European Heart Journal (2002) 23, 1655–1663
doi:10.1053/euhj.2002.3235, available online at http://www.idealibrary.com on
Fifteen percent of myocardial infarctions and
coronary revascularizations explained by family
history unrelated to conventional risk factors
The Reykjavik Cohort Study
M. B. Andresdottir1,2, G. Sigurdsson1,2, H. Sigvaldason1, and V. Gudnason1
1
Icelandic Heart Association-Research Institute, Reykjavik, Iceland; 2Department of Internal Medicine,
Landspitali-University Hospital, Reykjavik, Iceland
Aims To examine the relationship between history of
myocardial infarction in first-degree relatives and the risk of
developing coronary heart disease (myocardial infarction or
coronary revascularization).
Methods and Results A total of 9328 males and 10 062
females, randomly selected residents of the Reykjavik area,
aged 33–81 years, were examined in the period from 1967 to
1996 in a prospective cohort study. Cardiovascular risk
assessment was based on characteristics at baseline.
Information on history of myocardial infarction in firstdegree relatives was obtained from a health questionnaire.
Mean follow-up was 18 and 19 years for men and women,
respectively. During follow-up 2700 men and 1070 women
developed coronary heart disease. Compared with subjects
without a family history, the hazard ratio of coronary heart
disease was 1·75 (95% confidence interval, CI, 1·59–1·92)
for men and 1·83 (95% CI, 1·60–2·11) for women, with one
or more first-degree relatives with myocardial infarction.
The risk factor profile was significantly worse in individuals
with a positive family history. After allowance for these risk
factors, the hazard ratio was still highly significant, 1·66
Introduction
Coronary heart disease is the major cause of morbidity
and mortality in most parts of the industrialized world.
This remains true despite a decline in the incidence of
myocardial infarction and cardiovascular mortality over
the past decades, which is believed to be secondary to
the recognition of several risk factors and the public
Revision submitted 29 January 2002, and accepted 30 January
2002.
Correspondence: Margret B. Andresdottir, MD, PhD, Icelandic
Heart Association-Research Institute, Holtasmari 1, 210
Kopavogur, Iceland.
0195-668X/02/$35.00/0
(CI, 1·51–1·82) and 1·64 (CI, 1·43–1·89) for men and
women, respectively. Family history of myocardial infarction was attributed to 15·1% of all cases of coronary heart
disease in men and 16·6% in women, independent of other
known risk factors.
Conclusion Family history of myocardial infarction
increases the risk of developing coronary heart disease in
both men and women and is largely independent of other
classic risk factors. Approximately 15% of all myocardial
infarctions can be attributed to familial factors that have
not been measured in the study or remain to be elucidated.
(Eur Heart J, 2002; 23: 1655–1663, doi:10.1053/euhj.2002.
3235)
2002 The European Society of Cardiology. Published by
Elsevier Science Ltd. All rights reserved.
Key Words: The Reykjavik Study, family history, myocardial infarction, coronary heart disease, risk factors.
See doi: 10.1053/euhj.2002.3295 for the Editorial comment
on this article
awareness of these factors[1,2]. It is evident, however,
that coronary heart disease cannot be explained solely
by the traditionally known risk factors, a fact that
stimulates interest in further research. It has been known
for many years that coronary heart disease aggregates in
families[3,4–12]. However, since most of the well-known
risk factors, such as cholesterol, high blood pressure and
obesity are also known to aggregate in families[9,13–17]
the contribution of positive family history has been
debated, independent of these other risk factors. By
1967 there was speculation that the known degree of
resemblance of familial serum cholesterol and blood
pressure was insufficient to account for the familial
clustering of coronary heart disease[18]. Since then
2002 The European Society of Cardiology. Published by Elsevier Science Ltd. All rights reserved.
1656
M. B. Andresdottir et al.
several studies have addressed the familial occurrence of
classic risk factors as well as the independent contribution of a history of coronary heart disease in close
relatives in predicting future disease. Most have concluded that family history contributes risk independent
of other factors, although the size of this component has
not been elucidated[4–12,19,20]. The majority of these
studies concentrate on individuals who experience heart
attack early in life, e.g. prior to age 55, and many find no
increased risk in women or in the elderly.
We have calculated the risk associated with a reported
family history of myocardial infarction at any age in
first-degree relatives in a population-based prospective
cohort study of 19 390 male and female residents of the
Reykjavik area. The independence of this association
was assessed by allowing for other well-known risk
factors.
Methods
Study participants
The Reykjavik Study is a large, population-based cohort
study, which started in 1967. Men born 1907–1934 and
women born 1908–1935, residing in the Reykjavik area
and a few adjacent communities on 1 December 1966,
were randomly selected for participation. A total of
19 390 persons agreed to take part, which was approximately 70% of all those invited. The study cohort, which
was divided into six groups of men and women, was
investigated at the Heart Preventive Clinic in Reykjavik
during the period 1967–1996. The first examination of
each person occurred in 1967–1993 for men and in
1968–1996 for women. In this analysis, only data from
the first examination of each participant were used, and
the risk of developing coronary heart disease as defined
below was calculated. The follow-up period is from first
attendance until 31 December 1998. The Reykjavik
Study protocol is approved by the Data Protection
Committee of Iceland and the Director General of
Health.
