Physical activity and other risk factors in male twin ... - Atherosclerosis

Atherosclerosis 150 (2000) 193 – 200 www.elsevier.com/locate/atherosclerosis

Physical activity and other risk factors in male twin-pairs discordant for coronary heart disease Jaakko Kaprio a,*, Urho M. Kujala b, Markku Koskenvuo c, Seppo Sarna a b

a Department of Public Health, Uni6ersity of Helsinki, P.O. Box 41, Mannerheimintie 172, FIN-00014 Helsinki, Finland Unit for Sports and Exercise Medicine, Institute of Biomedicine, Uni6ersity of Helsinki, Mannerheimintie 17, FIN-00250 Helsinki, Finland c Department of Public Health, Uni6ersity of Turku, FIN-20520 Turku, Finland

Received 29 March 1999; received in revised form 6 August 1999; accepted 8 September 1999

Abstract Background: Twin-pairs discordant for coronary heart disease (CHD) can be used to examine the possible contribution of genetic and other familial factors to the relationship between CHD risk factors, in particular physical inactivity at leisure, and CHD. Methods and results: In 1975 at baseline, 8205 men aged 25 – 69 years of the Finnish Twin Cohort without overt CHD responded to a questionnaire. The outcome measures were hospitalization for CHD or death from CHD before age 70 between 1977 to 1995 based on reliable nationwide registries. Among all men, the age-adjusted relative risk of CHD was 0.52 (95% confidence interval (CI) 0.39, 0.71) in men participating in conditioning exercise compared to sedentary men, based on their questionnaire responses. The age, and covariate (body-mass index, smoking, hypertension, diabetes) adjusted-relative risk was 0.68 (CI 0.50, 0.92). Among the twin-pairs discordant for CHD during follow-up (N =311), the odds ratio (OR) of having CHD was 0.54 (CI 0.27, 1.09) for conditioning exercisers compared to sedentary men. The point estimates from discordant pairs analyses were of the same magnitude as obtained from the analyses of individuals, though statistically not significant due to smaller samples. In contrast, smoking (OR 3.21) and diabetes (OR 10.0) at baseline were also significant predictors of CHD. Conclusions: By studying twins we were able to examine genetic and other familial selection bias and found further supporting evidence that leisure conditioning physical exercise compared to sedentariness helps prevent CHD in men, but smoking was a stronger risk factor. © 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Coronary disease; Diabetes mellitus; Hypertension; Exercise; Genes; Smoking

1. Introduction A beneficial association between physical activity and future coronary heart disease (CHD) morbidity/mortality has been established [1 – 6]. Recently, to discount the hypothesis that genetic or other confounding factors explain this association, the effects of longitudinal changes in physical activity and fitness have been studied [7,8]. Paffenbarger et al. [7] found that increased physical activity and Blair et al. [8] that increased physical fitness were associated with lower mortality from CHD.

* Corresponding author. Tel.: +358-0-19127595; fax: + 358-019127600. E-mail address: [email protected] (J. Kaprio)

Smoking, hypertension and diabetes have rather consistently been exposed as strong risk factors for CHD in Western societies [9,10], but sedentary lifestyle is usually considered a somewhat weaker risk factor. High leisure physical activity is often accompanied by other beneficial health habits and there is no conclusive proof that physical activity itself is the primary preventive mechanism, or whether other health habits or genetic selection associated with the physically active lifestyle play the dominant role. One possible method to distinguish between physical activity and genetic factors in the etiology of CHD is to study family members, in particular monozygotic twins sharing the same genes. We investigated physical activity and other CHD risk factors with respect to genetic and other familial selection as predisposing or preventive factors for CHD in the Finnish Twin Cohort.

