcohol
drinking habits predicted from alcohol dehydrogenase genes. Pharmacogenomics J 2008;8:220–7.
12. The Copenhagen City Heart Study. Osterbroundersogelsen. A book of
tables with data from the first examination (1976–78) and a five year
follow-up (1981–83). The Copenhagen City Heart Study Group. Scand J
Soc Med 1989;41 (suppl): 161–0.
13. Schnohr P, Jensen G, Lange P, et al., The Copenhagen City Heart
Study—Østerbroundersøgelsen. Tables with data from the third examination 1991–9. Eur Heart J 2001;3 (suppl): H81–3.
14. Sosnowski RG, Tu E, Butler WF et al. Rapid determination of single base
mismatch mutations in DNA hybrids by direct electric field control. Proc
Natl Acad Sci USA 1997;94:1119–23.
15. Sethi AA, Tybjaerg-Hansen A, Andersen RV et al. Nanogen microelectronic chip for large-scale genotyping. Clin Chem 2004;50:443–6.
16. Jürgensen HJ, Frølund C, Gustafsen J et al. Registration of diagnoses in
the Danish National Registry of patients. Methods Inf Med 1986;25:
158–64.
17. Juel K, Helweg-Larsen K. The Danish registers of causes of death. Danish
Med Bull 1999;46:354–7.
18. Cavalli-Sforza L, Bodmer W. The Genetics of Human Populations. Dover:
New York, 1999.
19. Lewontin RC. The interaction of selection and linkage.I. General considerations: heterotic models. Genetics 1964;49:49–67.
20. Pritchard JK, Przeworski M. Linkage disequilibrium in humans: models
and data. Am J Hum Genet 2001;69:1–14.
21. Marks V, Cantor T, Mesko D et al. Springer: Heidelberg, 2002.
22. Burtis C, Ashwood E, Bruns D. Tietz Textbook of Clinical Chemistry and
Molecular Diagnosis, 4th edn Elsevier Saunders: St. Louis, 2006.
23. Lee KW, Lip GY. Effects of lifestyle on hemostasis, fibrinolysis, and
platelet reactivity: a systematic review. Arch Intern Med 2003;163:
2368–92.
24. Tunstall-Pedoe H, Kuulasmaa K, Mahonen M et al. Contribution of trends
in survival and coronary-event rates to changes in coronary heart disease
mortality: 10-year results from 37 WHO MONICA project populations.
Monitoring trends and determinants in cardiovascular disease. Lancet
1999;353:1547–57.
25. Gronbaek M, Heitmann BL. Validity of self-reported intakes of wine, beer
and spirits in population studies. Eur J Clin Nutr 1996;50:487–90.
26. Johansen D, Andersen PK, Jensen MK et al. Nonlinear relation between
alcohol intake and high-density lipoprotein cholesterol level: results
from the Copenhagen City Heart Study. Alcohol Clin Exp Res
2003;27:1305–9.
The American Journal of GASTROENTEROLOGY
7
LIVER AND BILIARY TRACT
Alcohol Intake and Liver Damage and Disease
American Journal of Epidemiology Advance Access published September 8, 2008
American Journal of Epidemiology
ª The Author 2008. Published by the Johns Hopkins Bloomberg School of Public Health.
All rights reserved. For permissions, please e-mail: [email protected].
Doi:10.1093/aje/kwn222
Original Contribution
Risk of Pancreatitis According to Alcohol Drinking Habits: A Population-based
Cohort Study
Louise Kristiansen, Morten Grønbæk, Ulrik Becker, and Janne Schurmann Tolstrup
Received for publication May 6, 2008; accepted for publication June 25, 2008.
The association between alcohol intake and pancreatitis has been examined previously in case-control studies,
mostly consisting of men. The significance of beverage type and drinking pattern is unknown. The objective of this
study was to assess the association between amount, type, and frequency of alcohol intake and risk of pancreatitis.
For this purpose, the authors used data on 17,905 men and women who participated in the Copenhagen City Heart
Study in 1976–1978, 1981–1983, 1991–1994, and 2001–2003 in Copenhagen, Denmark. Alcohol intake and
covariates were assessed by questionnaire. Information on pancreatitis was obtained from national registers.
A high alcohol intake was associated with a higher risk of pancreatitis. Hazard ratios associated with drinking
1–6, 7–13, 14–20, 21–34, 35–48, and >48 drinks/week were 1.1 (95% confidence interval (CI): 0.8, 1.6), 1.2 (95%
CI: 0.8, 1.8), 1.3 (95% CI: 0.8, 2.1), 1.3 (95% CI: 0.7, 2.2), 2.6 (95% CI: 1.4, 4.8), and 3.0 (95% CI: 1.6, 5.7),
respectively, compared with 0 drinks/week (Ptrend < 0.001). Associations were similar for men and women.
Drinking frequency did not seem to be independently associated with pancreatitis.
alcohol drinking; alcoholic beverages; alcohol-induced disorders; pancreatitis; prospective studies
Abbreviations: CI, confidence interval; ICD, International Classification of Diseases.
tect against gallstone disease (7, 8), the association between
alcohol and risk of pancreatitis might be tempered by gallstone disease.
In this study, we examined the association between alcohol intake and risk of pancreatitis in a large prospective
cohort consisting of men and women from the general Danish
population. Furthermore, we aimed at addressing whether
there may be specific effects of beverage type or drinking
frequency and whether the association is mediated by gallstone
disease.
As early as 1878, alcohol was proposed as a risk factor for
pancreatitis (1), and it is now considered well known that
alcohol increases the risk of pancreatitis. However, epidemiologic studies on the quantitative aspect of the association
between alcohol intake and pancreatitis are sparse. To date,
only 4 case-control studies (1–4) and 1 ecologic study (5) on
this subject have been published. In all these studies, an
association was found between alcohol intake and risk of
pancreatitis in men; however, only one of the case-control
studies included women, where surprisingly, no increased
risk of pancreatitis according to alcohol was observed (2).
No studies have assessed the risk of pancreatitis associated
with other dimensions of alcohol intake, such as beverage
type and drinking frequency. It is also uncertain whether
a threshold exists, that is, a level of alcohol intake under
which the risk of pancreatitis is not increased.
In addition to alcohol, gallstone disease is thought to be
an important risk factor for pancreatitis (6) and, because
studies indicate that a moderate intake of alcohol could pro-
MATERIALS AND METHODS
Study population
The data used in this study came from the 4 examinations
of the Copenhagen City Heart Study, performed in 1976–
1978, 1981–1983, 1991–1994, and 2001–2003. Enrolment
and examination procedures have been described in more
Correspondence to J. Tolstrup, National Institute of Public Health, Oster Farimagsgade 5a, DK-1399 Copenhagen K, Denmark (e-mail:
[email protected]).
1
2
Kristiansen et al.
Table 1. Baseline Characteristics by Alcohol Consumption in 18,035 Danish Men and Women Participating in the Copenhagen City Heart Study,
Denmark, 1976–2007
Alcohol Intake,
drinks/week
No. of
Participants
Beverage Type,
% of total intake
Beer
Wine
Spirits
Mean Age, years
(5th percentile,
95th percentile)
Smoking
Status, %
Current
Education,
% <8 years
Mean Body
Index, kg/m2
Physical
Inactivity, %
Low
Income, %
Past
Women
0
2,953
55 (33, 73)
54
12
62
21
26
49
1–6
4,244
32
58
10
49 (25, 68)
57
16
38
21
14
27
7–13
1,554
34
54
13
48 (26, 70)
59
18
27
23
15
25
14–20
488
34
55
11
50 (31, 69)
67
18
31
24
20
23
>20
334
33
47
20
50 (33, 68)
75
11
33
24
25
26
Men
0
54 (32, 75)
62
22
55
26
25
40
1–6
1,856
948
53
33
13
50 (25, 71)
62
20
41
23
18
26
7–13
1,989
56
30
15
50 (24, 70)
65
20
39
25
16
21
14–20
1,383
62
25
13
51 (26, 70)
67
22
37
25
15
20
21–34
1,119
63
24
13
51 (30, 69)
75
17
43
22
19
19
35–48
566
66
19
15
51 (33, 69)
78
16
49
26
23
19
>48
471
75
11
14
50 (24, 67)
84
11
50
26
32
27
detail elsewhere (9, 10). Subjects with pancreatitis before
baseline or missing information on alcohol intake were excluded, which left 17,905 participants eligible for further
analysis. The study was approved by the ethical committees
for the Copenhagen area (approval reference number: KF
100.2039/91).