Registration of coronary heart disease
amongst study participants
In this study coronary heart disease end-points were
defined as acute myocardial infarction or sudden cardiac
death as well as the need for coronary revascularization by coronary artery bypass graft (CABG) or
percutaneous transluminal coronary angioplasty
(PTCA).
All episodes of myocardial infarction amongst the
participants of the Reykjavik Study occurring through
31 December 1998 have been registered by the Icelandic
Heart Association, using the WHO’s MONICA criteria
(multinational monitoring of trends and determinants of
cardiovascular disease)[21]. Diagnostic criteria included
symptoms, electrocardiograms, enzyme activities, and
Eur Heart J, Vol. 23, issue 21, November 2002
necropsy findings compatible with definite or possible
myocardial infarction. Registrations included all occurrences of acute myocardial infarction (and sudden
ischaemic cardiac death). A complete registry of all
CABGs and PTCAs ever carried out on Icelanders has
been maintained at the Icelandic Heart Association and
was used in this study to identify end-points. Because of
a computerized national roster with a unique personal
identification number and low emigration figures (0·5%
of all those participating), the follow-up is nearly total.
Risk factors
The following baseline data were recorded for all participants: age, body mass index (BMI), systolic blood
pressure, serum total cholesterol level (mmol . l 1),
serum triglycerides (mmol . l 1), erythrocyte sedimentation rate (ESR, mm . h 1), glucose levels 90 min after
oral glucose load, smoking history (never, former,
current), use of antihypertensive medication (yes/no),
level of education (elementary school, high school,
college or university) and leisure-time physical activity.
Furthermore, symptoms of coronary heart disease of
participants prior to the entry were registered on the
basis of the answers to the Rose chest pain questionnaire[22], review of hospital records, physical examination and the electrocardiogram (ECG). A recognized
myocardial infarction was defined according to the
MONICA criteria and unrecognized myocardial infarction according to ECG changes fulfilling the criteria for
definite myocardial infarction without a history of
symptoms. In this study both categories were collapsed
into one. Angina pectoris was defined according to the
following criteria: the Rose questionnaire criteria for
angina pectoris were fulfilled and there were either
ischaemic ECG changes at rest or positive changes on an
exercise ECG or no ECG changes or symptoms of
angina pectoris according to the clinical judgement
of the examining doctor. Finally, there was a group of
participants not fulfilling the Rose chest pain questionnaire for angina pectoris, having no history or ECG
manifestation of myocardial infarction but manifesting
ST segment and/or T wave changes on a resting ECG
indicative of myocardial ischaemia.
Blood was drawn from the participants in the clinic
before noon, after overnight fasting. A 50-g glucose load
was given in a 250-ml solution to be swallowed in
3–5 min. A capillary blood sample was taken from an
ear lobe during the fasting state and 90 min after
the glucose load. Blood pressure was measured by a
mercury sphygmomanometer in the supine position after
5 min rest and the average of two separate measurements was used for the analysis. Height and weight were
measured and body mass index (BMI) calculated
(weight (kg)/height (m)2). Information on smoking
habits and the use of antihypertensive agents, education
and leisure-time activity was obtained from a health
questionnaire, which participants filled out and returned
before the visit. A detailed description of the questions
Family history and risk of myocardial infarction
concerning leisure-time physical activity has been published elsewhere[23]. In short, the questionnaire included
the age period when leisure time activity was practised,
number of hours per week as well as the type of activity
(vigorous activity or not). Since physical activity after
the age of 40 was shown to be the strongest protective
factor for cardiovascular mortality, irrespective of the
type of activity, we have included this in our analysis.
Participants were asked about any history of acute
myocardial infarction (heart attack) in their father,
mother or siblings (yes/no/do not know). Trained personnel re-evaluated the questionnaires with the participants during the visit. Participants were urged to contact
other family members in order to obtain a more accurate
family history of myocardial infarction. No attempt
was made to verify the family history of myocardial
infarction.