0021-9150/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 2 1 - 9 1 5 0 ( 9 9 ) 0 0 3 6 8 - 8

J. Kaprio et al. / Atherosclerosis 150 (2000) 193–200

194

2. Methods

2.1. Selection of subjects The Finnish Twin Cohort was compiled from the Central Population Registry of Finland via selection procedures described elsewhere [11]. In brief, the baseline sample comprised all same-sex twin-pairs born in Finland before 1958 with both co-twins alive in 1967. The original cohort also included other subjects who had the same family name and the same home parish, and were born on the same day, but were not twins. The target group of the present study was comprised of male cohort subjects who were alive and aged 25 –69 on 1 January 1976 (N= 14 516; Fig. 1). In autumn 1975, the subjects were mailed a questionnaire which included items on physical activity, body weight, height, alcohol use and smoking, physician diagnosed hypertension and diabetes mellitus, and occupation. Among those whose address could be identified (92.7% of men, Fig. 1), the response rate was 80.4% being higher for twins (85.5%) than for other subjects (60.2%) as those found to be singletons were no longer sent a reminder. Zygosity [monozygotic (MZ), dizygotic (DZ) or unclassified] was defined on the basis of questions on confusion by strangers in childhood and similarity in childhood appearance (‘like two peas in a pod’), which have been used in other large twin samples [12]. This method of zygosity diagnosis was found to be highly accurate when tested on a sample of 104 pairs [12]. The twofold ratio of DZ/MZ

Fig. 1. Flow chart of study subjects.

twins reflects the high frequency of fraternal twinning until the 1960s in Finland, as discussed elsewhere in detail [13]. We included subjects who supplied complete questionnaire data, were alive on 1 January 1977 and were free of cardiovascular diseases at baseline (Fig. 1). We excluded subjects with cardiovascular disease based on the following criteria: (1) physician-diagnosed angina pectoris, myocardial infarction or stroke in the baseline questionnaire, (2) angina-pectoris based on a standard angina pectoris and chest pain history questionnaire [14,15], or (3) having cardiovascular disease (except for hypertension or venous diseases) based on their hospital discharge reports between 1972 and 31 December 1976. The hospital discharge reports were collected from the reliable nationwide register kept by the National Board of Health, which includes private hospitals [16,17]. We excluded subjects with International Classification of Diseases (ICD), version 8 codes 390–399 or 410–449. In 1981, a questionnaire with identical physical activity items was re-administered to twins with a response rate of 91%. The questionnaire studies have been approved by the ethical committee of the Department of Public Health, University of Helsinki. Participants have also had multiple feedback during the course of the study. The Ministry of Social Affairs and Health, and the Data Protection Agency have given permission for the formation and follow-up of the cohort.

2.2. Assessment of physical acti6ity At baseline we assessed physical activity by asking the subjects about their leisure physical activity (monthly frequency, mean duration and mean intensity of physical activity bouts, as well as their opinion of their overall physical activity level), occupational physical activity and physical activity during the journey to work. We used three classifications of leisure physical activity habits [18]. First we classified the subjects in three physical acti6ity categories based on four structured questions. Those who reported exercising at least six times per month at a mean duration of at least 30 min, and with a mean intensity corresponding to at least vigorous walking or jogging, were classified as conditioning exercisers. Those who reported not partaking in leisure physical activity were considered sedentary. As a check, sedentary subjects also had to report in other questions that the intensity of any leisure activities did not exceed walking, and that they participated in leisure physical activities less than six times a month. Other subjects were classified as occasional exercisers (for number of subjects in each category, see Table 1). To verify the accuracy of our categorical classification we calculated a leisure acti6ity MET-index by assigning a

J. Kaprio et al. / Atherosclerosis 150 (2000) 193–200 Table 1 Characteristics of the men of the Finnish Twin Cohort, aged 25–69, with complete questionnaire data in 1975a Characteristics

Men (N = 8205)

Mean age (1 January 1976)

37.9 (10.5) 2

Mean Quetelet’s index (kg/m )

24.51 (2.93)

Cigarette habit (%) Never smokers Occasional smokers Former smokers Current smokers

29.4 4.0 25.7 41.0

Use of alcohol, geometric mean (g/day)

5.7 (6.2)

% reporting physician-diagnosed Hypertension Diabetes mellitus

9.5 1.0 12.7 23.4 44.1 8.4 9.3 2.2

Leisure physical acti6ity habits (%) Sedentary Occasional exercisers Conditioning exercisers Mean leisure activity MET-index

15.4 69.2 15.5 2.86 (3.35)

Work-related physical acti6ity (%) b Sedentary Standing or walking at work Light manual labor Heavy manual labor Never worked