Alcohol intake
Beer, wine, and spirits were categorized as 0, 1–6, 7–13,
and 14 drinks/week and total alcohol intake as 0, 1–6,
7–13, 14–20, 21–34, 35–48, and >48 drinks per week. Because drinking frequency was assessed separately for beer,
wine, and spirits, we could not determine overall drinking
frequency with certainty. We performed exploratory analyses defining drinking frequency from a combination of information on beverage type and beverage type-specific
drinking frequency, categorizing drinking frequency as rare
drinkers, monthly drinkers, weekly drinkers, almost daily
drinkers, and daily drinkers.
Covariates
Smoking (never smoker, former smoker, and smoker of
1–14 g, 15–24 g, and >24 g of tobacco/day), body mass
index (<20, 20–24, and 25 kg/m2), physical activity (sedentary, light, moderate, and heavy), school education (<8,
8–11, and >11 years of education, corresponding to lower
primary school, higher primary school, and secondary
school), and income (low, middle, and high income, corresponding to <$30,000, $30,000–$80,000, and >$80,000/
year in Denmark in 1991–1994) were considered to be potential confounders. Information on gallstone disease was
obtained from the Danish Hospital Discharge Register,
which contains data on all hospital admissions in Denmark.
Endpoints
Information on acute and chronic pancreatitis was obtained from the Danish Hospital Discharge Register and
the Danish Register of Causes of Death. For acute pancreatitis, the relevant International Classification of Diseases
(ICD), Eighth Revision, codes were 577.00–577.04, 577.08,
and 577.09, and the ICD, Tenth Revision, code was K85.9.
For chronic pancreatitis, the relevant ICD, Eighth Revision,
codes were 577.19 and 577.90–577.92, and the ICD, Tenth
Revision, codes were K86.0–K86.3, K86.8, and K86.9.
Statistical analysis
Participants accrued person-time from the time of their
first participation in the Copenhagen City Heart Study until
the time of pancreatitis diagnosis, date of death, emigration,
or end of follow-up (July 9, 2007), whichever occurred first.
We had follow-up information on 100% of the study participants. Data were analyzed by means of the Cox proportional hazards regression model, with delayed entry
implemented by using SAS/STAT, version 9.1, software
(SAS Institute, Inc., Cary, North Carolina). Age (in days)
was used as the underlying time axis.
Analyses were performed by using updated information on
alcohol and covariates. Tests for linear trend were performed
by treating the median value within categories of alcohol intake as a continuous variable and adding this to the model.
Using fractional polynomials, we performed analyses to
study the shape of the risk curve to see if there was any
1.6, 5.7
<0.001
3.0
2.0, 7.2
<0.001
3.8
14
1.3, 8.3
0.004
3.3
1.8, 11
<0.001
4.4
b
a
Adjusted for age and sex.
Adjusted for age, sex, smoking, education, and body mass index.
8
1.5, 7.3
<0.001
<0.001
3.3
4.3
9
>48
1.9, 9.3
4.3
13
Ptrend
0.7, 2.2
1.4, 4.8
2.6
1.7, 5.6
3.1
16
1.1, 6.6
2.7
1.3, 7.8
3.2
8
1.8, 7.1
3.5
0.8, 2.1
35–48
2.2, 8.5
1.3
0.8, 2.4
1.4
20
0.6, 3.1
1.3
0.6, 3.3
1.4
9
0.9, 3.2
1.7
2.0
17
21–34
1.1, 3.6
0.8, 1.8
1.2
1.3
0.8, 2.2
0.8, 1.8
1.2
1.4
26
46
0.6, 2.4
0.7, 3.2
1.5
1.2
0.6, 2.2
0.7, 3.2
1.5
1.1
17
12
0.7, 2.4
0.9, 2.3
1.4
1.3
0.8, 2.6
1.4
17
14–20
1.5
36
7–13
0.9, 2.4
Referent
0.8, 1.6
1.1
1.0
Referent
0.7, 1.6
1.1
1.0
52
61
0.7, 2.3
Referent
1.0
1.2
0.6, 2.1
Referent
1.0
1.2
25
18
Referent
0.7, 1.8
1.2
1.0
Referent
0.7, 1.8
1.2
44
1–6
1.0
35
0
Hazard
Ratiob
95%
Confidence
Interval
No. of
Cases
Hazard
Ratioa
95%
Confidence
Interval
Hazard
Ratiob
95%
Confidence
Interval
No. of
Cases
Hazard
Ratioa
Total Pancreatitis
95%
Confidence
Interval
Hazard
Ratiob
95%
Confidence
Interval
Hazard
Ratioa
No. of
Cases
The mean follow-up time in this study was 20.1 years
(range, 0–31 years). At the end of follow-up, 235 participants (113 women and 122 men) had developed pancreatitis,
and there were 171 cases of acute and 97 cases of chronic
pancreatitis.
An increasing risk of pancreatitis according to alcohol
intake was observed in both men and women, although this
was statistically insignificant in women (data not shown).
We performed further analyses including men and women in
the same model (Pinteraction between sex and alcohol in
a nested log likelihood test ¼ 0.83). A high alcohol intake
was associated with an increased risk of both acute and
chronic pancreatitis (Table 2). The hazard ratio for acute
and chronic pancreatitis combined (total pancreatitis) increased by 1.13 for every additional drink/day (95% confidence interval (CI): 1.06, 1.21). In multivariate-adjusted
models, smoking was responsible for most of the effect of
adjustment. The fully adjusted risk of pancreatitis in women
compared with men was 0.9 (95% CI: 0.7, 1.2). Including
variables for personal income and physical activity in the
adjusted model had little effect on risk estimates.
Separating never drinkers from the nondrinkers was not
possible in this study, and it is hence possible that the category of nondrinkers contains participants with a previously
high intake, resulting in a falsely high incidence rate of
pancreatitis in this category. To address this issue, analyses
were repeated, separating nondrinkers from consistent nondrinkers, that is, participants who participated in at least 2
examinations and reported no alcohol intake at every examination. During follow-up, 23 cases occurred in the category of
consistent nondrinkers. The hazard ratios for total pancreatitis
Chronic Pancreatitis
Amount of alcohol intake and risk of pancreatitis
Acute Pancreatitis
Table 1 shows the characteristics of the 9,573 women and
8,332 men participating in this study, categorized by amount
of alcohol intake (assessed on the time of their first participation in the Copenhagen City Heart Study). Among both
women and men, those in the higher alcohol categories were
more likely to smoke, whereas individuals in the lowest
alcohol category tended to be older, to have fewer years of
education, and to have lower income levels.
Alcohol
Intake,
drinks/week
RESULTS
Table 2. Risk of Acute Pancreatitis, Chronic Pancreatitis, and Total Pancreatitis According to Updated Consumption of Alcohol Intake, Copenhagen, Denmark, 1976–2007
indication of a threshold effect (11). The model with the best
fit was a model including alcohol as linear and square terms.
To examine the possibility that latent baseline symptoms
of pancreatitis might have reduced the amount of alcohol
consumed, thereby biasing results, we carried out analyses
in which the first 2 or 4 years of observation time were
excluded, and information was updated with a delay of 2
or 4 years, respectively.
The test for interaction between alcohol intake and smoking was performed by a nested log likelihood test, comparing a model containing the variables as single terms with
a model also including the interaction terms. For this purpose, alcohol was categorized as <7, 7–20, and >20 drinks/
week, and smoking was categorized as never smokers, former
smokers, and current smokers.
95%
Confidence
Interval
Pancreatitis and Drinking Habits
3
4
Kristiansen et al.
using consistent nondrinkers as the reference category were
similar to results from the main analysis (data not shown).
Excluding the first 2 or 4 years of observation time revealed results similar to those from the main analyses (data
not shown).
Modeling the association between alcohol and risk of
total pancreatitis for men and women separately and together, respectively, using fractional polynomials did not
indicate that there was a threshold in the risk of pancreatitis
according to alcohol (data not shown). The curves flattened
out at high intakes, but the test for linearity did not provide
evidence for departure from linearity (P ¼ 0.14 for women
and men combined).
Compared with that among never smokers, the hazard
ratio for pancreatitis was 1.6 (95% CI: 1.0, 2.5) among past
smokers and 1.5 (95% CI: 0.9, 2.3), 2.3 (95% CI: 1.5, 3.6),
and 3.1 (95% CI: 1.8, 5.3) among current smokers of 1–14,
15–24, and >24 g of tobacco/day, respectively. We found no
evidence of interaction between alcohol intake and smoking
(P ¼ 0.57).