Statistical analysis
Risk factor levels in participants with and without a
family history of myocardial infarction were compared,
using t-test or chi square test where applicable. The
significance level of P<0·05 was chosen. The distributions of triglycerides and ESR were very skewed, and
their logarithmic transforms proved to be far better
predictors for MI than the untransformed values. The
difference in predictive power measured by likelihood
ratio was between 30 and 50 for each gender between
transformed and untransformed values. The transforms
were therefore used (ln(triglycerides) and ln(ESR+1)
because ESR can take the value of zero. Hazard ratios
for developing coronary heart disease were computed by
Cox regression. The risk period was from examination until diagnosis of end-points, death or until
31 December 1998, whichever came first. As the increase
in risk from the answer ‘do not know’ to the question
about family history of myocardial infarction was about
half the risk increase from the answer ‘yes’, a variable
having the values 0, 0·5, 1 for ‘no’, ‘do not know’ and
‘yes’, respectively, was created and used in the subsequent calculations. Hazard ratios associated with a
family history of myocardial infarction were computed
first by adjusting only for age and the year of examination. Thereafter, each risk factor was entered separately, one at the time, and then in a multivariable model,
by controlling for all factors. For categorical variables,
values were weighted by hazard ratios. The risk percentage attributable to family history was calculated. This is
the proportion of all new cases in a given period that are
attributable to a risk factor. For a categorical risk factor
(family history), the attributable risk percentage was
calculated according to the following formula: AR=
(1–1/sum (Pi*Hi))*100, where AR is the attributable risk
(%), Pi is the prevalence (Table 1) of the risk factor
taking value number i and Hi is the corresponding
hazard ratio (Tables 2, 4 and 5). Interaction between
family history and the variables adjusted for was tested
by using the products of the family history variable and
1657
the other variables as predictors. The SPIDA (Statistical
Package for Interactive Data Analysis) software
package was used for the analyses[24].
Results
Of the 9328 male and 10 062 female participants,
1305 men (14%) and 1311 women (13%) had a history
coronary heart disease prior to entry into the study
(Table 1). As expected, more men were diagnosed with
myocardial infarction at entry; however, more women
had an angina diagnosis. A further analysis showed that
this was mainly due to a more frequent fulfilment of the
Rose chest pain questionnaire without ECG changes or
a confirmation of the angina diagnosis by a doctor (264
men, 398 women). Further baseline characteristics of the
participants are shown in Table 1. The females compared favourably with the male participants with respect
to BMI, triglycerides, systolic blood pressure, and smoking habits, factors all significantly higher in men. However, ESR and total cholesterol levels were significantly
higher in women and the use of antihypertensive drugs
more prevalent, compared with men; women had a
lower level of education. Furthermore, women reported
a positive family history of myocardial infarction significantly more often (21·6% and 16·8% of women and men,
respectively), (P<0·001). However, an equal number of
men and women reported not knowing of such history in
the family (20·3% and 19·0%, respectively). The average
follow-up for men was 18·3 years (SD 8·3 years) and for
women 19·2 years (SD 8·1 years).
During follow-up, 2700 of the male participants and
1070 of the females developed a myocardial infarction
(n=1609 males/652 females), sudden death (n=488
males/257 females) or underwent CABG (174 males/66
females) or PTCA (429 males/95 females). Table 2 shows
the hazard ratio of developing coronary heart disease
predicted by the risk factors shown, adjusted only for
age and the year of study. It is noteworthy that the
increase in risk associated with the answer ‘do not know’
regarding family history is approximately half that
obtained for the answer ‘yes’. The mean values of the
selected risk variables according to the reported family
history of myocardial infarction are shown in Table 3.
We have compared these values in participants reporting
a negative history with those reporting a positive one,
whereas the group reporting ‘do not know’ is considered
to be a mixture of both and is not included in this
comparison. Male participants reporting a positive
family history were significantly younger at examination
than those who reported not having such a history.
However, the reverse was seen in women. In men,
cholesterol, triglycerides, glucose tolerance after 90 min
and the use of antihypertensive treatment were inversely
related to family history of myocardial infarction,
whereas the differences observed for BMI, systolic
blood pressure, ESR and smoking habits did not reach
statistical significance. The findings were similar in
women.
Eur Heart J, Vol. 23, issue 21, November 2002
1658
M. B. Andresdottir et al.
Table 1
Baseline characteristics in the study population
Males (n=9328)
Females (n=10 062)
Risk factor (unit)
Mean/%
SD
Mean/%
SD
Age (year)
52·9
9·3
54·1
10·1
Family history of MI (%)
No
Yes
Do not know
62·9
16·8
20·3
BMI (kg . m 2)
Cholesterol (mmol . l 1)
Triglycerides (mmol . l 1)
Ln (triglycerides)*
Systolic blood pressure (mmHg)
Antihypertensive treatment (%)
Glucose tolerance 90 min (mmol . l 1)
ESR (mm . h 1)
Ln (ESR+1)*
Smoking (%)
Never
Former
Current
Education (%)
Elementary school
High school
College
University
Physical activity after the age of 40 (%)
CHD diagnosis (%)
MI
Angina pectoris
ST-T changes
25·8
6·37
1·26
0·12
140
7·4
5·72
7·1
1·71
59·4
21·6
19·0
3·5
1·07
0·27
0·47
19
2·01
8·3
0·89
25·1
6·62
1·07
0·03
138
11·9
5·89
12·3
2·30
21·6
25·1
53·3
44·8
16·0
39·2
35·2
43·7
12·0
9·3
54·3
37·5
6·6
1·6
15·6
17·3
3·5
5·1
5·4
1·3
5·5
6·2
4·3
1·24
0·55
0·43
21
1·84
11·6
0·78
MI=myocardial infarction; BMI=body mass index; ESR=erythrocyte sedimentation rate;
CHD=coronary heart disease.