29.0 19.2 37.7 14.1 0.6

b

of conditioning exercisers was 7.57 (standard error of the mean [S.E.M.] =0.14), of occasional exercisers 2.32 (S.E.M.= 0.03) and of sedentary men 0.54 (S.E.M. = 0.03). Assessment of work-related (current or latest) physical activity was based on responses to a question with four alternatives (Table 1). Changes in physical activity over a 6-year follow-up are shown in Table 2. Only about 2% of subjects who were sedentary in 1975 were conditioning exercisers in 1981, and about 2% who were conditioning exercisers in 1975 were sedentary in 1981, with most changes being between sedentary and occasional exercisers. The Spearman correlation between physical activity METindex in 1975 and 1981 was 0.45 [18].

2.3. Other CHD risk factors

Occupational group (%) Upper white-collar Clerical work Skilled workers Unskilled workers Farmers Others

a

195

S.D. in parentheses. No working history: 52 men.

multiple of resting metabolic rate (MET score) to each activity [19]. In this calculation we used MET values four (for exercise intensity corresponding to walking), six (vigorous walking to jogging), 10 (jogging), and 13 (running). We also included MET value four for physical activity during the journey to work, which was usually walking. The mean leisure activity MET-index

Using the subjects’ self reports, we obtained information on other predictors of CHD: Body-mass index [reported weight (kg) divided by the square of reported height (m)], occupational group, cigarette smoking, use of alcohol, physician diagnosed hypertension, and diabetes mellitus (Table 1). Classification by occupational group based on job title was into the following categories: executives, clerical staff, skilled workers, unskilled workers, and farmers [20]. The smoking status was classified from responses to a detailed smoking history [21]. The age of starting, age at possible cessation and the amount smoked were asked if the respondent had ever regularly smoked. The subjects were classified as current smokers, former smokers, occasional smokers or never smokers [21]. The life-long dose of smoking was calculated as pack-years (i.e. equivalent to smoking one pack (20 cigarettes) each day during a year). To validate the questionnaire, urine cotinine was measured in 10 pairs discordant for smoking: all smoking co-twins were positive, while their non-smoking co-twins showed no trace of smoking [22]. The use of alcohol was recorded in beverage-type specific items on frequency and quantity, converted into grams absolute alcohol and summed to yield an estimate of total consumption in grams per month [23].

Table 2 Leisure physical activity category in 1975 by physical activity category in 1981 among 6064 (twin) men responding to the Finnish Twin Cohort questionnaire studies in 1975 and 1981a Activity category in 1975

Activity category in 1981, N (row%) Sedentary

Occasional exercisers

Conditioning exercisers

Sedentary Occasional exercisers Conditioning exercisers

370 (40.7) 386 (9.2) 20 (2.1)

520 (57.3) 3346 (80.0) 447 (45.8)

18 (2.0) 449 (10.7) 508 (52.1)

Total

776

4313

975

a

Kappa statistic value for physical activity category in 1975 and 1981 was 0.35 (S.E. 0.01).

196


The questionnaire also included items for physician diagnosed diabetes mellitus and hypertension [24]. The check against registry data showed that in 83 (34 IDDM, 39 NIDDM, 10 other) of the 126 MZ or DZ twins who reported diabetes at baseline a record of diabetes was found from the Finnish registers for hospital inpatient discharges or reimbursable medications until 1985 [25]. Not all subjects with diabetes exercise their right to free-of-charge medications or need hospital treatment.

childhood environment and share part or all of their genes in common by descent. We estimated Odds Ratios (OR) for CHD discordant twin-pairs with respect to discordance of physical activity and other CHD risk factors using conditional logistic regression analysis [29].

3. Results

3.1. Indi6idual-based analyses 2.4. Outcome measures Our review of the hospital discharge reports covered records (dates of admissions and discharges as well as diagnoses) from 1 January 1972 to 31 December 1995. We used the ICD codes (ICD 8 until 1985 and ICD 9 from 1986 onwards) to form the diagnostic category of coronary heart disease (ICD 410 – 414). The overall correspondence in Finland between individual hospital discharge records and their written patient histories for different diagnoses is reportedly 94.4% [16]. When the discharge register data were compared to data based on re-examination of survivors 5 years after myocardial infarction, the specific diagnosis code had been recorded for 84.7% of the definite myocardial infarctions [17]. More recent studies have confirmed the continued high quality of the discharge diagnoses [26– 28]. Causes of death for the years 1977 – 1995 were available from the Cause-of-Death Bureau files at the Central Statistical Office of Finland. The cause of death was recorded using ICD codes by trained coders. In 53% of CHD deaths before age 70 the cause of death was based on forensic or medical autopsy. Follow-up of CHD morbidity and mortality started from 1 January 1977; we followed CHD hospital discharges and CHD deaths to the end of 1995.