Table 3. Risk of Total Pancreatitis According to Updated
Consumption of Individual Alcoholic Beverages, Copenhagen,
Denmark, 1976–2007
Total Pancreatitis
No. of
Cases
95%
Confidence
Interval
Hazard
Ratiob
95%
Confidence
Interval
Beer, drinks/
week
0
87
1.0
Referent
1.0
Referent
1–6
74
1.3
0.9, 1.8
1.3
0.9, 1.8
7–13
25
1.3
0.8, 2.2
1.2
0.7, 2.0
14
49
2.5
1.6, 3.8
2.0
Ptrend
1.3, 3.1
0.003
<0.001
Wine, drinks/
week
0
Gallstone disease, alcohol, and pancreatitis
In analyses including gallstone disease as a time-dependent
variable, the hazard ratios of pancreatitis according to alcohol
were slightly increased: For example, the risk associated with
drinking >48 drinks/week increased from 3.0 (95% CI: 1.6,
5.7) to 3.6 (95% CI: 1.9, 6.9) when including gallstones in
the model. In addition, the hazard ratio for continuous alcohol intake increased from 1.13 (95% CI: 1.06, 1.21) to 1.15
(95% CI: 1.08, 1.23) per drink/day. The hazard ratio of
pancreatitis according to gallstone disease was 11 (95%
CI: 7.5, 15).
Hazard
Ratioa
117
1.0
Referent
1.0
Referent
1–6
99
1.0
0.7, 1.4
1.1
0.8, 1.5
7–13
11
0.7
0.4, 1.3
0.8
0.4, 1.5
14
8
0.8
0.4, 1.7
0.9
0.4, 1.8
0.21
Ptrend
0.49
Spirits, drinks/
week
0
145
1.0
Referent
1.0
Referent
1–6
66
0.8
0.6, 1.1
0.7
0.5, 1.0
7–13
16
1.1
0.6, 1.8
1.0
0.6, 1.7
14
8
1.0
0.5, 2.0
0.9
Ptrend
0.59
0.4, 1.8
0.42
a
Adjusted for age, sex, and other beverage types.
Adjusted for age, sex, smoking, education, body mass index, and
other beverage types.
b
Analyses of beverage type and drinking frequency
We explored the relation between type of alcoholic beverage and risk of pancreatitis (Table 3). The adjusted hazard
ratio for drinking more than 14 beers/week was 2.0 (95%
CI: 1.3, 3.1). No association was observed regarding wine
and spirits. However, there were only a few participants in
the highest categories of both wine and spirits.
We found no evidence that drinking frequency was associated with risk of pancreatitis. In a comparison with never
drinkers, the adjusted hazard ratio was 0.8 (95% CI: 0.5,
1.3), 1.0 (95% CI: 0.7, 1.5), 1.1 (95% CI: 0.6, 1.8), and
1.3 (95% CI: 0.9, 1.9) for monthly, weekly, almost daily,
and daily alcohol drinking. With further adjustment for the
amount of alcohol intake (as a continuous variable), the
hazard ratio for daily drinking attenuated to 1.0 (95% CI:
0.6, 1.5). In this model, amount of alcohol intake remained
significantly associated with increased risk of pancreatitis
(P < 0.001), indicating that amount of alcohol is more
important than drinking frequency for the risk associated
with alcohol drinking.
DISCUSSION
In this prospective cohort study, we found that a high
alcohol intake was associated with increased risk of pancre-
atitis. Drinking frequency appeared to be unassociated with
the risk of pancreatitis. Further, our results indicate that
gallstones slightly temper the association between alcohol
and pancreatitis.
Limitations include a lack of information on potential
confounders, such as diet and coffee (4, 12). Furthermore,
the diagnosis of pancreatitis has not been validated. In the
Danish Hospital Discharge Register, only hospital admissions and not contacts with general practitioners are registered; hence, it is likely that we do not have information on
all cases occurring during the study period. Moreover, misclassification between the diagnoses of acute and chronic
pancreatitis has most likely occurred, as the symptoms and
diagnostic criteria are overlapping and the 2 diseases can
coexist (13). Such misclassification would result in similar
observed associations between alcohol and risk of acute and
chronic pancreatitis. Therefore, for these reasons and because of the limited number of cases, we recommend that
readers do not place too much emphasis on this result. However, results for the joint outcome of pancreatitis are valid.
We found an increased risk of pancreatitis when drinking
>14 drinks of beer/week, whereas no association was observed for wine or spirits. For wine, this result could hypothetically be due to a protective effect of the antioxidants
Pancreatitis and Drinking Habits
found in wine. However, the validity of these results is severely limited by the small number of cases in the high categories of wine and spirits and by the fact that wine drinking
is associated with a generally healthier lifestyle (14).
Results on drinking frequency are limited by the fact that
our measure of this drinking variable may be too crude to
pick up small effects. Unfortunately, information on binge
drinking (i.e., drinking a minimum number of drinks per
occasion) was not available, and we cannot comment on
this aspect with the present data.
Our data did not suggest a clear threshold effect of alcohol
intake on the risk of pancreatitis. However, this might be due
to misclassification. In the case of underreporting, the risk in
the reference category would increase, and this could lead to
a true threshold amount being either lowered or blurred.
Strengths include the large study size, the prospective
design, and the fact that we had complete follow-up information on all 17,905 participants. In addition, to the best of
our knowledge, this is the first study to investigate the association between alcohol intake and pancreatitis prospectively, even though it is considered well known that alcohol
is a strong risk factor for pancreatitis.
Our findings are consistent with previous results from
case-control studies (1–4), although the literature on alcohol
and pancreatitis is sparse, and only one previous case-control
study has included women (2). In that study, the risk of
pancreatitis was not associated with alcohol among women;
however, results were limited by the few female cases.
Different mechanisms have been suggested to explain the
toxic effect of alcohol. The pancreas can degrade alcohol by
both oxidative and nonoxidative metabolism— mechanisms
involving the syntheses of acetaldehyde and fatty-acid ethanol esters, respectively (15). The latter (i.e., fatty-acid ethanol esters) have been demonstrated to cause pancreatic
edema, intracellular trypsin activation, and the induction
of proinflammatory transcription factors in animal studies
(16–18). In addition, the toxic effect of alcohol may be due
to the induction of oxidative stress, that is, the imbalance
between formation and neutralization of reactive oxygen
species (19–21). On the other hand, there is some evidence
that a moderate alcohol intake protects against gallstone
disease (7, 8), which would lower the risk of pancreatitis.
This is in accordance with our finding that the risk of pancreatitis according to alcohol was increased when including
gallstones in the model.
In summary, we found that a high alcohol intake was
associated with an increased risk of pancreatitis in both
men and women, whereas drinking frequency did not seem
to be a risk factor for pancreatitis when the amount of alcohol intake was taken into consideration.
ACKNOWLEDGMENTS
Author affiliations: Centre for Alcohol Research, National
Institute of Public Health, University of Southern Denmark,
Copenhagen, Denmark (Louise Kristiansen, Morten Grønbæk,
Janne Schurmann Tolstrup); and Alcohol Unit, Hvidovre Hospital, Copenhagen, Denmark (Ulrik Becker).
5
The study is supported by grants from the Danish
National Board of Health and the Danish Medical Research
Council.
Conflict of interest: none declared.
REFERENCES
1. Durbec JP, Sarles H. Multicenter survey of the etiology of
pancreatic diseases. Relationship between the relative risk of
developing chronic pancreatitis and alcohol, protein, and lipid
consumption. Digestion. 1978;18(5–6):337–350.
2. Yen S, Hsieh CC, MacMahon B. Consumption of alcohol and
tobacco and other risk factors for pancreatitis. Am J Epidemiol.
1982;116(3):407–414.
3. Talamini G, Bassi C, Falconi M, et al. Alcohol and smoking as
risk factors in chronic pancreatitis and pancreatic cancer. Dig
Dis Sci. 1999;44(7):1303–1311.
4. Lin Y, Tamakoshi A, Hayakawa T, et al. Associations of alcohol drinking and nutrient intake with chronic pancreatitis:
findings from a case-control study in Japan. Am J Gastroenterol. 2001;96(9):2622–2627.
5. Ramstedt M. Alcohol and pancreatitis mortality at the population level: experiences from 14 Western countries. Addiction.
2004;99(10):1255–1261.
6. Steinberg W, Tenner S. Acute pancreatitis. N Engl J Med.
1994;330(17):1198–1210.
7. Cuevas A, Miquel JF, Reyes MS, et al. Diet as a risk factor for
cholesterol gallstone disease. J Am Coll Nutr. 2004;23(3):
187–196.
8. Kono S, Eguchi H, Honjo S, et al. Cigarette smoking, alcohol
use, and gallstone risk in Japanese men. Digestion. 2002;
65(3):177–183.