A significantly higher proportion of participants with
a positive family history had a higher educational level
than those with a negative history, and a significantly
higher proportion of those with a positive family history
reported leisure-time physical activity after the age of 40.
Furthermore, a coronary heart disease diagnosis at
baseline was found significantly more often for those
with a positive family history. To evaluate whether these
observed differences in classic risk factors accounted for
the higher risk of coronary heart disease in participants
with a positive family history, we performed a Cox
regression analysis of the risk of coronary heart disease
among those with a family history of myocardial infarction, compared with those without one, simultaneously
controlling for other risk factors. These results are
compared with the hazard ratio computed by controlling
only for age and year of study in order to examine the
evidence for confounding. In men, the latter hazard
ratio was 1·75 (CI 1·59–1·92), decreasing to 1·66 (CI
1·51–1·82) after allowing for the risk factors shown in
Table 4. The results from the Cox regression are shown
to illustrate the confounding effect of individual risk
factors. The main decrease in hazard ratio was seen
when cholesterol was entered into the model. In women
Eur Heart J, Vol. 23, issue 21, November 2002
the hazard ratio of a positive family history, adjusted for
age and year of study, was 1·83 (CI 1·60–2·11). Again,
the main decrease in the hazard ratio was obtained by
entering cholesterol, and triglycerides had a similar
confounding effect (Table 5). The multivariate hazard
ratio decreased to 1·64 (CI 1·43–1·89), still highly significant. Adjustment for prior coronary heart disease
diagnosis in addition to the above-mentioned factors
reduced the overall hazard ratio in men to 1·51 (95% CI
1·37–1·66) and in women to 1·57 (95% CI 1·36–1·81).
The risk attributable to having a positive family
history was similarly calculated before and after adjusting for the confounding risk factors. For men the
attributable risk was 16·8%, when adjusted only for age
and year of study, and decreased only by 1·7%, to 15·1%
after allowing for other risk factors. In women the
unadjusted, attributable risk of 20·5% decreased to
16·6% after allowing for other risk factors. This
3·9% decrease was not significant. The likelihood ratio
for each risk factor, as well as for all factors combined,
is also shown.
The following significant interactions between family
history and other variables were found. For males, one
SD increase in systolic blood pressure decreased the
Family history and risk of myocardial infarction
1659
Table 2 Hazard ratios (95% confidence intervals) of MItCABGtPTCA predicted
by various risk factors. Adjusted for age and year of study
Men (n=9328)
Women (n=10 062)
Risk factor (unit)
Hazard Ratio
95% CI
Hazard Ratio
95% CI
Age (year)
1·05
1·04–1·06
1·09
1·08–1·10
Family history of MI
No (reference)
Yes
Do not know
1
1·75
1·31
1·59–1·92
1·19–1·45
1
1·84
1·33
1·60–2·11
1·14–1·56
BMI (kg . m 2)
Cholesterol (mmol . l 1)
Ln (triglycerides)
Systolic blood pressure (mmHg)
Antihypertensive treatment
Glucose tolerance 90 min (mmol . l 1)
Ln (ESR+1)*
1·043
1·34
1·75
1·010
1·68
1·06
1·29
1·032–1·055
1·30–1·39
1·61–1·90
1·008–1·012
1·47–1·92
1·04–1·08
1·24–1·35
1·022
1·29
2·50
1·014
1·86
1·08
1·48
1·008–1·036
1·24–1·34
2·18–2·86
1·011–1·017
1·58–2·18
1·06–1·10
1·36–1·61
Smoking
Never (reference)
Former
Current
1
1·24
1·53
1·10–1·39
1·38–1·70
1
1·17
2·17
0·96–1·42
1·90–2·48
Education
Elementary school (reference)
High school
College
University
1
0·95
0·90
0·76
0·87–1·04
0·80–1·03
0·65–0·88
1
0·75
0·51
0·98
0·65–0·86
0·37–0·71
0·59–1·64
Physical activity after the age of 40
0·80
0·71–0·90
0·66
0·54–0·81
CHD diagnosis
None (reference)
Myocardial infarction
Angina pectoris
ST-T changes in ECG
1
4·01
2·26
1·76
3·43–4·69
1·97–2·60
1·52–2·05
1
4·47
1·83
1·80
3·21–6·23
1·48–2·25
1·47–2·19
MI=myocardial infarction; BMI=body mass index; ESR=erythrocyte sedimentation rate;
CHD, coronary heart disease.
