Parental History of Cardiovascular Disease and Risk of Stroke
A Prospective Follow-up of 14 371 Middle-aged Men and Women in Finland
Background and Purpose Studies on risk factors for stroke have been less intensive than those for coronary disease. Only a few studies have addressed the question of the role of heredity in the occurrence of stroke. We analyzed whether a positive parental history of cardiovascular disease predicts the risk of stroke independently from other risk factors and whether the role of parental history varies by age and stroke subtypes.
Methods This study was a prospective follow-up of 14 371 middle-aged men and women. A positive parental history of cardiovascular disease was defined as either stroke or coronary disease before the age of 60 years. The end point of the follow-up was an incident case of stroke. Multivariate analyses were performed with the Cox proportional hazards model.
Results The risk ratio of stroke after multifactorial adjustment (age, smoking, blood pressure, cholesterol, diabetes, and education) associated with a positive parental history of stroke was 1.89 (P=.004) in men and 1.80 (P=.007) in women. The association between parental history of stroke and the risk of stroke was stronger among subjects aged 25 to 49 years than among older subjects. Parental history of coronary disease was not associated with the risk of stroke in men, but in women it had a borderline significant association with the risk of ischemic stroke.
Conclusions A positive parental history of stroke predicted the risk of stroke independently from the other risk factors.
Although the role of high blood pressure in the causation of stroke has been known for a century, a systematic search for stroke risk factors began much later. Studies of risk factors for stroke have been less intensive than those for coronary heart disease, and some inconsistency still exists regarding the risk factors for stroke.1 Although most of the interest has focused on environmental influences, since these can be modified and controlled, it was recognized that stroke also has a hereditary basis that may interact with the environmental factors.2 3
Despite a great number of studies on familial clustering of coronary disease,4 5 relatively few studies have addressed the question of the role of heredity and the importance of family history in the occurrence of stroke. Results from these few studies are also somewhat inconclusive.6 7 8 9 10 11 12 13 14 Most of the previous studies have used case-control or cross-sectional design, and they often suffered from a small number of cases, resulting in a lack of statistical power to confirm or to deny the importance of family history on the risk of stroke. Furthermore, very few studies of the role of family history in the risk of stroke have included women.
The aim of the present study was to determine whether a positive parental history of cardiovascular disease, either stroke or coronary disease, predicts the risk of stroke among middle-aged men and women independently from smoking, serum cholesterol, systolic blood pressure, diabetes, and socioeconomic status in a large population-based cohort. We also analyzed whether the association between positive parental history varies between the different subtypes of stroke: ischemic stroke, SAH, and ICH.
Subjects and Methods
Participants and Surveys
Cardiovascular risk factor surveys were performed in 1982 and 1987 in the eastern Finnish provinces of North Karelia and Kuopio and in the southwestern region of Turku-Loimaa.15 16 In both years, the random sample included the group aged 25 to 64 years and was stratified such that in all three areas at least 250 subjects were chosen from each sex and 10-year age group according to the international WHO MONICA Project protocol.16 The 1982 and 1987 cohorts are combined in the analyses. The samples included 9789 men and 9538 women. The participation rate was 79% among men and 85% among women. The 168 subjects who participated in both 1982 and 1987 surveys were included in the first survey cohort only. Of 15 658 participants, 158 were excluded from this study because of a history of stroke before the risk factor survey. Another 1129 were excluded because of incomplete data on one or more risk factors. Thus, 6979 men and 7392 women are included in the present study.
Risk Factor Assessment
The survey methods, described elsewhere in detail, followed the WHO MONICA protocol.15 16 The surveys included a self-administered questionnaire, which was sent to the participants in advance and included questions about medical history, health behavior, socioeconomic status, and family history. At the study site, specially trained nurses measured blood pressure using a standardized protocol. Blood pressure was measured from the right arm of the subject with the use of the standard manometer. A venous blood specimen was taken after the blood pressure measurement. Serum total cholesterol was determined by an enzymatic assay method (CHOD-PAP, Boehringer Mannheim). All samples were analyzed in the same laboratory.