9. The Copenhagen City Heart Study. Osterbroundersøgelsen.
A book of tables with data from the first examination (1976–78)
and a five year follow-up (1981–83). The Copenhagen City
Heart Study Group. Scand J Soc Med. 1989;170(suppl 41):
1–160.
10. Schnohr P, Jensen G, Lange P, et al. The Copenhagen City
Heart Study. Østerbroundersøgelsen: tables with data from the
third examination 1991–1994. Eur Heart J. 2001;3(suppl H):
1–83.
11. Royston P, Ambler G, Sauerbrei W. The use of fractional
polynomials to model continuous risk variables in epidemiology. Int J Epidemiol. 1999;28(5):964–974.
12. Morton C, Klatsky AL, Udaltsova N. Smoking, coffee, and
pancreatitis. Am J Gastroenterol. 2004;99(4):731–738.
13. Kloppel G, Maillet B. Development of chronic pancreatitis
from acute pancreatitis: a pathogenetic concept. (In German).
Zentralbl Chir. 1995;120(4):274–277.
14. Tjonneland AM, Gronbaek MN, Stripp C, et al. The connection between food and alcohol intake habits among 48.763
Danish men and women. A cross-sectional study in the project
‘‘Food, cancer and health.’’ (In Danish). Ugeskr Laeger. 1999;
161(50):6923–6927.
15. Lange LG. Nonoxidative ethanol metabolism: formation of
fatty acid ethyl esters by cholesterol esterase. Proc Natl Acad
Sci U S A. 1982;79(13):3954–3957.
16. Gukovskaya AS, Mouria M, Gukovsky I, et al. Ethanol metabolism and transcription factor activation in pancreatic acinar
cells in rats. Gastroenterology. 2002;122(1):106–118.
17. Werner J, Saghir M, Warshaw AL, et al. Alcoholic pancreatitis
in rats: injury from nonoxidative metabolites of ethanol. Am J
Physiol Gastrointest Liver Physiol. 2002;283(1):G65–G73.
6
Kristiansen et al.
18. Haber PS, Wilson JS, Apte MV, et al. Fatty acid ethyl esters
increase rat pancreatic lysosomal fragility. J Lab Clin Med.
1993;121(6):759–764.
19. Nordmann R, Ribiere C, Rouach H. Implication of free radical
mechanisms in ethanol-induced cellular injury. Free Radic
Biol Med. 1992;12(3):219–240.
20. Cederbaum AI. Introduction-serial review: alcohol, oxidative
stress and cell injury. Free Radic Biol Med. 2001;31(12):
1524–1526.
21. Chowdhury P, Gupta P. Pathophysiology of alcoholic pancreatitis: an overview. World J Gastroenterol. 2006;12(46):
7421–7427.
ORIGINAL INVESTIGATION
Smoking and Risk of Acute and Chronic
Pancreatitis Among Women and Men
A Population-Based Cohort Study
Janne Schurmann Tolstrup, MSc, PhD; Louise Kristiansen, BSc;
Ulrik Becker, MD, DrMedSci; Morten Grønbæk, MD, PhD, DrMedSci
Background: Alcohol and gallstone disease are the
most established risk factors for pancreatitis. Smoking
is rarely considered to be a cause despite the fact that a
few studies have indicated the opposite. We aimed to
assess the independent effects of smoking on the risk of
pancreatitis.
Methods: We used data from an observational, population-based cohort study conducted in Denmark. Participants were 9573 women and 8332 men who were
followed up for a mean of 20.2 years. Participants underwent a physical examination and completed selfadministered questionnaires about lifestyle habits. Information on incident cases of acute and chronic
pancreatitis were obtained by record linkage with the
Danish national registries.
Results: A total of 235 cases of pancreatitis occurred during follow-up. A dose-response association between smok-
ing and risk of acute and chronic pancreatitis was observed in both men and women. For example, the hazard
ratio of developing pancreatitis was 2.6 (95% confidence interval [CI], 1.5-4.7) among women and 2.6 (95%
CI, 1.1-6.2) among men who smoked 15 to 24 grams of
tobacco per day. Alcohol intake was associated with an
increased risk of pancreatitis (hazard ratio, 1.09; 95% CI,
1.04-1.14 for each additional drink per day). The risk of
pancreatitis associated with smoking, however, was independent of alcohol and gallstone disease. Approximately 46% of cases of pancreatitis were attributable to
smoking in this cohort.
Conclusion: In this population of Danish men and
women, smoking was independently associated with increased risk of pancreatitis.
Arch Intern Med. 2009;169(6):603-609
T
Author Affiliations: Center for
Alcohol Research, National
Institute of Public Health,
University of Southern
Denmark (Drs Tolstrup and
Grønbæk and Ms Kristiansen),
and The Alcohol Unit, Hvidovre
Hospital (Dr Becker),
Copenhagen, Denmark.
HE INCIDENCE OF PANCREatitis has increased in recent decades.1-8 Pancreatitis can be divided into acute
and chronic pancreatitis;
however, disagreement exists on whether
2 distinct conditions really exist or whether
acute pancreatitis leads to chronic pancreatitis.9,10 The most common symptom
of both is severe abdominal pain. The mortality rate of acute pancreatitis is especially high and has not decreased since the
1970s, which is most likely because treatment has not improved and is mainly directed toward pain control.
Gallstone disease and excessive alcohol
use are described as being the most common causes of acute and chronic pancreatitis, respectively. In the medical literature, smoking is generally not considered
to be an important risk factor for pancreatitis, even though some case-control studies conducted in men, as well as a recent cohort study, suggest the opposite.11-18 Also,
evidence from experimental studies19-22 sug-
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 6), MAR 23, 2009
603
gests that smoking is associated with pancreas damage. In most populations, smoking is strongly associated with drinking
alcohol; hence, an independent effect of
smoking can be difficult to assess, especially from a case-control design because individuals with moderate to heavy alcohol
intake are dominant.
Smoking may also be associated with
risk of gallstone disease.23,24 If so, observed associations between smoking and
pancreatitis could be explained by an increased risk of gallstones in individuals
who smoke, which in turn renders these
individuals at high risk of pancreatitis.
In this study, we examined the association between smoking and risk of acute
and chronic pancreatitis in a large prospective cohort consisting of men and
women from the general population. Our
objective was to determine if smoking is
associated with an increased risk of acute
and chronic pancreatitis independently of
alcohol use and gallstone disease.
WWW.ARCHINTERNMED.COM
Downloaded from www.archinternmed.com at University Library of Southern Denmark, on April 28, 2010
©2009 American Medical Association. All rights reserved.
INFORMATION ON PANCREATITIS
AND GALLSTONE DISEASE
METHODS
STUDY POPULATION
Data used in this study came from the first 3 examinations of
the Copenhagen City Heart Study (CCHS), performed in
1976-1978, 1981-1983, and 1991-1994. The CCHS is a prospective cohort study. The participants were randomly chosen
from the general population of Copenhagen and included
14 233 men and women 20 to 95 years old in 1976-1978 (response rate, 74%). In 1981-1983, all previously invited individuals plus 500 new individuals aged 20 to 24 years were
invited to participate, resulting in 12 698 participants (response rate, 70%). In 1991-1994, all previously invited individuals plus 3000 new individuals aged 20 to 49 years were
invited to participate, and 10 135 of these participated (response rate, 61%). In total, 18 035 individuals participated in
1 or more examinations of the CCHS.
Before visiting the study clinic, participants completed a selfadministered questionnaire (including questions on alcohol intake, smoking, physical activity, education level, and income). At the clinic visit, physical examinations were performed
(including measurement of height, weight, forced expiratory
volume in 1 second [FEV1], and carbon dioxide in expired air)
and questionnaires were checked for missing information.
SMOKING
At each examination, participants were asked whether they
smoked or had been smoking previously and, if the response
was affirmative, about duration of smoking (in years). Current smokers were further asked about the usual amount of tobacco in categories of daily cigarettes, cheroots, cigars, and pipes.
Assuming 1 cigarette to be equivalent to 1 g of tobacco, 1 cheroot or 1 pipe to 3 g of tobacco, and 1 cigar to 5 g of tobacco,
participants were categorized in 5 groups (never-smokers, exsmokers, and smokers of 1-14, 15-24, and ⬎24 g/d of tobacco). Pack-years of smoking were calculated as (years of
smoking⫻daily grams of tobacco)/20. For current smokers, duration of smoking was updated every 5 years until the participant reported having quit or was censored. For ex-smokers, we
had information only on duration of smoking; therefore, packyears could not be calculated.
