Increased Stroke Risk Is Related to a Binge Drinking Habit
Laura Sundell, MD; Veikko Salomaa, MD, PhD; Erkki Vartiainen, MD, PhD;
Kari Poikolainen, MD, PhD; Tiina Laatikainen, MD, PhD
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
Background and Purpose—Heavy alcohol consumption increases the risk for all strokes, whereas moderate regular
alcohol consumption is associated with a lower risk for ischemic stroke. The purpose of this study was to evaluate the
effect of different drinking patterns on stroke risk, independent of average alcohol intake.
Methods—A prospective cohort study of 15 965 Finnish men and women age 25 to 64 years who participated in a national
risk factor survey and had no history of stroke at baseline were followed up for a 10-year period. The first stroke event
during follow-up served as the outcome of interest (N⫽249 strokes). A binge drinking pattern was defined as
consuming 6 or more drinks of the same alcoholic beverage in men or 4 or more drinks in women in 1 session.
Cox proportional-hazards models were adjusted for average alcohol consumption, age, sex, hypertension, smoking,
diabetes, body mass index, educational status, study area, study year, and history of myocardial infarction.
Results—Binge drinking was an independent risk factor for total and ischemic strokes. Compared with non– binge drinkers,
the hazard ratio for total strokes among binge drinkers was 1.85 (95% CI, 1.35 to 2.54) after adjusting for average
alcohol consumption, age, and sex; the association was diluted after adjustment for other risk factors. Compared with
non– binge drinkers, the risk for ischemic stroke was 1.99 (95% CI, 1.39 to 2.87) among binge drinkers; the association
remained statistically significant after adjustment for potential confounders.
Conclusions—This study found that a pattern of binge drinking is an independent risk factor for all strokes and ischemic
stroke. (Stroke. 2008;39:3179-3184.)
Key Words: alcohol drinking 䡲 stroke 䡲 population 䡲 risk factors
T
he role of alcohol consumption in stroke risk has been
widely studied. Many cohort and case-control studies
have found an increased risk for hemorrhagic and ischemic
strokes to be related to heavy alcohol consumption.1–3 A U- or
J-shaped association has been observed between alcohol
consumption and ischemic stroke, wherein light to moderate
alcohol consumption has been shown to protect against
ischemic stroke.1– 4 The protective effect of light to moderate
alcohol consumption against ischemic stroke has been assumed
to be mediated through the effects of alcohol on lipid peroxidation, increased HDL cholesterol, alcohol-induced decreases in
platelet aggregation, and increased fibrinolytic activity.
The association of alcohol consumption with hemorrhagic
stroke is more linear than its association with ischemic
stroke.1 Alcohol consumption increases the risk of both
intracerebral (ICH) and subarachnoid (SAH) hemorrhages.4 –7
Increased risk for ICH due to alcohol consumption has been
attributed to alcohol-induced hypertension and impaired hemostasis. A transient increase in blood pressure during heavy
alcohol intake and withdrawal may contribute to the rupture
of small cerebral arteries or to an aneurysm, suggesting that
episodic heavy drinking is a particular risk factor for SAH.6
The effects of binge drinking on health are suspected to be
different from those of regular drinking. Although binge
drinking is an independent risk factor for ischemic heart
disease mortality,8,9 only a few studies have assessed the
effects of different drinking patterns on stroke risk. Finnish
researchers Hillbom and Kaste were the first to show that
acute alcohol intoxication predisposes to ischemic strokes
and aneurysmal SAHs within 24 hours.10,11 We have studied
the long-term predictive value of binge drinking on stroke
incidence.
A Swedish cohort study reported a higher risk of ischemic
stroke among infrequent drinkers, occasional binge drinkers,
and men who sometimes felt intoxicated than among lifelong
abstainers.12 Regular light to moderate alcohol consumption
was found to protect against ischemic stroke, but irregular or
sporadic alcohol consumption reduced this benefit.2
The purpose of this prospective cohort study was to
investigate the risk of total stroke, ischemic stroke, and
hemorrhagic stroke in relation to alcohol drinking patterns
and to determine whether a binge drinking pattern is a risk
factor for stroke after adjusting for average alcohol consumption and other potential confounders.
Received March 19, 2008; final revision received April 20, 2008; accepted May 5, 2008.
From the Department of Health Promotion and Chronic Disease Prevention (L.S., V.S., E.V., T.L.), National Public Health Institute (KTL), Helsinki;
the University of Kuopio (L.S.), Kuopio; and the Finnish Foundation for Alcohol Studies and Department of Mental Health and Alcohol Research (K.P.),
National Public Health Institute (KTL), Helsinki, Finland.
Correspondence to Laura Sundell, MD, National Public Health Institute, Mannerheimintie 166, FI-00300 Helsinki, Finland. E-mail [email protected]
© 2008 American Heart Association, Inc.