Smoking was assessed in the surveys with a set of standardized questions in a self-administered questionnaire. In our present analyses those ex-smokers who had gone without smoking for at least 6 months were considered nonsmokers, and those who had stopped smoking during the last 6 months were considered smokers. Occurrence of diabetes was asked in the questionnaire. Self-reporting included all types of diabetes and both diet- and drug-treated cases. The data were complemented by information from the Social Insurance Institution’s register on persons receiving free-of-charge medication for diabetes. The records covered the period of 1969 to 1989 and included only drug-treated cases. The length of formal education, obtained from the questionnaire, was used as the indicator of socioeconomic status. To adjust for the changes in educational system over time, the number of school years was divided into tertiles separately for each birth year, and the birth cohort–specific distribution was used as the indicator of socioeconomic status. Data on parental history of cardiovascular disease were obtained by the questionnaire. Positive parental history of stroke was defined as fatal or nonfatal stroke before the age of 60 years and positive parental history of coronary disease as fatal or nonfatal myocardial infarction or angina pectoris before the age of 60 years (Table 1⇓). Missing answers to the question about parental history (≈6%) were considered to indicate negative parental histories.
Mortality data were obtained from the Central Statistical Office of Finland. Data on nonfatal stroke events were received from the National Hospital Discharge Register. Mortality data and hospital discharge register data were linked to the risk factor data by means of the identification numbers assigned to all residents of Finland. The linkage data covered all deaths and hospital discharges from cardiovascular causes through the end of 1994. The 8th revision of the ICD was used in Finland from 1969 to 1986; the 9th revision was used beginning in 1987.
The end point during the follow-up was an incident case of stroke, defined as either the first nonfatal cerebrovascular event or stroke death without a history of nonfatal attack. ICD codes 430 to 438 were classified as stroke events. In the analysis by stroke subtype, ICD code 430 was classified as SAH, ICD codes 431 and 432 (from 1987) as ICH, and ICD codes 432 (before 1987) and 433 to 436 as ischemic strokes. Subjects with ICD codes 437 and 438, as well as subjects with more than one type-specific stroke diagnosis, were excluded from the type-specific analyses. During the follow-up, 149 896 person-years were accumulated. The number of stroke events during the follow-up was 249 among men and 204 among women.
ANOVA was used to test the difference in risk factors between subjects with and without a positive parental history of stroke and coronary disease. Age was used as a covariate in the analyses. Multivariate analyses were performed with the Cox proportional hazards model.17 RRs and their CIs are based on this model.
First we analyzed only age- and study year–adjusted RRs of stroke associated with parental history of stroke and coronary disease. To assess the extent to which the risk associated with a positive parental history may be mediated through the known cardiovascular risk factors, a further adjustment was made for smoking, serum total cholesterol, systolic blood pressure, diabetes, and relative length of education. Analyses were done including all incident cases of stroke and also separately for the following subtypes: ischemic stroke, SAH, and ICH. In the type-specific analysis, the comparisons were done between subjects with that specific type of stroke and those without (any type of) stroke.
To assess whether the risk of stroke depends on the type of parental history, the histories were classified into three categories: family history in at least one parent, family history in the father only, and family history in the mother only. To analyze whether the effect of positive parental history on the risk of stroke depends on age, a stratified analysis was done in age groups 25 to 49 years and 50 to 64 years. Interactions between parental history and the other risk factors were analyzed by adding all first-level interactions between parental history and smoking, cholesterol, systolic blood pressure, diabetes, and relative length of education, one at time, into the model containing all main variables. Analyses were done with SAS statistical programs.18
The levels of major cardiovascular risk factors were quite similar among subjects with and without a positive parental history of stroke (Table 2⇓). Only systolic blood pressure was slightly higher among women who had a positive parental history of stroke than among women who had a negative parental history. With regard to parental history of coronary disease, mean cholesterol and systolic blood pressure were higher and diabetes was more common among men who had a positive parental history than among men who had a negative parental history. Among women, cholesterol was higher and relative length of education was shorter among subjects who had a positive parental history of coronary disease than among women who had a negative parental history. Prevalence of smoking did not differ between subjects with and without a positive parental history of coronary disease.
Age-adjusted incidence of stroke among subjects with and without a positive parental history of stroke was 663 and 366 per 100 000 person-years in men and 463 and 251 per 100 000 person-years in women, respectively. The incidence of stroke among subjects with and without a positive parental history of coronary disease was 421 and 369 per 100 000 person-years in men and 296 and 255 per 100 000 person-years in women, respectively.
The age- and study year–adjusted RR of stroke associated with a positive parental history of stroke was 1.90 (P=.003) among men and 1.73 (P=.012) among women (Table 3⇓). Adjustment for the other risk factors changed the RRs only slightly, to 1.89 (P=.004) among men and 1.80 (P=.007) among women. In the analysis by stroke subtype, the RRs after multifactorial adjustment for men and women were 1.51 (P=.136) and 1.79 (P=.025) for ischemic stroke, 5.01 (P=.012) and 2.77 (P=.109) for SAH, and 3.86 (P=.014) and 0.75 (P=.782) for ICH, respectively.