COVARIABLES
For the purpose of this study, sex, smoking, education level,
income, body mass index (BMI), and physical activity were considered potential confounders. Education level was categorized as less than 8 years, 8 to 11 years, and more than 11 years
of education corresponding to lower primary school, higher primary school, and secondary school, respectively. Income was
categorized as low, middle, and high income. The BMI was calculated as weight in kilograms divided by height in meters
squared (measured at the clinic visit) and thereafter categorized as less than 20, 20 through 24.9, and 25 or greater. Physical activity was categorized as sedentary (light physical activity for ⬍2 hours per week), light (light physical activity for 2-4
hours per week), moderate (light physical activity for ⬎4 hours
per week or strenuous activity for 2-4 hours per week), and
heavy (strenuous physical activity at least 4 hours per week).
Concerning alcohol intake, participants were asked how often
they drank beer, wine, and spirits in categories of “never/
hardly ever,” “monthly,” “weekly,” and “daily” and how many
drinks they drank of each type of beverage per week. The 3 types
of alcoholic beverage were added up to a measure of total alcohol intake.
Information on pancreatitis was obtained from the Danish Hospital Discharge Register25 and the Danish Registers of Causes of
Death,26 which contain data on all hospital admissions and causes
of death in Denmark, respectively. Information on incident cases
of pancreatitis among the study participants in these registries was
identified through linkage by the unique identification number,
which is allocated to every Danish inhabitant by the Central Population registry. The exact diagnosis for pancreatitis has not been
validated, but the validity of the Danish Hospital Discharge Register is generally considered to be high.27 For acute pancreatitis,
the relevant International Classification of Diseases, Eighth Revision (ICD-8)28 codes were 577.00, 577.01, 577.02, 577.03, 577.04,
577.08, and 577.09, along with International Statistical Classification of Diseases, 10th Revision (ICD-10)29 code K85.9. For chronic
pancreatitis, the relevant ICD-8 codes were 577.19, 577.90, 577.91,
and 577.92, and the ICD-10 codes were K86.0, K86.1, K86.2,
K86.3, and K86.9. Lastly, information on gallstone disease was
also obtained from the Danish Hospital Discharge Register (ICD-8
and ICD-10 code K80).
STATISTICAL ANALYSIS
Participants accrued person-time from the time of their first
participation in the CCHS until the time of their first admission to a hospital because of pancreatitis, date of death, emigration, or end of follow-up ( July 9, 2007), whichever occurred first. We had follow-up information on 100% of the study
participants. Analyses were performed for acute and chronic
pancreatitis separately and combined (total pancreatitis). For
the analysis of acute pancreatitis, participants who had a previous diagnosis of chronic pancreatitis were censored at the time
of the chronic pancreatitis (n=11). Participants with missing
information on smoking, alcohol intake, BMI, or school education level (n=125) or with a diagnosis of pancreatitis before
baseline (n=3) were excluded from the study, which left 17 905
eligible for analysis. Data were analyzed by means of the Cox
proportional hazards regression model, with delayed entry implemented (using the SAS/STAT program software, version 9.1;
SAS Institute Inc, Cary, North Carolina). To ensure maximal
adjustment for confounding by age, age (in days) was used as
the underlying time axis. The Cox proportional hazards assumption was examined graphically and statistically by introducing interaction terms between time and alcohol consumption in the model; no violations against the assumptions were
detected.
Primary analyses were performed using updated measures
of alcohol consumption and other covariables, in which we prospectively assessed the risk of pancreatitis in betweenexamination increments, based on information on alcohol consumption and other covariables derived from the preceding
questionnaire. Also, analyses were performed including only
information on smoking and covariables from the first examination in which every participant had joined.
We estimated the population-attributable risk related to smoking categories (never-smoker, ex-smoker, and current smokers
of 1-14, 15-24, and ⱖ25 g/d of tobacco) as ⌺Pei (RRi−1)/1⫹⌺Pei
(RRi兺−1), where Pei is the prevalence of the i’th smoking category and RRi is the relative risk associated with this category.30
Interaction between alcohol intake and smoking was determined with the use of a nested log-likelihood test, comparing
a model containing the variables as single terms with a model
also including the interaction terms. For this particular purpose, alcohol was categorized into 3 groups (⬍7, 7-20, and ⬎20
drinks per week) and smoking was categorized into 2 groups
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 6), MAR 23, 2009
604
WWW.ARCHINTERNMED.COM
Downloaded from www.archinternmed.com at University Library of Southern Denmark, on April 28, 2010
©2009 American Medical Association. All rights reserved.
Table 1. Baseline Characteristics of 9573 Women and 8332 Men Participating in the Copenhagen City Heart Study
by Smoking Status a
Smoking Status
Characteristic
All Patients
No. (%)
Mean age (range), y
Duration of smoking, y
FEV1, % predicted
Carbon dioxide in expired air, ppm
Alcohol intake, drinks/wk b
BMI
Physical inactivity, %
Low income, %
Education level ⬍8 y, %
9573
52 (26-70)
25 (5-43)
80 (63-94)
3 (1-25)
3 (0-16)
24 (19-33)
19
33
43
No. (%)
Mean age (range), y
Duration of smoking, y
FEV1, % predicted
Carbon dioxide in expired air, ppm
Alcohol intake, drinks/wk b
BMI
Physical inactivity, %
Low income, %
Education level ⬍8 y, %
8332
52 (26-70)
30 (7-53)
79 (59-94)
4 (1-28)
11 (0-49)
25 (20-32)
19
24
42
Never
Former
Women
2665 (28)
1430 (15)
53 (25-72)
53 (27-72)
11 (2-40)
82 (68-95)
81 (64-93)
2 (1-4)
2 (1-4)
2 (0-13)
3 (0-16)
24 (19-34)
24 (19-34)
17
14
36
30
41
37
1-14 g/d
of Tobacco
15-24 g/d
of Tobacco
ⱖ25 g/d
of Tobacco
2843 (30)
53 (27-69)
25 (6-45)
80 (61-93)
8 (2-22)
3 (0-16)
23 (19-32)
18
34
47
2255 (24)
48 (27-66)
25 (10-43)
79 (61-93)
15 (3-32)
3 (0-21)
23 (19-32)
23
30
45
380 (4)
51 (31-67)
29 (10-44)
79 (59-93)
19 (2-41)
5 (0-38)
24 (19-33)
28
29
38
1950 (23)
54 (26-73)
34 (8-57)
79 (57-95)
7 (2-23)
10 (0-42)
25 (20-32)
16
28
45
2542 (31)
51 (28-69)
33 (10-51)
78 (57-94)
14 (3-33)
12 (0-51)
25 (20-32)
21
23
46
1163 (14)
50 (33-68)
34 (15-52)
78 (58-94)
17 (5-41)
18 (0-77)
26 (21-33)
27
17
45
Men
1080 (13)
41 (23-69)
82 (69-96)
2 (1-5)
8 (0-36)
25 (20-32)
13
24
27
1597 (19)
56 (30-73)
22 (2-50)
80 (60-94)
2 (1-5)
11 (0-42)
26 (21-33)
17
25
43
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); FEV1, forced expiratory volume in 1 second.
a Data are medians (5th-95th percentiles) unless otherwise indicated.
b One drink corresponds to 12 g of pure alcohol.
(never-smokers and ex-smokers and current smokers) to increase the statistical power of the test.
RESULTS
BASELINE CHARACTERISTICS
Table 1 gives the characteristics of 17 905 men and women
categorized by smoking status as reported at the time of
their first participation in the CCHS (1976-1978, 19811983, or 1991-1994). Overall, 58% of the women and 68%
of the men were current smokers, 15% of the women and
19% of the men were ex-smokers, and 28% of the women
and 13% of the men had never smoked. As expected, lung
function as measured by FEV1 (in percentage of the expected value) was highest in participants who never smoked
and was lowest in participants who were current smokers. Also, carbon dioxide in expired air was similar in neversmokers and ex-smokers and increased with increasing
amounts of tobacco used per day. On average, current smokers drank more alcohol and were less physically active compared with never-smokers.
SMOKING STATUS, PACK-YEARS,
AND RISK OF PANCREATITIS
The mean follow-up in this study was 20.2 years (range,
0-28.3 years). At the end of follow-up, a total of 235 participants (113 women and 122 men) had developed pancreatitis, and there were 160 cases of acute and 97 cases
of chronic pancreatitis (the number of cases of acute and
chronic pancreatitis does not equal the total number of
cases because some people had both acute and chronic
pancreatitis and they were only counted once in the total
number of cases).