Stroke is available at http://stroke.ahajournals.org
DOI: 10.1161/STROKEAHA.108.520817
3179
3180
Stroke
December 2008
Subjects and Methods
Subjects
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
The study population consisted of subjects examined in the national
FINRISK surveys in the years 1987, 1992, and 1997. A stratified,
random sample of people age 25 to 64 years living in the provinces
of North Karelia and Kuopio, southwestern Finland, the HelsinkiVantaa region, and the Oulu region was drawn from the Finnish
population register. Stratification was performed by sex and by
10-year age group so that in the year 1987, the study sample
comprised 500 men and women in each age group from the province
of North Karelia and 250 men and women in each age group from the
province of Kuopio and southwestern Finland. In 1992, the study
sample comprised 250 men and women in each age group from the
provinces of North Karelia and Kuopio, southwestern Finland, and
the Helsinki-Vantaa region. In 1997, the study sample comprised 250
men and women in each 10-year age group from all 5 geographically
defined areas, leading to a total sample size of 26 000 subjects. Of
these subjects, 19 689 participated in the baseline survey. The
response rate was 81% in 1987, 76% in 1992, and 72% in 1997,
yielding an overall response rate of 76%. Those subjects (n⫽480)
who participated in the baseline survey more than once were counted
eligible only in the first study year. Subjects with previous stroke
(n⫽318) were excluded from the analyses. The total number of
subjects was 18 891. The number of those subjects who reported
alcohol consumption during the previous year was 16 143. Information on alcohol drinking patterns and on all other covariates was
obtained for 15 256 subjects.
Baseline Survey
The baseline risk factor survey was performed with standardized
methods according to the WHO MONICA protocol.13 A selfadministered questionnaire with questions on health behavior, diseases, symptoms, socioeconomic factors, and medical history was
mailed to the subjects with an invitation to attend a health examination. Laboratory tests and physical measurements were performed at
the health examination site.
Average alcohol consumption was assessed by a self-administered
questionnaire regarding the usual quantity and frequency of consumption during the past 12 months before the study. Although
changes in alcohol consumption may take place over time, it is
generally assumed that such data on average alcohol consumption
yield a reasonably good estimate of the long-term average alcohol
intake. Consumption was assessed separately for beer, wine, and
spirits with the following questions: “How often do you usually drink
beer (similar questions for wine and spirits)? ” and “How much beer
do you usually drink at a time (similar questions for wine and
spirits)?” Pure alcohol content per 1 standard drink of alcoholic
beverage was estimated to be 12.5 g for beer, 12.0 g for wine, and
12.0 g for spirits, and 1 standard drink was estimated to contain 33
cL for beer, 12 cL for wine, and 4 cL for spirits.
Heavy drinkers were defined as those consuming ⬎350 g/wk for
men and ⬎210 g/wk for women. Moderate drinkers were defined as
those drinking below these limits but ⬎230 g/wk for men and ⬎150
g/wk for women, and the rest were defined as light drinkers. A binge
drinking pattern was defined for men as consuming 6 or more drinks
of the same beverage type on 1 occasion, and for women, 4 drinks.
The widely used AUDIT screening questionnaire includes a question
on consuming 6 or more drinks at a time, and the criterion for binge
drinking for men was chosen to follow this definition. 14 The
questionnaire did not allow us to identify those subjects who
consumed different types of alcoholic beverages on 1 occasion and
reached a level of binge drinking only when the number of drinks of
different beverages was summed. However, an earlier study in
Finland8 showed that 83% of all binge drinkers reported spirit
consumption and 52% reported consuming only spirits. Only 10% of
binge drinkers reported consuming both spirits and beer, 11% spirits
and wine, and 9% spirits, wine, and beer. Seventeen percent
preferred beer or wine during heavy drinking occasions, and among
them, 8% reported only beer consumption, 7% only wine consumption, and 2% wine and beer consumption. These results suggest that
most binge drinkers consume only spirits on heavy drinking
occasions.
Education was assessed with a question about the total years of
full-time education. To control for the effect of birth year on
educational status, we used the following education categories: low,
medium, and high. These categories were calculated with birth
year–specific tertiles of total years of education.
Physical measurements (height, weight, blood pressure) were
performed by trained nurses at the examination site. Body mass
index (BMI, kg/m2) was calculated as a measure of obesity. Blood
pressure was measured from the right arm of the participant after 15
minutes of rest in a sitting position. Appearance of the Korotkoff
sounds was recorded as the systolic blood pressure and the fifth
phase as the diastolic blood pressure. The measurement was repeated
after 1 minute, giving 2 systolic and 2 diastolic blood pressure
values. The latter blood pressure value was used in this study.
Participants were classified as having hypertension if their blood
pressure was ⱖ140/90 mm Hg at baseline or if they were receiving
antihypertensive drug therapy and had taken their medication within
the previous week.