2.5. Statistical analyses We computed the incidence rate of CHD, defined as the first occurrence during follow-up of hospitalization or death from CHD, and CHD mortality rates as events per person-year. Rate ratios (RR) for rates in physically active subjects compared to rates in sedentary subjects and their statistical significance were modelled using Poisson regression [29] to adjust for age and other covariates. Because most of the study subjects were twins, twinpairs discordant for physical activity habits and for CHD were examined to determine whether the CHD morbidity of physically active subjects differed from that of their age-matched sedentary sibs (either monozygotic or dizygotic co-twins). These co-twins represent persons who have nearly always shared the same

Of the men initially free of CHD (Fig. 1), 8205 had complete data on all covariates (age, body-mass index, smoking, hypertension, diabetes) and leisure physical activity (Table 1). Of these men, 723 had CHD events and 240 CHD deaths during the follow-up up to the age of 69 years. As expected, increased body-mass index, smoking, hypertension and diabetes were strong predictors for CHD (Table 3). In men the age-adjusted risk for CHD was explained by activity category (Table 3). There was no significant interaction between physical activity and calendar time, meaning that baseline physical activity predicted CHD similarly during the first and second halves of the follow-up. There was no significant interaction between smoking and physical activity indices (P= 0.26). There was no association between baseline occupational physical demands and CHD events.

3.2. Pairwise analyses The twin–co-twin intraclass correlation of leisure activity MET-index was 0.25 for male twin-pairs (0.41 for MZ-pairs and 0.20 for DZ-pairs), which indicates a moderate genetic influence on leisure activity. The OR of incident CHD was 0.54 (95% CI 0.27– 1.09) in the twin who was a conditioning exerciser at baseline compared to the sedentary co-twin, and after adjustment for body-mass index, smoking, hypertension and diabetes the OR was 0.81 (95% CI 0.38–1.72) (Tables 4 and 5). Among the 70 monozygotic discordant pairs, the crude OR was 0.50 (95% CI 0.15–1.66) in the occasional exerciser compared to the sedentary co-twin and 0.21 (95% CI 0.04–1.31) in the conditioning exerciser compared to the sedentary co-twin. After adjustment for all covariates the respective ORs were 0.62 (95% CI 0.18–2.18 for occasional and 0.40 (95% CI 0.06–2.78) for conditioning exercisers compared to sedentary among the MZ pairs. For the male twins-pairs discordant for CHD on follow-up (Fig. 1; N=311), pairwise analyses showed that smoking and diabetes clearly increased the risk for CHD (Table 5). Among MZ twin-pairs discordant for smoking (N= 18) the crude OR of CHD was 5.00 (95% CI 1.45–17.3) and among MZ twin-pairs discordant for


197

Table 3 Relative risks of CHD events before the age of 70 with follow-up from 1 January 1977 through 31 December 1995 according to different risk factors in 1975a

Acti6ity category Sedentary Occasional exercisers Conditioning exercisers P for trend Hypertension (yes) Diabetes (yes) Smokers (current/other) Body-mass index B27 ]27 and B30 ]30

N with CHD/total number of subjectsb

Relative riskc (95% CI)

Relative riskd (95% CI)

146/1256 516/5653 61/1268 129/775 19/79 405/3355

1 0.77 (0.64–0.92) 0.52 (0.39–0.71) P =0.0001 1.87 (1.53–2.16) 2.47 (1.56–3.89) 2.13 (1.84–2.47)

1 0.84 (0.70–1.01) 0.68 (0.50–0.92) P=0.010 1.86 (1.53–2.21) 2.36 (1.50–3.73) 2.18 (1.87–2.53)

532/6714 134/1115 57/348

1.0 1.23 (1.02–1.49) 1.72 (1.31–2.26)