After multifactorial adjustment, the RRs of any type of stroke associated with positive parental history of coronary disease were 1.10 (P=.506) among men and 1.21 (P=.226) among women (Table 3⇑). The RRs for ischemic stroke were 1.10 (P=.574) and 1.41 (P=.055), respectively. Positive parental history of coronary disease was not associated with ICH and SAH in either sex.
In the analysis by type of parental history, RRs of stroke associated with a positive paternal history of stroke after multifactorial adjustment were 2.17 (P=.012) among men and 2.15 (P=.014) among women (Table 4⇓). The respective RRs associated with a positive maternal history of stroke were 1.83 (P=.042) and 1.67 (P=.075). In the analysis by age group, RRs of stroke associated with positive parental history of stroke after multifactorial adjustment were 2.82 (P=.011) among men aged 25 to 49 years and 1.65 (P=.054) among men aged 50 to 64 years (interaction between age group and parental history, P=.240) (Table 5⇓). Among women the respective RRs were 2.63 (P=.030) and 1.60 (P=.062) (interaction between age group and parental history, P=.155). None of the interactions between parental history of stroke and the other risk factors were statistically significant.
These results show that a positive parental history of stroke is associated with increased risk of stroke and that the risk is independent of the known major cardiovascular risk factors and socioeconomic status. The risk associated with a positive parental history can be due to either genetic factors or familial shared environmental factors. Part of the risk may be mediated through other risk factors, such as smoking, high blood pressure, high cholesterol, diabetes, and low socioeconomic status. Among our study cohorts, however, the levels of other risk factors did not differ much between subjects with and without a parental history of stroke.
Our data on parental history did not include the stroke type. We can assume, however, that most of the parents had had an ischemic stroke. Therefore, our finding of a strong association between parental history of stroke and risk of SAH and ICH in men is somewhat surprising. The association may be due to a very strong familial component of the risk of SAH and ICH, or it may demonstrate that atherosclerosis in cerebral arteries increases not only the risk of cerebral infarction but also the risk of bleeding.
Positive parental history of coronary disease was associated with the risk of ischemic stroke among women but not men. Also among women, the RR of ischemic stroke associated with parental history of coronary disease was lower then the RR associated with parental history of stroke. Thus, the familial component of the risk of ischemic stroke may be partly common with atherosclerosis and thrombosis of arteries in general but partly specific for cerebral arteries only.
The risk of stroke tended to be higher among people with a paternal (as opposed to maternal) history of stroke. This may demonstrate a real difference in stroke risk associated with paternal and maternal family histories, but it may also demonstrate differences in the validity of reported parental histories. Because the risk of stroke, particularly ischemic stroke, in middle-aged people is higher among men than women,19 and because our study only considered parental histories before the age of 60 years, it is possible that the reported parental history indicated the hereditary component of stroke more specifically among men than women.
The relative risk of stroke associated with a positive parental history of stroke tended to be higher among subjects aged 25 to 49 years than among those aged 50 to 64 years. The absolute risk of cardiovascular disease is low among young subjects, and it increases with age, partly because of aging itself but also because of exposure to several environmental risk factors, such as smoking and an unhealthy diet. The risk of disease associated with the hereditary factors does not change, but it is not completely independent of age either. Even though the relative risk associated with a positive parental history may be higher among younger subjects, the absolute risk attributed to a positive parental history is higher among older subjects because of the larger number of events among them.
In the Copenhagen City Heart Study, Boysen et al11 followed 13 088 men and women who were at least 35 years of age and free of stroke at entry for 5 years. After adjustment for other risk factors for stroke, a positive family history of stroke was not a significant predictor of stroke in this study. In contrast to our results, results from the Framingham Offspring Study demonstrated an association between parental history of coronary disease and cerebrovascular disease.13 Similar to our study, the Framingham Offspring Study showed a higher risk of cerebrovascular disease associated with paternal than with maternal history of stroke, even though the statistical power of the study was weak because of a relatively young cohort and a small number of cases. The predictive value of positive maternal, but not paternal, history of stroke on the incidence of stroke among 789 men aged 54 years at baseline was found in a Swedish study.10 The Rancho Bernardo Study is the only other prospective study that thus far has evaluated the importance of family history of stroke on the risk of stroke among men and women separately.8 This study showed that a positive family history of stroke in any first-degree relative was an independent predictor of coronary mortality among men aged 50 to 64 years and stroke mortality among women aged 50 to 79 years at baseline. In a univariate analysis, Graffagnino et al14 demonstrated that a significantly greater number of stroke patients had a positive family history of stroke and heart disease compared with age- and sex-matched community control subjects. After multifactorial adjustment, however, family history of cardiovascular disease was no longer an independent risk factor for stroke, although it had a significant positive predictive value for hypertension, ischemic heart disease, and dyslipidemias.