Smoking was assessed by structured self-administered questions
about smoking habits. Based on their responses, participants were
classified into 2 categories: current smokers and nonsmokers. Current smokers referred to those who had smoked regularly for at least
1 year and had smoked daily during the previous month. Other
respondents were classified as nonsmokers (never-smokers, occasional smokers, ex-smokers). Diabetes was assessed on the basis of
hospital discharge register data and self-reported diagnoses of
diabetes. Subjects were classified as having diabetes if they reported
either physician-diagnosed diabetes or impaired glucose tolerance or
if they had been hospitalized with a diagnosis of diabetes (ie, the
National Hospital Discharge Register indicated ICD-9 code 250 or
ICD-10 codes E10 or E11). Causes of death and hospital discharges
were classified with ICD-9 coding until the end of 1995 and with
ICD-10 coding from the beginning of 1996 because use of these
codes is standard in Finland.
Subjects were classified as having a history of myocardial infarction if on the questionnaire they reported myocardial infarction
diagnosed by a physician or if they had been hospitalized with ICD-9
diagnosis codes 410, 411, or 412 or with ICD-10 codes I21, I22,
I20.0, or I25.2 before the baseline examination. The hospitalizations
were identified by record linkage of the study data with the National
Hospital Discharge Register.
Criteria for previous stroke were either self-reported stroke diagnosed by a physician before the baseline examination or hospitalization with a diagnosis of stroke, ie, ICD-9 codes 430, 431, 436,
4330A, 4331A, 4339A, 4340A, 4341 or 4349A or ICD-10 codes I63,
I60, I64, or I61, referring to ICH, ischemic stroke, and SAH
diagnosed before the baseline study year. Subjects with previous
strokes (n⫽318) were excluded from the statistical analyses.
Follow-Up
The study cohort was followed up for fatal and nonfatal strokes.
Follow-up information was based on the Finnish National Hospital
Discharge Register for nonfatal stroke events and on the National
Causes of Death Register for fatal events. Survey data were linked to
both the hospital discharge register and the causes of death register
through identification numbers assigned to every resident of Finland.
The follow-up period was 10 years if the baseline risk factor survey
was conducted in 1987 or 1992 and 9 years if the baseline survey was
carried out in 1997. Fatal and nonfatal stroke events were analyzed
together. The coverage of the follow-up was 100% for stroke events
that occurred in Finland during the follow-up period. In total, the
follow-up period comprised 149 425 person-years.
The outcome event was defined as either the first fatal or nonfatal
stroke. ICD codes 431 and I61 were classified as ICH. ICD codes
430 and I60 were classified as SAH. ICD codes 436, I63, 4340A,
4331A, 4339A, 4330A, 4341A, and 4349A were classified as
ischemic stroke.
Sundell et al
Increased Stroke Risk Related to Binge Drinking
Table 1.
Statistical Methods
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
The differences in background characteristics of the study participants with and without a pattern of binge drinking were analyzed
with t tests and ␹2 tests as appropriate. The Cox proportional-hazards
model served to assess the independent contribution of a binge
drinking pattern to the risk of stroke and to evaluate relative risks
adjusted for average alcohol consumption and other risk factors. The
main interest was to evaluate the contribution of the binge drinking
habit to the risk of stroke among those subjects who consumed
alcohol. Therefore, those who reported consuming alcohol in the
previous year participated in the analyses, and unless otherwise
indicated, the results were applied to that study sample. To evaluate
the risk of the binge drinking habit on the entire population sample,
we included those who quit consuming alcohol or had never
consumed alcohol.
The analyses were carried out for all stroke events, ischemic
stroke events, and hemorrhagic stroke events. Total stroke events
(n⫽249) refer to ICHs (n⫽33), SAHs (n⫽37), and ischemic strokes
(n⫽179) analyzed together. Of all strokes, 26 were fatal and 223
nonfatal. Because of the small number of hemorrhagic stroke events,
ICHs and SAHs were combined to assess the risk of hemorrhagic
stroke. Ischemic strokes were analyzed as a single group.
Long-term average alcohol consumption, age, study year, and
BMI were included in the models as continuous variables. Because
of a skewed distribution, logarithmic transformation of average
alcohol consumption was used in the models. Sex, education, study
area, hypertension, diabetes, smoking, and a history of coronary
heart disease were included as categorical variables. The basic model
included average alcohol consumption, sex, and age as covariates.
Next, potential confounders were added to the model 1 by 1.
Confounders included blood pressure, diabetes, smoking, study area,
study year, education, and a history of myocardial infarction.
To assess the contribution of average alcohol consumption to the
risk of stroke, Cox proportional-hazards models were performed by
including logarithmically transformed alcohol consumption and
other covariates, except the drinking pattern, into the models. To
evaluate whether stroke risk for binge drinkers differed among men
and women, we evaluated a possible interaction between drinking
pattern and sex. We found no such interaction (P⫽0.3), so analyses
were carried out for men and women together. The proportionalhazards assumption was examined graphically and found to be valid.
All analyses were performed with the SAS statistical package,
version 8.2 (SAS Inc, Cary, NC).
Results
The study cohort included 3558 subjects (1053 women and
2505 men) with a binge drinking pattern and 12 407 subjects
with no binge drinking pattern (Table 1). The subjects with a
binge drinking pattern were younger, consumed on average
significantly more alcohol, had a higher BMI, more often had
hypertension, smoked notably more, and had a lower educational status than did subjects with no binge drinking pattern.