1.0 1.21 (1.00–1.47) 1.66 (1.26–2.19)

a

Poisson regression model. Based on 8177 men, as 28 subjects turned 70 during 1976. c Age-adjusted. d With age, body-mass index, smoking, hypertension and diabetes in the model. b

Table 4 Leisure physical activity category in 1975 among 311 same-sex twin-pairs and among 70 same-sex MZ twin-pairs of the Finnish Twin Cohort discordant for CHD during follow-up Activity category of the healthy twin

Sedentary Occasional exerciser Conditioning exerciser

Activity category of the twin with incident CHD; N all pairs (N MZ pairs) Sedentary

Occasional exerciser

Conditioning exerciser

25 (9) 43 (8) 0 (0)

33 (4) 149 (37) 32 (7)

2 (0) 19 (3) 8 (2)

Table 5 Odds ratios of CHD before the age of 70 among 311 male twin-pairs discordant for CHD during the follow-up of the Finnish Twin Cohort from 1977 through 1995, according to patterns of physical activity in 1975 and other characteristics in 1975 Characteristic

N pairs

Total N

311

Acti6ity category Sedentary Occasional exerciser Conditioning exerciser P for trend Concomitants Current smoker (yes) Hypertension (yes) Diabetes (yes) BMI, per unit a

101 62 11 310

Odds ratio, unadjusted (95% CI)

Odds ratioa (95% CI)

Odds ratiob (95% CI)

1 0.82 (0.52–1.29) 0.54 (0.27–1.09) 0.10

1 0.86 (0.54–1.38) 0.66 (0.32–1.38) 0.29

1 0.86 (0.53–1.39) 0.81 (0.38–1.72) 0.54

3.21 (2.03–5.07) 1.70 (1.01–2.84) 10.0 (1.28–78.1) 1.09 (1.02–1.17)

1.60 (0.93–2.75) 13.1 (1.60–107) 1.11 (1.03–1.20)

3.55 (2.19–5.76) 1.31 (0.75–2.28) 20.7 (2.31–185) 1.14 (1.05–1.23)

Adjusted for smoking. Physical activity adjusted for body-mass index, diabetes, hypertension and smoking. Each concomitant adjusted for other concomitants and physical activity. b

198


hypertension (N = 13) the respective OR was 0.86 (95% CI 0.29–2.56). For BMI, the MZ pair OR was 1.07 (95% CI 0.88–1.37) per unit of BMI.

4. Discussion In this analysis of the Finnish Twin Cohort, we found further evidence that high leisure physical activity compared to sedentariness helps prevent CHD in men. The baseline health of the cohort was assessed by a questionnaire and by investigating their hospital records. About 3% of the subjects had a cardiovascular diagnosis in their hospital discharge reports between 1 January 1972 and 1 January 1977 and were excluded from the study. Between 1972 and 1977, 39% of subjects had been hospitalized for other reasons but their discharge reports did not include cardiovascular disease, which is further evidence for their cardiovascular health. CHD as an outcome measure was partly based on the Finnish hospital discharge registry, which data are sufficiently valid for research on predictors of disease [16,17,26–28]. The higher twin– co-twin correlation of physical activity index in MZ than DZ twins shows that there is also a genetic contribution to physical activity habits, as reported earlier [30]. Our analyses on individuals was based on the statistical assumption of independent observations, which was not strictly true, as 30.7% of subjects were age-matched sibs of other study subjects. Statistical models for analysing binary outcomes in sibships, known as random effects logistic regression models, assume either that the covariates are identical for both sibs in a pair or, if not identical, that preferably all of the sibships are full pairs [31]. Neither assumption was tenable in this data set, so we used standard regression models, which may have lead to slightly wider confidence intervals if the twin dependence had been accounted for. However, the correlations between twins for physical activity patterns was overall very modest, and the majority of pairs with CHD during follow-up were discordant (71% of MZ pairs, and 83% of DZ pairs). Thus we also observed greater concordance of MZ pairs than of DZ pairs, as found also in the Swedish Twin Registry [32] and in US WWII veteran twins [33]. The analyses on individuals demonstrate that in the present cohort we found the same relationship for men between leisure physical activity and CHD established in other epidemiological studies [2 – 4]. The risk of CHD death was very similar to the risk for CHD events overall with respect to physical activity indices at baseline. Also, the CHD mortality rates for men in the present study did not differ by zygosity (non-twin, unknown zygosity, MZ, DZ) in the Poisson regression analysis. The mortality of Danish adult twins has been found to be no different from that of singletons, and no differ-