These somewhat inconsistent results of studies regarding the role of parental history of cardiovascular disease in the risk of stroke can reflect real differences between populations, but they also may be due to the methodological differences between different studies. Showing that a positive family history is an independent risk factor for stroke requires the identification of positive versus negative family history, identification of stroke events during the follow-up, assessment of other stroke risk factors, and control for the covariation of other risk factors and family history in the analyses.
Our study based the identification of the family history status on the data obtained from the self-reported questionnaires. Because this information was not verified, the accuracy of recollection may be suspect and likely to result in some degree of misclassification. There are no studies concerning the reliability of reporting family history of stroke, but in a myocardial infarction study the specificity of reporting a positive family history of myocardial infarction was high (≈97%), although the sensitivity was lower (≈67%).20 Thus, there may have been a bias whereby some subjects with a positive family history were falsely classified as having a negative family history. Such a misclassification may have resulted in an underestimation of the true risk associated with a positive parental history. It is possible that the reliability of reporting may vary between age groups. Because of the greater time elapsed since the parental event, the reporting may be less reliable in the older than the younger subjects. Therefore, it may explain part of the difference in stroke risk associated with a positive family history between age groups.
Stroke diagnoses were not separately validated in our study, but we can provide indirect evidence of the quality of the data. Community-based stroke registers have been operating in the North Karelia province and the Turku-Loimaa area since 1982 and in the Kuopio province since 1983. A comparison of our incidence data with the stroke register data of the 1980s showed that the overall stroke incidence in the 1980s is in relatively good agreement with our data and that the distribution of events among sexes and between different types of stroke are fairly similar in both sources of data.19 21 In large cohort studies, however, misclassification of outcome parameters is difficult to avoid completely. The effect of this possible misclassification is that the true association between positive parental history and stroke would be somewhat stronger than that shown in our study.
Among our study cohorts, the levels of major known cardiovascular risk factors were fairly similar among subjects with and without a positive family history of stroke. Other factors may also be associated with the risk of stroke but were not included in our analyses, such as hemostatic factors, high levels of lipoprotein(a), low levels of high-density lipoprotein, and genetic variation of angiotensin-converting enzyme.22 23 24 25 26 These factors and their possible familial occurrence may explain part of the risk associated with positive parental history.
The practical implication of positive family history as a risk factor for stroke is somewhat controversial. In healthcare practice, a positive family history of stroke can be used as an argument in individual counseling to convince people at high risk of stroke to reduce their levels of other cardiovascular risk factors by changing their lifestyle or to start preventive treatments such as acetylsalicylic acid. In the future, genetic research may also be able to identify specific subtypes of stroke with particularly high familial risk, which could be used for development of targeted intervention methods in the prevention of stroke. The practical importance of family history for stroke prevention at the population level, however, may be limited. The genetic or familial effects most probably operate through several pathological mechanisms, which are likely to be less important than the classic major risk factors: high blood pressure, smoking, diabetes, and high cholesterol. Also, whatever the mechanisms are, the possibilities of modifying them, at least in the near future, may be limited. Therefore, stroke prevention even among high-risk individuals is likely to be based on management of the major known modifiable risk factors. Moreover, in populations in which the levels of known major risk factors are high and in which stroke is relatively common, preventive activities should be targeted to the whole community and not only to high-risk individuals.27
In conclusion, our results support the hypothesis that positive parental history is an independent risk factor for stroke. This finding provides additional support for the suggested role of the genetic component in the etiology of stroke, which may be further clarified by future genetic research. The practical implications of positive family history in stroke prevention will also be seen in the future.
Selected Abbreviations and Acronyms
|ICD||=||International Classification of Diseases, Injuries, and Causes of Death|
|MONICA||=||Monitoring Trends and Determinants in Cardiovascular Disease|
|WHO||=||World Health Organization|
Reprint requests to Dr Pekka Jousilahti, Department of Epidemiology and Health Promotion, National Public Health Institute, Mannerheimintie 166, FIN-00300 Helsinki, Finland.