Hazard ratio was calculated per unit for continuous variables and relative to the reference value for
categorical variables.
hazard ratio for family history by 12% (P=0·003) and in
females by 15% (P=0·01). An increase in ln(triglycerides) of one SD decreased the hazard ratio for family
history by 18% (P=0·03). Furthermore, a coronary
heart disease diagnosis at entry decreased the hazard
ratio for family history in men by 27% (P=0·004).
Discussion
This is, to our knowledge, the largest prospective cohort
study confirming the independent contribution of family
history to myocardial infarction. Based on approximately 20 000 randomly selected inhabitants of the
Reykjavik area, men and women with a history of
myocardial infarction in first-degree relatives are at 75%
and 83% higher risk, respectively, of coronary heart
disease than those who report a negative family history.
Allowance for potential confounding risk factors did not
alter the risk significantly (66% in men and 64% in
women).
We found that several of the risk factors were significantly worse in individuals with a family history of
myocardial infarction, compared with those without
one. These findings are in line with the preliminary
results of our study of risk factors in the offspring of the
cohort presented here[25]. A number of other studies
have also demonstrated a familial aggregation of risk
factors[9,13–17]. It is not clear, however, how much of this
familial resemblance in risk factors can be attributed to
genetic similarities, and what the role of environmental
factors shared by the families may be. Twin studies
indicate that the resemblance in risk factors is influenced
by both genetic and environmental factors[26,27].
Our results show that although the selected risk
factors are worse in participants with a positive family
history, they can only partly explain the risk associated
with a positive family history of myocardial infarction.
Several studies have addressed the issue of family history
as an independent risk factor. Most have concluded that
family history contributes a risk component that is
independent of other factors[4–12,19,20], i.e. findings similar to ours. However, whereas we find a higher risk of
Eur Heart J, Vol. 23, issue 21, November 2002
1660
M. B. Andresdottir et al.
Table 3 Mean values (SD) of selected variables in the participants of the Reykjavik
Study according to their reported family history of myocardial infarction
Family history of MI
Males
Variable (unit)
Age (years)
BMI (kg . m 2)
Cholesterol (mmol . l 1)
Triglycerides (mmol . l 1)
Glucose tolerance 90 min (mmol . l 1)
Systolic blood pressure (mmHg)
Antihypertensive treatment (%)
ESR (mm . h 1)
Females
Yes
n=1564
No
n=5870
Yes
n=2173
No
n=5978
52·2 (9·3)
25·9 (3·5)
6·49 (1·08)
1·32 (0·76)
5·80 (2·06)
140 (19)
9·6
7·2 (8·2)
52·3 (9·1)
25·7 (3·4)
6·35 (1·07)‡
1·24 (0·72)‡
5·64 (1·95)†
140 (19)
6·0‡
7·0 (8·1)
54·6 (9·9)
25·2 (4·2)
6·72 (1·25)
1·12 (0·60)
5·96 (1·87)
138 (20)
15·0
12·4 (11·1)
53·1 (9·8)‡
25·0 (4·1)
6·57 (1·23)‡
1·03 (0·51)‡
5·82 (1·0)†
137 (21)
9·6‡
12·3 (12·2)
Smoking (%)
Current smoker
Former smoker
53·3
26·5
54·0
23·6†
39·4
16·6
39·0
15·2
Education (%)
High school
College
University
44·1
13·7
12·9
44·1
12·3†
9·7‡
42·2
7·5
2·3
37·2‡
6·7
1·5†
Physical activity after the age of 40 (%)
17·8
15·6*
19·9
16·5†
7·7
8·4
5·2
1·9‡
3·9‡
5·2
2·5
8·2
5·5
0·8‡
3·9‡
6·1
CHD diagnosis (%)
Myocardial infarction
Angina pectoris
ST-T changes in ECG
MI=myocardial infarction; BMI=body mass index; ESR=erythrocyte sedimentation rate;
CHD=coronary heart disease. Individuals answering yes or no to a family history of myocardial
infarction are compared. P-values are *<0·05; †<0·01 and ‡<0·001.
Table 4 Hazard ratio (95% CI), attributable risk and likelihood ratio (chi square)
of MItCABGtPTCA predicted by a family history of myocardial infarction. Always
adjusted for age and year of study. Other risk factors were entered into the model one
at the time and all together. Males (n=9 328)
Hazard
ratio*
95% CI
Attributable
risk %*
Likelihood ratio
(chi square)*
Age and year of study
Total cholesterol
Ln triglycerides
Systolic blood pressure
Antihypertensive medication
Glucose tolerance 90 min
BMI
Ln(ESR+1)
Smoking
Education
Physical activity after the age of 40
1·75
1·69
1·72
1·74
1·73
1·73
1·74
1·73
1·75
1·76
1·75
1·59–1·92
1·54–1·86
1·57–1·89
1·58–1·90
1·57–1·89
1·58–1·90
1·58–1·91
1·58–1·90
1·59–1·92
1·60–1·93
1·60–1·92
16·8
15·7
16·8
16·6
16·3
16·4
16·6
16·4
16·8
17·0
16·8
131·6
117·2
124·2
128·0
125·1
127·3
128·2
127·2
130·8
132·7
132·2
All risk factors above
1·66
1·51–1·82
15·1
107·9
Risk factors
BMI=body mass index; ESR=erythrocyte sedimentation rate.