We performed analyses by modeling the risk of acute
and chronic pancreatitis combined (total pancreatitis) according to categories of smoking status (Table 2). Similar risk estimates were observed in women and men: for
example, the hazard ratio of developing pancreatitis was
2.6 (95% confidence interval [CI], 1.5-4.7) among women
and 2.6 (95% CI, 1.1-6.2) among men who smoked 15
to 24 grams per day of tobacco. Combining women and
men (P value for interaction between sex and smoking
status in nested log likelihood test was .70) in an analysis adjusted for age and sex and in an analysis further adjusted for alcohol, education level, and BMI, current smokers had a higher risk of both acute and chronic pancreatitis
compared with never-smokers, and risk estimates were
similar for the 2 outcomes (Table 3). For ex-smokers,
however, the hazard ratio for acute pancreatitis was 2.3
(95% CI, 1.3-4.1), whereas the hazard ratio for chronic
pancreatitis was 0.9 (95% CI, 0.4-2.0). For total pancreatitis, adjusted hazard ratios were 1.7 (95% CI, 1.0-2.7),
1.5 (95% CI, 0.9-2.5), 2.5 (95% CI, 1.5-3.9), and 3.3 (95%
CI, 1.9-5.8) among ex-smokers and current smokers of
1 to 14, 15 to 24, and 25 g/d or more of tobacco, respectively. In the multivariable-adjusted models, alcohol was
responsible for most of the effect of adjustment. The adjusted hazard ratio for amount of alcohol intake was 1.09
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 6), MAR 23, 2009
605
WWW.ARCHINTERNMED.COM
Downloaded from www.archinternmed.com at University Library of Southern Denmark, on April 28, 2010
©2009 American Medical Association. All rights reserved.
Table 2. Risk of Total Pancreatitis by Updated Information on Smoking Status in Men and Women a
Smoking Status
Women
No. of cases
Crude hazard ratio (95% CI)
Adjusted hazard ratio (95% CI) b
Men
No. of cases
Crude hazard ratio (95% CI)
Adjusted hazard ratio (95% CI) b
Never
Former
1-14 g/d of
Tobacco
15-24 g/d of
Tobacco
ⱖ25 g/d of
Tobacco
20
1.0
1.0
23
1.4 (0.8-2.5)
1.4 (0.8-2.6)
30
1.7 (1.0-3.0)
1.7 (0.9-3.0)
35
2.8 (1.6-5.0)
2.6 (1.5-4.7)
5
2.5 (0.9-6.6)
2.3 (0.8-6.2)
6
1.0
1.0
31
2.2 (0.9-5.4)
2.1 (0.9-5.2)
18
1.6 (0.6-4.1)
1.5 (0.6-3.9)
37
2.9 (1.2-6.7)
2.6 (1.1-6.2)
30
4.9 (2.0-12.0)
4.1 (1.7-9.9)
Abbreviation: CI, confidence interval.
a Results from Cox proportional hazards regression analysis with age as the underlying time scale.
b Adjusted for education level (⬍8 years, 8-11 years, ⬎11 years), body mass index (calculated as weight in kilograms divided by height in meters squared;
⬍20, 20-24.9, ⱖ25), and alcohol intake (continuous).
Table 3. Risk of Acute, Chronic, and Total Pancreatitis by Updated Information on Smoking Status in Men and Women Combined a
Smoking Status
Acute pancreatitis
No. of cases
Crude hazard ratio (95% CI)
Adjusted hazard ratio (95% CI) b
Chronic pancreatitis
No. of cases
Crude hazard ratio (95% CI)
Adjusted hazard ratio (95% CI) b
Total pancreatitis
No. of cases
Crude hazard ratio (95% CI)
Adjusted hazard ratio (95% CI) b
Never
Former
1-14 g/d of
Tobacco
15-24 g/d of
Tobacco
ⱖ25 g/d of
Tobacco
16
1.0
1.0
45
2.4 (1.3-4.2)
2.3 (1.3-4.1)
34
1.9 (1.1-3.5)
2.0 (1.1-3.6)
44
2.8 (1.6-5.1)
2.8 (1.5-5.0)
21
4.3 (2.2-8.5)
3.8 (1.9-7.5)
11
1.0
1.0
13
0.9 (0.4-2.0)
0.9 (0.4-2.0)
18
1.3 (0.6-2.7)
1.1 (0.5-2.3)
34
2.4 (1.2-4.9)
2.0 (1.0-4.1)
21
4.2 (1.9-9.0)
3.3 (1.5-7.3)
26
1.0
1.0
54
1.7 (1.0-2.7)
1.7 (1.0-2.7)
48
1.6 (1.0-2.6)
1.5 (0.9-2.5)
72
2.6 (1.7-4.1)
2.5 (1.5-3.9)
35
3.9 (2.3-6.7)
3.3 (1.9-5.8)
Abbreviation: CI, confidence interval.
a Results from Cox proportional hazards regression analysis with age as the underlying time scale.
b Adjusted for sex, education level (⬍8 years, 8-11 years, ⬎11 years), body mass index (calculated as weight in kilograms divided by height in meters squared;
⬍20, 20-24.9, ⱖ25), and alcohol intake (continuous).
(96% CI, 1.04-1.14) for each additional drink per day.
The inclusion of variables for personal income and physical activity to the fully adjusted model had negligible effect
on the size and precision of the risk estimates. The fully
adjusted risk of pancreatitis in women compared with
men was 0.9 (95% CI, 0.6-1.1). Repeating the analyses
without updating information on smoking and other variables did not change our results (data not shown).
To explore if the relatively high risk among exsmokers was due to sick-quitters (ie, participants who have
given up smoking because of early symptoms of pancreatitis), we performed additional analyses omitting the first
2 variables for follow-up and updating smoking variables
and covariables with a delay of 2 years. However, this only
affected risk estimates slightly. Compared with the hazard ratio estimated for the entire follow-up period (1.7; 95%
CI, 1.0-2.7), the hazard ratio was 1.6 (95% CI, 1.0-2.6) when
omitting the first 2 years of follow-up.
Hazard ratios of acute, chronic, and total pancreatitis
according to pack-years of smoking for current smokers
and smoking duration for ex-smokers are given in
Table 4. Dose-response associations were observed and
again the associations for risk of acute and chronic pancreatitis were similar. For example, hazard ratios of acute
and chronic pancreatitis were 3.2 (95% CI, 1.6-6.2) and
3.1 (95% CI, 1.3-7.1) among participants who had smoked
45 to 59 pack-years compared with never-smokers. Assuming that the multivariable-adjusted hazard ratios represent biologically causal associations between smoking status and risk of pancreatitis, we estimated that
approximately 46% of cases of pancreatitis were attributable to smoking in this cohort.
We repeated the analyses stratifying by amount of alcohol intake: participants who in each examination reported drinking maximally 2 drinks per day for women
and 3 drinks per day for men were categorized as consistent light to moderate drinkers, and participants who
in at least 1 examination reported drinking above these
limits were categorized as heavy drinkers. Most of the
cases occurred among the consistent light to moderate
drinkers (162 of 235 cases totally). Among these, hazard ratios were 1.6 (95% CI, 0.9-2.6), 1.5 (95% CI, 0.9-
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 6), MAR 23, 2009
606
WWW.ARCHINTERNMED.COM
Downloaded from www.archinternmed.com at University Library of Southern Denmark, on April 28, 2010
©2009 American Medical Association. All rights reserved.
Table 4. Risk of Acute, Chronic, and Total Pancreatitis by Updated Information on Pack-Years
of Smoking (Smoking Duration Among Ex-smokers) in Men and Women Combined a
Adjusted Hazard Ratio (95% CI) b
Never smokers
Current smokers, pack-years
⬍15
15-29
30-44
45-59
ⱖ60
Ex-smokers, No. of years smoked
⬍30
ⱖ30
Total No. of
Cases
Acute
Pancreatitis
Chronic
Pancreatitis
Total
Pancreatitis
26
1.0
1.0
1.0
21
38
43
27
23
1.3 (0.7-2.7)
2.2 (1.2-3.8)
2.8 (1.6-5.1)
3.2 (1.6-6.2)
2.9 (1.4-6.3)
1.1 (0.4-2.6)
1.2 (0.5-2.6)
2.3 (1.1-4.8)
3.1 (1.3-7.1)
4.1 (1.8-9.7)
1.4 (0.8-2.4)
1.7 (1.0-2.8)
2.5 (1.5-4.2)
3.2 (1.8-5.7)
3.7 (2.0-6.8)
23
27
2.0 (1.1-3.7)
2.7 (1.5-5.1)
0.8 (0.3-2.3)
1.1 (0.4-3.2)
1.5 (0.9-2.7)
2.2 (1.3-3.9)
Abbreviation: CI, confidence interval.
a Results from Cox proportional hazards regression analysis with age as the underlying time scale.
b Adjusted for sex, education level (⬍8 years, 8-11 years, ⬎11 years), body mass index (calculated as weight in kilograms divided by height in meters squared;
⬍20, 20-24.9, ⱖ25), and alcohol intake (continuous). Seven cases were excluded because of missing data on duration of smoking.