- Received January 31, 1997.
- Revision received April 2, 1997.
- Accepted April 21, 1997.
- Copyright © 1997 by American Heart Association
Dyken ML, Wolf PA, Barnett HJM, Bergan JJ, Hass WK, Kannel WB, Kuller L, Kurtzke JF, Sundt TM. Risk factors for stroke: a statement by the Subcommittee on Risk Factors and Stroke of the Stroke Council. Stroke. 1984;15:1105-1111.
Goldbourt U, Neufeld HN. Genetic aspects of arteriosclerosis. Arteriosclerosis. 1986;6:357-377.
Motulsky AG, Burke W. Genetics of hypertension: an overview. In: Berg K, Bulyzhenkov V, Christen Y, Corvol P, eds. Genetic Approaches to Coronary Heart Disease and Hypertension. Berlin, Germany: Springer-Verlag; 1991:1-7.
Perkins KE. Family history of coronary heart disease: is it an independent risk factor? Am J Epidemiol. 1986;124:182-194.
Marshall J. Familial incidence of cerebral hemorrhage. Stroke. 1973;4:38-41.
Herman B, Schmitz PIM, Leyten ACM, van Lujik JH, Frenken CWGM, Op de Coul AAW, Schulte BPM. Multivariate logistic analysis of risk factors for stroke in Tilburg, the Netherlands. Am J Epidemiol. 1983;118:514-525.
Khaw KT, Barrett-Connor CE. Family history of stroke as an independent predictor of ischemic heart disease in men and stroke in women. Am J Epidemiol. 1986;123:59-66.
Diaz JF, Hachinski VC, Pederson LL, Donald A. Aggregation of multiple risk factors for stroke in siblings of patients with brain infarction and transient ischemic attacks. Stroke. 1986;17:1239-1242.
Boysen G, Jorgen N, Appleyard M, Sorensen PS, Boas J, Somnier F, Jensen G, Schnor P. Stroke incidence and risk factors for stroke in Copenhagen, Denmark. Stroke. 1988;19:1345-1353.
Brass LM, Shaker LA. Family history in patients with transient ischemic attacks. Stroke. 1991;22:837-841.
Kiely DK, Wolf PA, Cupples A, Beiser AS, Myers RH. Familial aggregation of stroke: the Framingham Study. Stroke. 1993;24:1366-1371.
Graffagnino C, Gasecki AP, Doig GS, Hachinski C. The importance of family history in cerebrovascular disease. Stroke. 1994;25:1599-1604.
Vartiainen E, Puska P, Jousilahti P, Korhonen HJ, Tuomilehto J. Twenty year trends in coronary risk factors in North Karelia and in other areas in Finland. Int J Epidemiol. 1994;23:495-504.
Cox DR. Regression models and life tables. J R Stat Soc B. 1972;34:187-220.
SAS Institute, Inc. SAS/STAT User’s Guide, Version 6. 4th ed, vol 1-2. Cary, NC: SAS Institute, Inc; 1990.
Tuomilehto J, Sarti C, Narva EV, Salmi K, Sivenius J, Kaarsalo E, Salomaa V, Torppa J. The FINMONICA Stroke Register: community-based stroke registration and analysis of stroke incidence in Finland, 1983-1985. Am J Epidemiol. 1992;135:1259-1270.
Kee F, Tiret L, Robo JY, Nicaud V, McCrum E, Evans A, Cambien F. Reliability of reported family history of myocardial infarction. BMJ. 1993;307:1528-1530.
Sarti C, Tuomilehto J, Sivenius J, Kaarsalo E, Narva EV, Salmi K, Torppa J, Salomaa V. Declining trends in incidence, case-fatality and mortality of stroke in three geographic areas of Finland during 1983-1989: results from the FINMONICA Stroke Register. J Clin Epidemiol. 1994;47:1259-1269.
Shintani S, Kikuchi S, Hamaguchi H, Shiigai T. High serum lipoprotein(a) levels are an independent risk for cerebral infarction. Stroke. 1993;24:965-969.
Lindenstrøm E, Boysen G, Nyboe J. Influence of total cholesterol, high density lipoprotein cholesterol, and triglycerides on risk of cerebrovascular disease: the Copenhagen City Heart Study. BMJ. 1994;309:11-15.
Vartiainen E, Sarti C, Tuomilehto J, Kuulasmaa K. Do changes in cardiovascular risk factors explain changes in mortality from stroke in Finland? BMJ. 1995;310:901-904.