Hazard ratio and attributable risk calculated per unit risk factor for continuous variables and
compared with the reference for categorical values, as shown in Table 2.
*All results are significant with a P-value of <0·001.
coronary heart disease in both men and women with a
positive family history and independent of the age of
relatives at myocardial infarction, many studies find this
Eur Heart J, Vol. 23, issue 21, November 2002
association only in men[3,7,9,28] or only in individuals
developing a cardiovascular event at a younger age[3,4,7].
The lack of association in women or the elderly in these
Family history and risk of myocardial infarction
1661
Table 5 Hazard ratio (95% CI), attributable risk and likelihood ratio (chi square)
of MItCABGtPTCA predicted by family history of myocardial infarction. Always
adjusted for age and year of study. Other risk factors were entered into the model one
at the time and all together. Females (n=10 062)
Hazard
ratio*
95% CI
Attributable
risk %*
Likelihood ratio
(chi square)*
Age and year of study
Total cholesterol
Ln triglycerides
Systolic blood pressure
Antihypertensive medication
Glucose tolerance 90 min
BMI
Ln(ESR+1)
Smoking
Education
Physical activity after the age of 40
1·83
1·75
1·76
1·80
1·80
1·82
1·83
1·81
1·80
1·86
1·84
1·60–2·11
1·52–2·02
1·53–2·02
1·56–2·07
1·56–2·07
1·59–2·10
1·59–2·10
1·57–2·08
1·57–2·07
1·63–2·14
1·60–2·12
20·5
18·9
19·1
19·9
19·9
20·3
20·5
20·1
19·9
21·1
20·6
69·1
58·3
59·6
64·5
64·2
57·8
58·2
66·2
64·8
71·2
69·5
All risk factors above
1·64
1·43–1·89
16·6
45·7
Risk factors
BMI=body mass index; ESR=erythrocyte sedimentation rate. Hazard ratio and attributable risk
calculated per unit risk factor for continuous variables and compared with the reference for
categorical values, as shown in Table 2. *All results are significant with a P-value of<0·001.
studies may reflect insufficient numbers of those studied
or too few end-points. Furthermore, the different endpoints that are used in the various studies may account
for differences between studies, since the risk factors for
coronary heart disease may differ from risk factors for
myocardial infarction, where thrombosis is involved.
The attributable risk was calculated in order to
quantify the risk associated with family history in our
study cohort. The calculations take into account the
prevalence of the factor studied in the population, in this
case the family history. We found that a positive family
history attributed to one in every seven cases of myocardial infarctions or revascularizations (15%), unrelated to
other risk factors. To our knowledge, no other prospective cohort study has presented similar data. However,
one case control study from Argentina has reported that
family history accounted for 14% of all cases of MI in
men and 26% in women[19].
In our study the multivariate analysis included the
traditional risk factors, i.e. total cholesterol, triglycerides, blood glucose level, systolic blood pressure and
smoking habits. We also included some of the ‘newer’
risk factors, such as erythrocyte sedimentation rate, level
of education and leisure-time physical activity. All of
these factors combined only explained a small part of
the risk associated with a positive family history. HDLcholesterol measurements were started in 1986 and were
thus not available for all participants. We have calculated from our data that at least one-half of the association of low HDL and coronary heart disease can be
explained by high triglycerides, and we therefore assume
that HDL-cholesterol levels in plasma would not influence the effect of the family history any more than the
triglycerides.
Several factors evolving in recent years are shown to
be involved in the pathogenesis of atherosclerosis,
thrombosis and vasoconstriction, which can contribute
to the development of coronary heart disease. These
newer factors include elevated total plasma homocysteine level[29], lipoprotein(a)[30], abnormalities in
haemostatic parameters (e.g. plasma fibrinogen and
plasminogen activator inhibitor type I), inflammatory
markers (CRP) and infections such as Chlamydia
pneumoniae[31]. None of these factors were measured in
the study. Furthermore, several gene polymorphisms
have been associated with the risk of cardiovascular
diseases, although their contribution is still a matter of
controversy and needs to be established as such. These
include a homozygous deletion of the angiotensinconverting enzyme (ACE) gene (DD)[32] and a variation
in the angiotensinogen gene (T235)[33].
Whether the increased risk of myocardial infarction in
those with a family history of myocardial infarction can
be explained by some of these newer potential risk
factors remains to be studied. Alternatively, one or a few
strong independent genes render individuals with a
positive family history prone to the development of
coronary artery disease.