ences were seen by zygosity [34]. The design of using discordant pairs to test for genetic selection effects in relation to smoking and health has been used earlier for the present cohort [22], and in the US WWII veteran twin series [35], but this is the first such analysis of physical activity at leisure. Expectedly, in the present study smoking and diabetes strongly increased CHD risk also in the discordant pair analysis, and BMI was also a significant predictor. These findings further confirm that the relationship between these factors is not mediated by common genetic or shared familial influences. Interestingly, baseline occupational physical activity did not associate with future manifestations of CHD even though it causes much higher total energy expenditure than leisure physical activities. Compulsory longlasting and often static work-related physical strain may have physiological effects different from dynamic and short-lasting leisure time exercise or dynamic work-related physical activity. Sedentary habits are probably as hazardous for women as for men. The statistical power to analyze the associations in women in a pairwise manner in the Finnish Twin Cohort was too low due to the smaller number of CHD cases and less discordance in physical activity habits. Also, vigorous physical activity was less common in women than men in 1975 in Finland. Baseline physical activity category strongly predicted future CHD events in the whole cohort of men. Analyses in discordant twin-pairs showed that physical activity category associated with CHD rather similarly in monozygotic and dizygotic twin-pairs, as it did in the whole cohort, even though not all relative risks were statistically significant. This consistency in results for individuals and pairs, and with similar results in MZ and DZ pairs suggests that genetic selection cannot fully account for the well-established epidemiological relationship between leisure-time physical activity and CHD. After adjustment for different risk factors the relative risks by physical activity generally became less significant compared to the unadjusted relative risks. We also adjusted the results for pack-years smoked, but this changed the RRs only minimally. However, this attenuation of relative risks after adjustment for other covariates may reflect the role of some factors, such as BMI, in the causal pathway between physical activity and CHD. Our findings strengthen the evidence that in male subjects physical activity offers protection against CHD. This is supported by the recent intervention study of Niebauer et al. [36], in which physical activity was the best predictor of retardation of angiographically assessed coronary artery disease. Our findings also provide evidence that genetic selection bias does not fully underlie the finding of a beneficial effect of leisure


physical activity when intensity, frequency and duration of activity are considered. However both sedentariness and high body-mass index were weaker risk factors than smoking and diabetes.

Acknowledgements Supported by the Academy of Finland (grants 38332 and 42044), the Finnish Ministry of Education and the Paavo Nurmi Foundation. We thank Richard Burton for his language revision of the manuscript.

References [1] Morris JN, Everitt MG, Pollard R, Chave SPW. Vigorous exercise in leisure-time: Protection against coronary heart disease. Lancet 1980;ii:1207–10. [2] Powell KE, Thompson PD, Caspersen CJ, Kendrick JS. Physical activity and the incidence of coronary heart disease. Annu Rev Public Health 1987;8:253–87. [3] Berlin JA, Colditz GA. A meta-analysis of physical activity in the prevention of coronary heart disease. Am J Epidemiol 1990;132:612 – 28. [4] Bouchard C, Shephard RJ, Stephens T. Physical activity, fitness, and health. Proceedings and consensus statement. Champaign, IL: Human Kinetics Publishers, 1993. [5] Physical Activity and Health. A report of the Surgeon General, U.S. Department of Health and Human Services, Center for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, The President’s Council on Physical Fitness and Sports, 1996. [6] Fletcher GF. How to implement physical activity in primary and secondary prevention. A statement for healthcare professionals from the task force on risk reduction, American Heart Association. Circulation 1997;96:355–7. [7] Paffenbarger RS Jr, Hyde RT, Wing AL, Lee I-M, Jung DL, Kampert JB. The association of changes in physical-activity level and other lifestyle characteristics with mortality among men. N Engl J Med 1993;328:538–45. [8] Blair SN, Kohl HW, Barlow CE, Paffenbarger RS, Gibbons LW, Macera CA. Changes in physical fitness and all-cause mortality. A prospective study of healthy and unhealthy men. JAMA 1995;273:1093–8. [9] Dawber TR. The Framingham Study. Cambridge, MA: Harvard University Press, 1980. [10] Stamler J. Established major coronary risk factors. In: Marmot M, Elliott P, editors. Coronary Heart Disease Epidemiology. Oxford: Oxford University Press, 1992:35–66. [11] Kaprio J, Sarna S, Koskenvuo M, Rantasalo I. The Finnish Twin Registry: Formation and compilation, questionnaire study, zygosity determination procedures and research program. Prog Clin Biol Res 1978;24B:179–84. [12] Sarna S, Kaprio J, Sistonen P, Koskenvuo M. Diagnosis of twin zygosity by mailed questionnaire. Hum Hered 1978;28:241 – 54. [13] Rose RJ, Koskenvuo M, Kaprio J, Sarna S, Langinvainio H. Shared genes, shared experiences, and similarity of personality: Data from 14,288 adult Finnish co-twins. J Pers Soc Psychol 1988;54:161 – 71. [14] Rose G, Blackburn H. Cardiovascular survey methods. World Health Organization Monograph Series No 65. Geneva: WHO, 1968.