Binge drinkers also had fewer myocardial infarctions. Drinking patterns differed slightly between geographic areas.
Subjects living in the province of North Karelia had binge
drinking patterns less frequently than did those living in other
study areas. No difference was observed in diabetes incidence
between the groups. Subjects with a binge drinking pattern
had an increased risk for all strokes and ischemic stroke,
whereas no such association was found between that drinking
pattern and hemorrhagic stroke.
Compared with subjects with no binge drinking pattern, the
hazard ratio (HR) for any stroke was 1.85 (95% CI, 1.35 to
2.54) among binge drinkers after adjusting for average
alcohol consumption, sex, and age (Table 2). The HR
remained significantly higher for binge drinkers after control-
3181
Characteristics of Participants by Drinking Pattern
Characteristics
No. of participants (%)
No. of ischemic strokes
No. of hemorrhagic
strokes
Binge
Drinking Pattern
3558 (22)
No Binge Drinking
Pattern
12 407 (77)
122
57
50
20
P Value
⬍0.001
Men, n (%)
2505 (70)
5510 (44)
⬍0.001
Age, y (SD)*
41.3 (10.5)
44.8 (11.2)
⬍0.0001
Alcohol intake, g/wk (%)
Light drinkers†
Moderate drinkers‡
76.1%
96.4%
9.6%
2.4%
14.3%
1.2%
⬍0.0001
26.9 (4.6)
25.9 (4.26)
⬍0.0001
History of myocardial
infarction, n (%)
70 (2)
330 (2.7)
0.019
Hypertension, n (%)
1614 (45)
Heavy drinkers§
BMI, kg/m2 (SD)*
5125 (41)
⬍0.0001
Diabetes, n (%)
138 (3.9)
471 (3.8)
Smoking, n (%)
1945 (55)
2690 (22)
⬍0.0001
NS
Lower, n (%)
1751 (50)
4677 (38)
⬍0.001
Medium, n (%)
1048 (30)
3739 (31)
Higher, n (%)
717 (20)
3814 (32)
Education
Values are frequencies (percentages), and the ␹2 test was used as the
statistical test.
*Values are mean (SD), and Student’s t test was used as the statistical
method.
†Total alcohol consumption 0 –230 g/wk for men and 0 –150 g/wk for
women.
‡Total alcohol consumption 231–350 g/wk for men and 151–210 g/wk for
women.
§Total alcohol consumption ⬎350 g/wk for men and ⬎210 g/wk for
women.
ling for average alcohol consumption, age, sex, study area,
hypertension, smoking, diabetes, and BMI. The risk was
diluted after further adjustments for education and a history
of myocardial infarction. The risk for ischemic stroke was
significantly higher for binge drinkers than for subjects with
no heavy drinking pattern. The HR was 1.99 (95% CI, 1.39 to
2.87) after controlling for average alcohol consumption, sex,
and age (Table 3). The HR remained significant (HR⫽1.56;
95% CI, 1.06 to 2.31) after adjusting for other covariates. The
HR for hemorrhagic stroke was 1.50 (95% CI, 0.79 to 2.83)
after adjusting for alcohol consumption, sex, and age.
Furthermore, we found no association between average
alcohol consumption and any stroke. The HR for any stroke
was 1.07 (95% CI, 0.99 to 1.16) after adjusting for sex and
age. The HRs were 1.09 (95% CI, 0.99 to 1.21) for ischemic
stroke and 1.00 (95% CI, 0.86 to 1.17) for hemorrhagic
stroke. After nondrinkers were included, the risk for any
stroke and for ischemic stroke for binge drinkers increased.
The HR for any stroke was 1.47 (95% CI, 1.06 to 2.04)
after adjusting for all covariates. The respective HR for
ischemic stroke was 1.72 (95% CI, 1.18 to 2.50). No
association was found for hemorrhagic stroke (HR⫽0.9;
95% CI, 0.47 to 1.79).