There are two main limitations to our study. First, as
we have commented, HDL-cholesterol, a well-known
risk factor, is not included in the study. It is unlikely,
however, that the inclusion of HDL-cholesterol would
have influenced the risk for family history considerably,
for reasons mentioned above. Second, the reported
family history of myocardial infarction has not been
verified, although all participants were urged to contact
other family members about such a history. However,
some 20% reported not knowing about the family history. The compensating advantage is, however, that this
method corresponds to how family history is assessed
clinically. As shown in Table 2, the increase in risk of
developing myocardial infarction for this latter group is
Eur Heart J, Vol. 23, issue 21, November 2002
1662
M. B. Andresdottir et al.
very close to half the increase in risk for those who
answered ‘yes’ to a family history. This indicates that the
group reporting ‘do not know’ includes a fairly evenly
distributed cohort of individuals with and without a
family history and allows us to include this group in the
analysis. Studies where the reported family history has
been validated find a sensitivity of 59–77% and specificity of 85–96%[10,28,34]. The relatively low sensitivity
indicates that a substantial proportion of relatives with
myocardial infarction was under-reported, and the
results thus biased towards an underestimation of the
effect of a positive family history.
It has been suggested that persons with a positive
family history may be more susceptible to the detrimental effect of other risk factors. The data are, however,
conflicting. We found no positive interaction of the risk
factors with family history of MI, which would be
expected if persons with a family history were more
sensitive to the risk factors than those without it. Myers
et al.[28] concluded similarly, however, Roncaglioni
et al.[20] and Ciruzzi et al.[19] found an almost multiplicative effect on risk ratio for several risk variables, and
Barrett-Connor et al.[7] found an increased susceptibility
to the detrimental effect of smoking. The issue thus
remains unsettled.
In conclusion, about 15% (one out of seven) of
myocardial infarction and coronary revascularizations
are explained by family history that cannot be related to
conventional risk factors. Our study thus indicates the
presence of some major (one or more) familial factors
that are still unknown or not measured in the study.
We gratefully acknowledge the work of I. Gudmundsdottir, I.
Stefansdottir and L. Jonsdottir, who have recorded the coronary heart disease end-points in the entire study population. M.
Andresdottir has been supported by a grant from the Icelandic
Research Council.
References
[1] Sigfusson N, Sigvaldason H, Steingrimsdottir L et al. Decline
in ischaemic heart disease in Iceland and change in risk factor
levels. BMJ 1991; 302: 1371–5.
[2] Tunstall-Pedoe H, Vanuzzo D, Hobbs M et al. Estimation of
contribution of changes in coronary care to improving survival, event rates, and coronary heart disease mortality across
the WHO MONICA Project populations. The Lancet 2000;
355: 688–700.
[3] Phillips RL, Lilienfeld AM, Diamond EL, Kagan A.
Frequency of coronary heart disease and cerebrovascular
accidents in parents and sons of coronary heart disease index
cases and controls. Am J Epidemiol 1974; 100: 87–100.
[4] Sholtz RI, Rosenman RH, Brand RJ. The relationship of
reported parental history to the incidence of coronary heart
disease in the Western Collaborative Group Study. Am J
Epidemiol 1975; 102: 350–6.
[5] Nora JJ, Lortscher RH, Spangler RD, Nora AH, Kimberling
WJ. Genetic- epidemiologic study of early-onset ischemic
heart disease. Circulation 1980; 61: 503–8.
[6] Ten Kate LP, Boman H, Daiger SP, Motulsky AG. Familial
aggregation of coronary heart disease and its relation to
known genetic risk factors. Am J Cardiol 1982; 50: 945–53.
[7] Barrett-Connor E, Khaw K-T. Family history of heart attack
as an independent predictor of death due to cardiovascular
disease. Circulation 1984; 69: 1065–9.
Eur Heart J, Vol. 23, issue 21, November 2002
[8] Friedlander Y, Kark JD, Stein Y. Family history of myocardial infarction as an independent risk factor for coronary
heart disease. Br Heart J 1985; 53: 382–7.
[9] Khaw K-T, Barrett-Connor E. Family history of heart attack:
a modifiable risk factor? Circulation 1986; 74: 239–44.
[10] Hopkins PN, Williams RR, Kuida H et al. Family history as
an independent risk factor for incident coronary artery disease
in a high-risk cohort in Utah. Am J Cardiol 1988; 62: 703–7.
[11] Schildkraut JM, Myers RH, Cupples LA, Kiely DK, Kannel
WB. Coronary risk associated with age and sex of parental
heart disease in the Framingham Study. Am J Cardiol 1989;
64: 555–9.
[12] Colditz GA, Rimm EB, Giovannucci E, Stampfer MJ, Rosner
B, Willett WC. A prospective study of parental history of
myocardial infarction and coronary artery disease in men. Am
J Cardiol 1991; 67: 933–8.