GALLSTONES AS A POTENTIAL
MEDIATOR OF THE ASSOCIATION BETWEEN
SMOKING AND PANCREATITIS
We performed analyses of gallstones as a potential mediator of the association between smoking and risk of pancreatitis. In total, 562 women and 251 men had gallstones before the end of follow-up. As expected, gallstones
were associated with subsequent risk of pancreatitis (adjusted hazard ratio for total pancreatitis, 11; 95% CI, 7.715). However, adjusting for gallstones had little influence on our results, indicating that gallstone disease does
not mediate the association between smoking and pancreatitis. For example, the adjusted hazard ratio of total
pancreatitis among participants who smoked 15 to 24 g/d
of tobacco was 2.5 (95% CI, 1.5-3.9) (Table 3) and 2.3
(95% CI, 1.5-3.7) after further adjustment for gallstones.
In separate analyses of acute and chronic pancreatitis, including gallstones in the model did not affect our results.
To see if smoking is associated with increased risk of
gallstone-associated pancreatitis, we performed exploratory analyses restricted to participants who had gallstones before or concomitant with pancreatitis diagnosis. This accounted for 831 participants of whom 65
developed pancreatitis during follow-up. In this subgroup, the adjusted hazard ratio of total pancreatitis was
1.4 (95% CI, 0.7-2.7) among ex-smokers and smokers
of 1 to 14 g/d of tobacco combined and was 0.9 (95% CI,
7
6
Never smokers
Former smokers
Current smokers
5
Hazard Ratio
2.4), 2.3 (95% CI, 1.4-3.8), and 2.9 (95% CI, 1.5-5.8) in
ex-smokers and current smokers of 1 to 14, 15 to 24, and
25 g/d or more of tobacco, respectively. Corresponding hazard ratios among the heavy drinkers were 1.7 (95% CI, 0.55.2), 1.4 (95% CI, 0.5-4.6), 2.4 (95% CI, 0.8-6.8), and 3.3
(95% CI, 1.1-10).
Incidence rates for total pancreatitis by smoking and
alcohol are shown in the Figure. Within each category of
alcohol intake, the incidence rate was higher among exsmokersandcurrentsmokerscomparedwithnever-smokers.
We found no evidence of interaction between alcohol intake and smoking on the risk of pancreatitis (P=.70).
4
3
2
1
0
<7 Drinks/wk
7-20 Drinks/wk
>20 Drinks/wk
Alcohol Intake
Figure. Hazard ratios of pancreatitis by smoking status and weekly alcohol
intake in men and women combined. Hazard ratios are adjusted for age, sex,
body mass index (calculated as weight in kilograms divided by height in
meters squared), and education level. One drink corresponds to 12 g of pure
alcohol. The P value for interaction as estimated from the nested
log-likelihood test was .70.
0.4-2.1) among smokers of 15 g/d of tobacco or more.
Lastly, analyses were performed on participants free of
gallstone disease and who were consistent light to moderate drinkers (idiopathic pancreatitis). In this subgroup, results were similar to main results.
COMMENT
In this large population-based study with long followup, we found that participants who at baseline reported
smoking or being previous smokers had higher risks of
developing acute and chronic pancreatitis compared with
nonsmokers. We found comparable effect sizes in women
and men and for acute and chronic pancreatitis. Furthermore, results strongly indicate that associations are
independent of alcohol intake and of gallstone disease,
which are risk factors considered to account for most cases
of pancreatitis.
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 6), MAR 23, 2009
607
WWW.ARCHINTERNMED.COM
Downloaded from www.archinternmed.com at University Library of Southern Denmark, on April 28, 2010
©2009 American Medical Association. All rights reserved.
Most previous studies on smoking and pancreatitis find
an increased risk among smokers; however, 2 casecontrol studies31,32 did not observe an increased risk, but
these studies included only individuals with alcohol use
disorders (ie, persons with a heavy alcohol intake and
thus a high risk of pancreatitis). Hence, relative risk estimates of pancreatitis according to smoking may be
smaller than for individuals with a lower alcohol intake.
Only 2 of the 4 studies12,18 in which women are included found an increased risk of pancreatitis according
to smoking in women, and 1 of these studies12 found only
a weak association. In the present study, we observed similar risks in men and women. Recently, Lindkvist et al11
found an increased relative risk of 2.14 for developing
acute pancreatitis in current smokers in a prospective
cohort study. In that study, a dose-response effect was
also observed after controlling for alcohol intake,
whereas no adjustment for gallstone disease was performed. We did not observe an increased risk of pancreatitis according to smoking among participants with
preceding or concomitant gallstone disease. This finding is in accordance with results of a study18 that found
that smoking was only associated with alcohol-associated and idiopathic pancreatitis, which also agrees with
our findings.
The risk of total pancreatitis in ex-smokers was of similar size to the risk among light current smokers (1-14 g/d
of tobacco). This relatively high risk in the ex-smokers
did not seem to be attributable to sick-quitters because
omitting 2 and 4 years of follow-up, respectively, had only
limited effect on the size of the hazard ratio. Lung function (as measured by FEV1) within smoking categories
ranged in the expected order (decreasing in the order of
never-smokers, ex-smokers, and current smokers of 1-14,
15-24, and ⬎25 g/d of tobacco, respectively). This finding further indicates that there was not a high proportion of sick-quitters among the ex-smokers. Also, the
amount of carbon dioxide in expired air was similar among
never-smokers and ex-smokers, which indicates that there
was not a substantial fraction of current smokers who
were misclassified as ex-smokers.
A high proportion of participants were smokers in this
study cohort. A total of 62% were current smokers at their
first enrollment in the CCHS, and we found that approximately 46% of the cases of pancreatitis in this cohort could
be attributed to smoking. Currently, the proportion of
smokers in the Danish population, as in most Western
populations, has decreased to approximately 30%, and
never-smokers and ex-smokers account for approximately 39% and 31%, respectively.33
This study has the advantages of a prospective design, large size, and complete follow-up information on
all participants. A limitation of the study is that the diagnoses of acute and chronic pancreatitis in the Danish
Hospital Discharge Register have not been validated. Only
hospital admissions are registered, and contacts to general practitioners are not included. Hence, it is likely that
not all cases of pancreatitis occurring during the study
period are categorized as such in this study. Also, some
misclassification between the diagnoses of acute and
chronic pancreatitis have probably occurred because the
symptoms and diagnostic criteria of acute and chronic
pancreatitis are overlapping and the 2 diseases can coexist.34 Such misclassification would result in similar observed associations between smoking and risk of acute
and chronic pancreatitis, whereas the results for the joint
outcome of pancreatitis are valid. Furthermore, we did
not have information on the cause of the pancreatitis, and
risk factors for gallstone-related pancreatitis may not be
the same as for alcohol-related or idiopathic pancreatitis.11,18 We addressed this issue in exploratory analyses,
defining gallstone-related pancreatitis as cases of pancreatitis with preceding or concomitant gallstone disease and idiopathic pancreatitis as cases of pancreatitis
with no history of gallstone disease and heavy alcohol
intake. We also observed in accordance with previous
studies18 that smoking does not seem to be associated with
increased risk of gallstone-related pancreatitis.
We observed an 11-fold risk of pancreatitis among participants with gallstone disease. However, this result
should be interpreted with caution because almost all patients with pancreatitis are examined by abdominal ultrasonography at hospitalization, and gallstones will be
detected also in patients with pancreatitis of other causes
with prevalent gallstones unrelated to the attack of pancreatitis. Hence, a part of the rather large relative risk of
pancreatitis according to gallstones may be owing to detection bias.
Apart from the epidemiologic evidence of an association between smoking and development of acute and
chronic pancreatitis, a biological effect of smoking seems
plausible because both animal studies and human studies19-21,35,36 have demonstrated changes of the pancreas and
in pancreatic functioning after exposure to tobacco smoke.
In conclusion, we found that smoking was associated with an increased risk of acute and chronic pancreatitis in both men and women. The risk associated with
smoking was independent of alcohol and gallstone disease, which are risk factors suggested to be the main causes
of pancreatitis.