3182
Stroke
December 2008
Table 2. HRs (95% CIs) for Any Stroke Event Among Binge
Drinkers Compared With Non–Binge Drinkers (Nⴝ15 256)
Table 3. HRs (95% CIs) for Ischemic Stroke Events Among
Binge Drinkers Compared With Non–Binge Drinkers (Nⴝ15 256)
Model 1
HR (95% CI)
Model 2
HR (95% CI)
Binge drinking pattern
1.85 (1.35–2.54)
1.39 (0.99–1.95)
Binge drinking pattern
1.99 (1.39–2.87)
1.56 (1.06–2.31)
Long-term average alcohol
consumption*
1.00 (0.92–1.01)
0.98 (0.89–1.07)
Long-term average alcohol
consumption*
1.02 (0.92–1.13)
1.00 (0.89–1.12)
Age (y)
1.10 (1.09–1.21)
1.09 (1.07–1.11)
Age (y)
1.13 (1.11–1.15)
1.12 (1.09–1.14)
Sex (0, 1)
0.52 (0.39–0.71)
0.61 (0.44–0.83)
Sex (0, 1)
0.49 (0.34–0.71)
0.57 (0.39–0.84)
Study area
Model 1
HR (95% CI)
Model 2
HR (95% CI)
Study area
North Karelia
1.00
North Karelia
1.00
Kuopio region
1.07 (0.76–1.53)
Kuopio region
1.02 (0.67–1.550)
Southwestern Finland
0.90 (0.62–1.32)
Southwestern Finland
0.79 (0.51–1.25)
Helsinki region
1.21 (0.79–1.84)
Helsinki region
1.22 (0.76–1.98)
Oulu region
0.67 (0.32–1.39)
Oulu region
0.78 (0.35–1.72)
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
Hypertension (0, 1)
1.64 (1.20–2.24)
Hypertension (0, 1)
1.34 (0.94–1.91)
Smoking (0, 1)
2.21 (1.67–2.93)
Smoking (0, 1)
1.99 (1.43–2.78)
Diabetes (0, 1)
2.01 (1.35–2.98)
Diabetes (0, 1)
2.04 (1.30–3.20)
BMI (kg/m2)
1.04 (1.01–1.08)
BMI (kg/m2)
1.06 (1.03–1.10)
Level of education (1, 2, 3)
1.03 (0.87–1.22)
Level of education (1, 2, 3)
1.03 (0.84–1.23)
History of myocardial
infarction (0, 1)
1.74 (0.94–1.01)
History of myocardial
infarction (0, 1)
1.49 (0.91–2.48)
Study year
0.98 (0.94–1.01)
Study year
0.97 (0.93–1.02)
Model 1 was adjusted for average long-term alcohol consumption, age, and sex.
Model 2 was adjusted for covariates in model 1 and additionally for study
area, study year, hypertension, smoking, diabetes, BMI, education, and
coronary heart disease.
*Logarithmically transformed.
Model 1 was adjusted for average long-term alcohol consumption, age, and sex.
Model 2 was adjusted for covariates in model 1 and additionally for study
area, study year, hypertension, smoking, diabetes, BMI, education, and
coronary heart disease.
*Logarithmically transformed.
Discussion
was strongly related to stroke risk. That study suggested that
the serum ␥-glutamyl transferase level predicts stroke risk
better than does self-reported alcohol consumption.
In this prospective cohort study, we found that alcohol
consumption with a binge drinking pattern was an independent risk factor for all strokes and ischemic stroke after
adjusting for average alcohol consumption and other potential
confounders. We found an increased risk for total stroke and
ischemic stroke but not for hemorrhagic stroke. However,
ischemic stroke comprised ⬎70% of all strokes, and therefore
the increased stroke risk relates mostly to ischemic stroke.
Many previous studies have found an increased risk for
hemorrhagic stroke among heavy drinkers. In this study, the
number of cases of hemorrhagic stroke was small, and the
statistical power was probably insufficient to detect any
association between binge drinking and hemorrhagic stroke.
Owing to the small number of hemorrhagic strokes, we
combined ICHs and SAHs and analyzed them together.
However, the natural cause and pathogenesis of ICH and
SAH differ, and this may also dilute the association between
binge drinking and hemorrhagic stoke. Furthermore, numerous cohort studies have found a positive association between
average alcohol consumption and stroke. In this study,
however, we found no association between average alcohol
consumption and any type of stroke. This could be due to the
small number of subjects reporting a heavy average alcohol
consumption in our study sample. This was tested in a
previous study with the same cohorts.15 Self-reported alcohol
consumption was not associated with stroke risk, but excessive alcohol consumption as assessed by laboratory markers
Mechanisms
The role of hypertension as a risk factor for stroke, both
ischemic and hemorrhagic, is well known.16 Hypertension is
considered a major risk factor for all types of stroke, and
other risk factors are thought to act in concert with hypertension.17 Alcohol consumption has been shown to increase
blood pressure.18 –20 Long-term habitual alcohol consumption
results in a dose-dependent gradual elevation in blood pressure. However, drinking large amounts of alcohol on 1
occasion was found to increase blood pressure more than the
same amount of alcohol consumed over a longer period.21
Acute alcohol consumption and withdrawal have been shown
to cause a transient rise in systolic and diastolic blood
pressure in normotensive subjects, suggesting that heavy
episodic drinking has an adverse affect on blood pressure
independent of the average intake level.22 In this study, we
found that binge drinking increased ischemic stroke risk
independently of the average alcohol consumption and other
confounders, including hypertension. Because of the known
causal pattern of association between alcohol consumption,
blood pressure, and stroke, controlling for hypertension may
obscure the relation between alcohol consumption and stroke.
Therefore, we conducted the analyses with and without
hypertension as a covariate. The HR was not attenuated after
Sundell et al
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
controlling for hypertension, suggesting that binge drinking
may increase stroke risk through a mechanism other than
hypertension. This finding is consistent with the findings of
previous case-control studies, which found recent heavy
alcohol consumption to be a risk factor for ICH23 and for
ischemic stroke due to cardiogenic brain embolism24 after
controlling for hypertension.