[13] Bao W, Srinivasan SR, Valdez R, Greenlund KJ, Wattigney
WA, Berenson GS. Longitudinal changes in cardiovascular
risk from childhood to young adulthood in offspring of
parents with coronary artery disease The Bogalusa Heart
Study. JAMA 1997; 278: 1749–54.
[14] Shear CL, Webber LS, Freedman DS, Srinivasan SR,
Berenson GS. The relationship between parental history of
vascular disease and cardiovascular disease risk factors in
children: the Bogalusa Heart Study. Am J Epidemiol 1985;
122: 762–71.
[15] Moll PP, Sing CF, Weidman WH et al. Total cholesterol and
lipoproteins in school children: prediction of coronary heart
disease in adult relatives. Circulation 1983; 67: 127–34.
[16] Laskarzewski P, Morrison JA, Horvitz R et al. The relationship of parental history of myocardial infarction, hypertension, diabetes and stroke to coronary heart disease risk factors
in their adult progeny. Am J Epidemiol 1981; 113: 290–306.
[17] Glueck CJ, Laskarzewski PM, Suchindran CM et al. Progeny’s lipid and lipoprotein levels by parental mortality. The
Lipid Research Clinics Program Prevalence Study. Circulation
1986; 73: I51–I61.
[18] Epstein FH. Risk factors in coronary heart disease —
environmental and hereditary influences. Isr J Med Sci 1967;
3: 594–607.
[19] Ciruzzi M, Schargrodsky H, Rozlosnik J et al. Frequency of
family history of acute myocardial infarction in patients with
acute myocardial infarction. Argentine FRICAS (Factores de
Riesgo Coronario en America del Sur) Investigators. Am J
Cardiol 1997; 80: 122–7.
[20] Roncaglioni MC, Santoro L, D’Avanzo B et al. Role of family
history in patients with myocardial infarction. An Italian
case-control study. GISSI-EFRIM Investigators. Circulation
1992; 85: 2065–72.
[21] Tunstall-Pedoe H. Diagnosis, measurement and surveillance
of coronary events. Int J Epidemiol 1989; 18: S169–73.
[22] Rose G, McCartney P, Reid DD. Self-administration of
questionnaire on chest pain and intermittent claudication. Br J
Soc Med 1977; 31: 42–8.
[23] Agnarsson U, Thorgeirsson G, Sigvaldason H, Sigfusson N.
Effects of lesure-time physical activity and ventilatory
function on risk for stroke in men: The Reykjavik Study. Ann
Intern Med 1999; 130: 987–90.
[24] Gebski V, Leung O, McNeil D, Lunn D. SPIDA User’s
Manual, Version 6. Macquarie University, University of NSW
Australia, 1992.
[25] Andresdottir MB, Sigurdsson G, Sigvaldason H, Agnarsson
U, Gudnason V. Relation of risk factor levels in adult
offspring to parental history of myocardial infarction: The
Reykjavik Offspring Study (Abstr). Atherosclerosis 2000; 151:
127.
[26] Heller DA, de Faire U, Pedersen NL, Dahlen G, McClearn
GE. Genetic and environmental influences on serum lipid
levels in twins. N Engl J Med 1993; 328: 1150–6.
[27] Hong Y, de Faire U, Heller DA, McClearn GE, Pedersen NL.
Genetic and environmental influences on blood pressure in
elderly twins. Hypertension 1994; 24: 663–70.
Family history and risk of myocardial infarction
[28] Myers RH, Kiely DK, Cupples LA, Kannel WBM. Parental
history is an independent risk factor for coronary artery
disease: the Framingham Study. Am Heart J 1990; 120:
963–9.
[29] Eikelboom JW, Lonn E, Genest J, Jr., Hankey G, Yusuf S.
Homocyst(e)ine and cardiovascular disease: a critical review
of the epidemiologic evidence. Ann Intern Med 1999; 131:
363–75.
[30] Seman LJ, DeLuca C, Jenner JL et al. Lipoprotein(a)cholesterol and coronary heart disease in the Framingham
Heart Study. Clinical Chemistry 1999; 45: 1039–46.
1663
[31] Braunwald E. Shattuck lecture — cardiovascular medicine
at the turn of the millennium: triumphs, concerns, and
opportunities. N Engl J Med 1997; 337: 1360–9.
[32] Cambien F, Poirier O, Lecerf L et al. Deletion polymorphism
in the gene for angiotensin-converting enzyme is a potent risk
factor for myocardial infarction. Nature 1992; 359: 641–4.
[33] Katsuya T, Koike G, Yee TW et al. Association of angiotensinogen gene T235 variant with increased risk of coronary
heart disease. The Lancet 1995; 345: 1600–3.
[34] Kee F, Tiret L, Robo JY et al. Reliability of reported family
history of myocardial infarction. BMJ 1993; 307: 1528–30.
Eur Heart J, Vol. 23, issue 21, November 2002