Accepted for Publication: October 19, 2008.
Correspondence: Janne Schurmann Tolstrup, MSc, PhD,
Center for Alcohol Research, National Institute of Public Health, University of Southern Denmark, Øster
Farimagsgade 5a, DK-1399 Copenhagen K, Denmark (jst
@niph.dk).
Author Contributions: Study concept and design: Tolstrup and Becker. Acquisition of data: Becker and Grønbæk. Analysis and interpretation of data: Tolstrup, Kristiansen, and Becker. Drafting of the manuscript: Tolstrup.
Critical revision of the manuscript for important intellectual content: Tolstrup, Kristiansen, Becker, and Grønbæk. Statistical analysis: Tolstrup. Obtained funding: Grønbæk. Study supervision: Becker and Grønbæk.
Financial Disclosure: None reported.
Funding/Support: This study was supported by grants
from the Danish National Board of Health and the Danish Medical Research Council.
Role of the Sponsors: The funding organizations did not
have any role in either the design or conduct of the study
or preparation, review, or approval of the manuscript.
Additional Contributions: We thank the participants of
the CCHS for their time and cooperation.
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 6), MAR 23, 2009
608
WWW.ARCHINTERNMED.COM
Downloaded from www.archinternmed.com at University Library of Southern Denmark, on April 28, 2010
©2009 American Medical Association. All rights reserved.
REFERENCES
1. O’Sullivan JN, Nobrega FT, Morlock CG, Brown AL Jr, Bartholomew LG. Acute
and chronic pancreatitis in Rochester, Minnesota, 1940 to 1969. Gastroenterology.
1972;62(3):373-379.
2. Corfield AP, Cooper MJ, Williamson RC. Acute pancreatitis: a lethal disease of
increasing incidence. Gut. 1985;26(7):724-729.
3. Lankisch PG, Schirren CA, Schmidt H, Schonfelder G, Creutzfeldt W. Etiology
and incidence of acute pancreatitis: a 20-year study in a single institution. Digestion.
1989;44(1):20-25.
4. Wilson C, Imrie CW. Changing patterns of incidence and mortality from acute
pancreatitis in Scotland, 1961-1985. Br J Surg. 1990;77(7):731-734.
5. McKay CJ, Evans S, Sinclair M, Carter CR, Imrie CW. High early mortality rate from
acute pancreatitis in Scotland, 1984-1995. Br J Surg. 1999;86(10):1302-1305.
6. Eland IA, Sturkenboom MJ, Wilson JH, Stricker BH. Incidence and mortality of
acute pancreatitis between 1985 and 1995. Scand J Gastroenterol. 2000;35
(10):1110-1116.
7. Tinto A, Lloyd DA, Kang JY, et al. Acute and chronic pancreatitis–diseases on
the rise: a study of hospital admissions in England 1989/90-1999/2000. Aliment
Pharmacol Ther. 2002;16(12):2097-2105.
8. Goldacre MJ, Roberts SE. Hospital admission for acute pancreatitis in an English population, 1963-98: database study of incidence and mortality. BMJ. 2004;
328(7454):1466-1469.
9. Apte MV, Pirola RC, Wilson JS. Molecular mechanisms of alcoholic pancreatitis.
Dig Dis. 2005;23(3-4):232-240.
10. DiMagno MJ, DiMagno EP. Chronic pancreatitis. Curr Opin Gastroenterol. 2005;
21(5):544-554.
11. Lindkvist B, Appelros S, Manjer J, Berglund G, Borgstrom A. A prospective cohort study of smoking in acute pancreatitis. Pancreatology. 2008;8(1):63-70.
12. Yen S, Hsieh CC, MacMahon B. Consumption of alcohol and tobacco and other
risk factors for pancreatitis. Am J Epidemiol. 1982;116(3):407-414.
13. Lowenfels AB, Zwemer FL, Jhangiani S, Pitchumoni CS. Pancreatitis in a native
American Indian population. Pancreas. 1987;2(6):694-697.
14. Bourliere M, Barthet M, Berthezene P, Durbec JP, Sarles H. Is tobacco a risk factor for chronic pancreatitis and alcoholic cirrhosis? Gut. 1991;32(11):13921395.
15. Talamini G, Bassi C, Falconi M, et al. Cigarette smoking: an independent risk factor in alcoholic pancreatitis. Pancreas. 1996;12(2):131-137.
16. Talamini G, Bassi C, Falconi M, et al. Alcohol and smoking as risk factors in chronic
pancreatitis and pancreatic cancer. Dig Dis Sci. 1999;44(7):1303-1311.
17. Lin Y, Tamakoshi A, Hayakawa T, Ogawa M, Ohno Y; Research Committee on
Intractable Pancreatic Diseases. Cigarette smoking as a risk factor for chronic
pancreatitis: a case-control study in Japan. Pancreas. 2000;21(2):109-114.
18. Morton C, Klatsky AL, Udaltsova N. Smoking, coffee, and pancreatitis. Am J
Gastroenterol. 2004;99(4):731-738.
19. Wittel UA, Hopt UT, Batra SK. Cigarette smoke-induced pancreatic damage:
experimental data. Langenbecks Arch Surg. 2008;393(4):581-588.
20. Wittel UA, Pandey KK, Andrianifahanana M, et al. Chronic pancreatic inflammation induced by environmental tobacco smoke inhalation in rats. Am J Gastroenterol.
2006;101(1):148-159.
21. Chowdhury P, Doi R, Tangoku A, Rayford PL. Structural and functional changes
of rat exocrine pancreas exposed to nicotine. Int J Pancreatol. 1995;18(3):
257-264.
22. Doi R, Chowdhury P, Nishikawa M, et al. Carbachol and cholecystokinin enhance accumulation of nicotine in rat pancreatic acinar cells. Pancreas. 1995;
10(2):154-160.
23. Jørgensen T, Kay L, Schultz-Larsen K. The epidemiology of gallstones in a 70year-old Danish population. Scand J Gastroenterol. 1990;25(4):335-340.
24. Jørgensen T. Gall stones in a Danish population: relation to weight, physical activity, smoking, coffee consumption, and diabetes mellitus. Gut. 1989;30(4):
528-534.
25. Jürgensen HJ, Frølund C, Gustafsen J, Mosbech H, Guldhammer B, Mosbech J.
Registration of diagnoses in the Danish National Registry of patients. Methods
Inf Med. 1986;25(3):158-164.
26. Juel K, Helweg-Larsen K. The Danish registers of causes of death. Dan Med Bull.
1999;46(4):354-357.
27. Mosbech J, Jorgensen J, Madsen M, Rostgaard K, Thornberg K, Poulsen TD.
The national patient registry: evaluation of data quality [in Dutch]. Ugeskr Laeger.
1995;157(26):3741-3745.
28. World Health Organization. International Classification of Diseases, Eighth Revision (ICD-8). Geneva, Switzerland: World Health Organization; 1969.
29. World Health Organization. International Statistical Classification of Diseases, 10th
Revision (ICD-10). Geneva, Switzerland: World Health Organization; 1992.
30. Wacholder S, Benichou J, Heineman EF, Hartge P, Hoover RN. Attributable risk:
advantages of a broad definition of exposure. Am J Epidemiol. 1994;140(4):
303-309.
31. Haber PS, Wilson JS, Pirola RC. Smoking and alcoholic pancreatitis. Pancreas.
1993;8(5):568-572.
32. Lévy P, Mathurin P, Roqueplo A, Rueff B, Bernades P. A multidimensional casecontrol study of dietary, alcohol, and tobacco habits in alcoholic men with chronic
pancreatitis. Pancreas. 1995;10(3):231-238.
33. Ekholm E, Kjoller M, Davidsen M, et al. The National Health Interview Surveys.
Copenhagen, Denmark: National Institute of Public Health; 2006.
34. Kloppel G, Maillet B. Development of chronic pancreatitis from acute pancreatitis: a pathogenetic concept [in German]. Zentralbl Chir. 1995;120(4):274-277.
35. Balldin G, Borgstrom A, Eddeland A, Genell S, Hagberg L, Ohlsson K. Elevated
serum levels of pancreatic secretory proteins in cigarette smokers after secretin
stimulation. J Clin Invest. 1980;66(1):159-162.
36. Bynum TE, Solomon TE, Johnson LR, Jacobson ED. Inhibition of pancreatic secretion in man by cigarette smoking. Gut. 1972;13(5):361-365.
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 6), MAR 23, 2009
609
WWW.ARCHINTERNMED.COM
Downloaded from www.archinternmed.com at University Library of Southern Denmark, on April 28, 2010
©2009 American Medical Association. All rights reserved.