Heavy, chronic alcohol consumption is associated with
cardiomyopathy,25 which is known to predispose to cardiac
arrhythmias and heart failure, both of which increase the risk
for embolic stroke.26,27 Furthermore, heavy episodic drinking,
even of short duration, and the subsequent withdrawal period
have been shown to trigger cardiac arrhythmias of both
ventricular and supraventricular origin. These arrhythmias are
often called“ holiday heart.” The most common arrhythmias
observed after heavy drinking include atrial fibrillation and
flutter.28 If sustained, arrhythmias predispose to embolic
stroke. Even though heavy alcohol drinking predisposes to
cardiac arrhythmias, arrhythmias are often of relatively short
duration and cease after cessation of drinking. Therefore, they
may be considered a relatively minor risk factor for embolic
stroke if found in an otherwise healthy heart without preceding heart failure or other abnormalities. A Finnish casecontrol study29 reported increased ischemic stroke risk among
persons who consumed 151 to 300 g and ⬎300 g of alcohol
within the week preceding the onset of stroke. Consumption
of ⬎40 g of alcohol within the preceding 24 hours increased
the risk for ischemic stroke due to cardiogenic embolism with
a predisposing high-risk source, referring to atrial fibrillation,
recent myocardial infarction together with an akinetic or
hypokinetic left ventricular segment, dilated cardiomyopathy,
left atrial thrombus, or patent foramen ovale. Furthermore,
heavy alcohol intake within 24 hours before the onset of
stroke was found to be a risk factor for ischemic stroke due to
large-artery atherosclerosis. This was suspected to be the effect
of acute heavy drinking on blood circulation, predisposing the
dislodging of a local thrombus from the artery wall.
Light to moderate alcohol consumption has been related to
lower rates of ischemic heart disease and ischemic stroke.
The beneficial effect of moderate consumption is associated
with a reduction in atherosclerosis. However, an angiographic
study30 measured coronary occlusion in relation to the total
amount of alcohol consumed, as well as to the pattern of
alcohol intake. Regular drinkers showed significantly lower
coronary occlusion scores than did occasional drinkers or
nondrinkers, whereas a variable drinking pattern was associated with higher occlusion scores. Accelerated atherosclerosis
associated with a pattern of occasional drinking has been
linked to adverse lipoprotein composition, where the lowest
total to HDL cholesterol ratio occurred among regular drinkers. Chronic heavy alcohol consumption is associated with a
variety of blood clotting and platelet abnormalities. Effects on
platelets and the coagulation cascade are considered a crucial
mechanism for the precipitation of acute strokes by alcohol.
In alcoholic patients with no cirrhosis of the liver, the
primary effect is on blood platelets, where both qualitative
and quantitative abnormalities appear due to alcohol ingestion. These alterations cause a prolongation of the bleeding
time and predispose to hemorrhagic complications, including
Increased Stroke Risk Related to Binge Drinking
3183
hemorrhagic stroke. Alcoholics with cirrhosis have decreased
levels of clotting factors in the presence of both platelet
abnormalities and coagulation defects.31 In terms of ischemic
stroke, acute alcohol consumption has been found to be
associated with increased platelet aggregation and
thromboxane formation. Furthermore, acute alcohol ingestion
has been shown to decrease fibrinolytic activity, to increase
factor VIII complex, and to shorten bleeding time.32 Also, the
alcohol withdrawal period has been shown to be associated
with alterations in platelet count and function.33 Reactive
thrombocytosis and increased platelet aggregability during
withdrawal may explain why drinkers with a habit of binge
drinking have an increased risk for ischemic stroke.
Strengths and Limitations
Self-reported alcohol consumption is usually underestimated.34,35 Underestimation has been found to be a consequence mainly of underreported drinking frequencies, probably because people tend to forget light drinking sessions.36
Amounts of alcohol consumed at 1 time are reported more
accurately. In the present study, we found no association
between average alcohol consumption and stroke. This could
be due to the underestimation of alcohol intake. Elevated
␥-glutamyl transferase levels indicating heavy alcohol consumption were found to be associated with increased stroke
risk in a previous study of the same cohort.15 Another
weakness is the lack of data on possible changes in alcohol
intake over time.
In this study, heavy episodic drinkers, also known as binge
drinkers, were defined as those men who consumed 6 or more
drinks of the same alcoholic beverage type (beer, wine, or
spirits) on 1 occasion and those women who consumed 4 or
more drinks. Subjects who reached these limits by consuming
drinks from more than 1 beverage type category on 1
occasion had to be classified as non– binge drinkers, because
we did not have the information necessary to identify them
separately. This probably diluted the observed associations
between binge drinking and stroke, and the true associations
might be stronger.
The risk for ischemic stroke among binge drinkers was
attenuated only slightly after adjustment for average alcohol
consumption, hypertension, age, sex, smoking, diabetes, obesity, study area, and education. Even though these are
considered major risk factors for stroke, especially for ischemic stroke, other relevant confounders that we were unable to
control for may exist. The use of oral contraceptives, hormone replacement therapy, and migraine are also considered
risk factors for stroke.37–39 Their contribution to stroke is
closely related to that of other risk factors (mostly hypertension and smoking) that we controlled for. However, some
evidence suggests that migraine may be an independent risk
factor stroke, but this evidence seems to apply mostly to
subjects who experience migraine with aura and to oral
contraceptive users.39
The strengths of this study lie in the large general population cohort followed up for the first stroke event by record
linkage to the National Hospital Discharge Register and the
National Causes of Death Register. These registers have good
coverage, and the accuracy of stroke diagnosis is high.40,41 All
3184
Stroke
December 2008
symptomatic stroke patients are hospitalized in Finland, and
nearly all are diagnosed by routine computed tomography or
magnetic resonance imaging examination, which is required for
precise specification of stroke subtype.42 Our end point definition included symptomatic strokes, whereas subjects with mild,
subclinical cases who were not hospitalized were not included.
The large general population cohort, including information
on alcohol consumption, drinking patterns, and other risk
factors, allowed us to study the effect of a binge drinking
habit on stroke risk after controlling for the effect of habitual
alcohol consumption. Because of the heterogeneity of the
nondrinking group, including lifelong teetotallers and abstainers who had quit drinking for health reasons, we compared drinkers with and without a binge drinking pattern.
15.
16.
17.
18.
19.
20.
21.
Conclusions
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
This prospective cohort study showed that a binge drinking
pattern is an independent risk factor for stroke; the highest
risk was noted for ischemic stroke. A previous study found an
increased risk for ischemic heart disease among binge drinkers.8 These results suggest that a binge drinking pattern
adversely affects cardiovascular outcomes independently of
average alcohol consumption.
22.
23.
24.
25.
Sources of Funding
26.
This study was supported by grants from the Finnish Foundation for
Alcohol Studies and the Yrjö Jahnsson Foundation.
27.
Disclosures
None.
28.
29.
References
1. Reynolds K, Lewis LB, Nolen JD, Kinney LG. Alcohol consumption and
risk of stroke: a meta-analysis. JAMA. 2003;289:579 –588.
2. Palomäki H, Kaste M. Regular light to moderate intake of alcohol and the risk of
ischemic stroke: is there a beneficial effect? Stroke. 1993;24:1828–1832.
3. Sacco RL, Elkind M, Boden-Albada B. The protective effect of moderate
alcohol consumption on ischemic stroke. JAMA. 1999;281:53– 60.
4. Stampfer MJ, Colditz GA, Walter BS, Willett C. A prospective study of
moderate alcohol consumption and the risk of coronary heart disease and
stroke in women. N Engl J Med. 1988;319:267–273.
5. Klatsky AL, Armstrong M, Friedman GD. Alcohol drinking and risk of
hemorrhagic stroke. Neuroepidemiology. 2002;21:115–122.
6. Longstreth WT, Nelson LM, Koepsell TD, Van Belle G. Cigarette smoking,
alcohol use, and subarachnoid hemorrhage. Stroke. 1992;23:1242–1249.
7. Donahue RP, Abbott RD, Reed DM. Alcohol and hemorrhagic stroke: the
Honolulu Heart Program. JAMA. 1986;255:2311–2314.
8. Laatikainen T, Manninen L, Poikolainen K, Vartiainen E. Increased
mortality related to heavy alcohol intake pattern. J Epidemiol Community
Health. 2003;57:379 –384.
9. McElduff P, Dobson AJ. How much and how often: population based
case-control study of alcohol consumption and risk of a major coronary
event. BMJ. 1997;314:1159 –1164.
10. Hillbom M, Kaste M. Does ethanol intoxication promote brain infarction
in young adults? Lancet. 1978;2:1181–1183.
11. Hillbom M, Kaste M. Does alcohol intoxication precipitate aneurysmal
subarachnoid haemorrhage? J Neurol Neurosurg Psychiatry. 1981;44:
523–526.
12. Hansagi H, Romelsjö A, Gerhardsson de Verier M, Andearsson S,
Leifman A. Alcohol consumption and stroke mortality-20 year follow-up
of 15077 men and women. Stroke. 1995;26:1768 –1773.
13. World Health Organization. The World Health Organization MONICA
project (monitoring trends and determinants in cardiovascular disease).
J Clin Epidemiol. 1988;41:105–114.
14. Saunders JB, Aasland OG, Babor TF, De la Fuente JR, Grant M. Development of the alcohol use disorder identification test (AUDIT): WHO
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
collaborative project on early detection of persons with harmful alcohol
consumption. Addiction. 1993;88:791– 804.
Jousilahti P, Rastenyte D, Tuomilehto J. Serum ␥-glutamyl transferase, selfreported alcohol drinking and the risk of stroke. Stroke. 2000;31:1851–1855.
Al-Roomi K, Heller RF, Holland T. The importance of hypertension in
the aetiology of infarctive and haemorrgahic stroke: the Lower Hunter
Stroke Study. Med J Aust. 1992;157:425– 455.
Goldstein LB, Adams R. Primary prevention of ischemic stroke:
guidelines from the American Heart Association/American Stroke Association Stroke Council. Stroke. 2006;37:1583–1633.
Beaglehole R, Jackson R. Alcohol, cardiovascular diseases and all causes
of death: a review of the epidemiological evidence. Drug Alcohol Rev.
1992;11:275–290.
Saunders JB. Alcohol: an important cause of hypertension. BMJ. 1987;
294:1045–1046.
Puddey IB, Beilin LJ, Vandongen R, Rouse IL, Rogers P. Evidence for a
direct effect of alcohol consumption on blood pressure in normotensive
men: a randomized controlled trial. Hypertension. 1985;7:707–713.
Wannamathee G, Shaper AG. Alcohol intake and variations in blood
pressure by day of examinations. J Hum Hypertens. 1991;5:59 – 67.
Clarck LT, Friedman HS. Hypertension associated with alcohol withdrawal: assessment of mechanisms and complications. Alcoholism Clin
Exp Res. 1985;9:125–130.
Juvela S, Hillbom M, Palomäki H. Risk factors for spontaneous intracerebral hemorrhage. Stroke. 1995;26:1558 –1564.
Hillbom M, Haapaniemi H, Juvela S, Palomaki H, Numminen H, Kaste
M. Recent alcohol consumption, cigarette smoking and cerebral
infarction in young adults. Stroke. 1995;26:40 – 45.
McKenna CJ, Codd MB, McCann HA, Sugrue DD. Alcohol consumption
and idiopathic dilated cardiomyopathy: a case-control study. Am Heart J.
1998;135:833– 837.
Demakis JG, Rahimtoola SH, Sutton GC, Gunnar RM. The natural course
of alcoholic cardiomyopathy. Ann Intern Med. 1974;80:293–297.
Caplan LR, Hier DB, Cruz ID. Cerebral embolism in the Michael Reese
Stroke Registry. Stroke. 1983;14:530 –536.
Greenspond AJ, Schaal SF. The ‘holiday heart’: electrophysiologic studies of
alcohol effects in alcoholics. Ann Intern Med. 1983;98:135–139.
Hillbom M, Numminen H, Juvela S. Recent heavy drinking and embolic
stroke. Stroke. 1999;30:2307–2312.
Gruchow HW, Hoffmann RG, Anderson AJ. Effects of drinking patterns on
the relationship between alcohol and coronary occlusion. Atherosclerosis.
1982;43:393–404.
Dale CH. Effect of alcoholism on hemostasis. Hematol Semin Haematal.
1980;17:137–147.
Hillbom M, Kaste M, Rasi V. Can ethanol intoxication affect hemocoagulation
to increase the risk of brain infarction in young adults? Neurology. 1983;33:
381–384.
Fink R, Hutton R. Changes in the blood platelets of alcoholics during
alcohol withdrawal. J Clin Pathol. 1983;36:337–340.
Duffy J, Alanko T. Self-reported consumption measures in sample surveys: a
simulation study af alcohol consumption. J Official Stat. 1992. 1992;8:327–350.
Simpura J. Finnish drinking habits: results from interview surveys held in
1968, 1976 and 1984. Report No. 1987;951-9192-36-0. Helsinki: The
Finnish Foundation for Alcohol Studies. 1987.
Lemmens P, Tan ES, Knibbe RA. Measuring quantity and frequency of
drinking in a general population survey: a comparison of five indices.
J Stud Alcohol. 1992;53:476 – 486.
Bushnell CD. Review, oestrogen and stroke in women: assessment of
risk. Lancet Neurol. 2005;4:743–751.
Lampi C, Marecek S. Migraine and stroke—why do we talk about it? Eur
J Neurol. 2006;13:215–219.
MacClellan LR, Giles W, Cole J. Probable migraine with visual aura and
risk of ischemic stroke: the Stroke Prevention in Young Women Study.
Stroke. 2007;38:2438 –2445.
Sivenius J, Tuomilehto J, Immonen-Räihä P, Kaarisalo M, Sarti C,
Torppa J, Kuulasmaa K, Mähönen M, Lehtonen A, Salomaa V. Continuous 15-year decrease in incidence and mortality of stroke in Finland:
The FINSTROKE Study. Stroke. 2004;35:420 – 425.
Leppälä JM, Virtamo J, Heinonen OP. Validation of stroke diagnosis in
the National Hospital Discharge Register and the Register of Causes of
Death in Finland. Eur J Epidemiol. 1999;15:155–160.
Tolonen H, Salomaa V, Torppa J, Sivenius J, Immonen-Räihä P,
Lehtonen A. The validation of the Finnish Hospital Discharge Register
and Causes of Death Register data on stroke diagnoses. Eur J Cardiovasc
Prev Rehabil. 2007;14:380 –385.
Increased Stroke Risk Is Related to a Binge Drinking Habit
Laura Sundell, Veikko Salomaa, Erkki Vartiainen, Kari Poikolainen and Tiina Laatikainen
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
Stroke. 2008;39:3179-3184; originally published online October 2, 2008;
doi: 10.1161/STROKEAHA.108.520817
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2008 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/39/12/3179
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Stroke is online at:
http://stroke.ahajournals.org//subscriptions/