199

[15] Pyo¨ra¨la¨ K, Punsar S, Reunanen A, Heinonen OP, Puro K, Aromaa A. Kansanela¨kelaitoksen sepelvaltimotautitutkimus. Rintakipuoireiden ja EKG-lo¨ydo¨sten esiintyvyys alkututkimuksessa. Publications of the Social Insurance Institution of Finland, Helsinki, 1975;AL1. [16] Keskima¨ki I, Aro S. Accuracy of data on diagnosis, procedures and accidents in the Finnish hospital discharge register. Int J Health Sci 1991;2:15 – 21. [17] Helio¨vaara M, Reunanen A, Aromaa A, Knekt P, Aho K, Suhonen O. Validity of hospital discharge data in a prospective epidemiological study on stroke and myocardial infarction. Acta Med Scand 1984;216:309 – 15. [18] Kujala UM, Kaprio J, Sarna S, Koskenvuo M. Relationship of leisure-time physical activity and mortality. The Finnish Twin Cohort. JAMA 1998;279:440 – 4. [19] Wilson PWF, Paffenbarger RS, Morris JN, Havlik RJ. Assessment methods for physical activity and physical fitness in population studies: Report of a NHLBI workshop. Am Heart J 1986;111:1177– 92. [20] Central Statistical Office of Finland: Alphabetical list of occupations and classification of social class. In Finnish. (Tilastokeskus. Aakkosellinen ammattihakemisto seka¨ sosiaaliryhmittely, 1972). Helsinki, 1972. [21] Kaprio J, Koskenvuo M. A prospective study of psychological and socioeconomic characteristics, health behavior and morbidity in cigarette smokers prior to quitting compared to persistent smokers and non-smokers. J Clin Epidemiol 1988;41:139– 50. [22] Lassila R, Seyberth HW, Haapanen A, Scheer H, Koskenvuo M, Laustiola K. Vasoactive and atherogenic effects of cigarette smoking: A study of monozygotic twins discordant for smoking. Br Med J 1988;297:955 – 7. [23] Kaprio J, Koskenvuo M, Langinvainio H, Romanov K, Sarna S, Rose RJ. Genetic influences on use and abuse of alcohol: A study of 5638 adult Finnish twin brothers. Alcohol Clin Exp Res 1987;11:349 – 56. [24] Kujala UM, Kaprio J, Taimela S, Sarna S. Prevalence of diabetes, hypertension, and ischemic heart disease in former elite athletes. Metabolism 1994;43:1255 – 60. [25] Kaprio J, Tuomilehto J, Koskenvuo M, Romanov K, Reunanen A, Eriksson J, Stenga˚rd J, Kesa¨niemi YA. Concordance for type 1 (insulin-dependent) and type 2 (non-insulin-dependent) diabetes mellitus in a population-based cohort of twins in Finland. Diabetologia 1992;35:1060 – 7. [26] Pietila¨ K, Tenkanen L, Ma¨ntta¨ri M, Manninen V. How to define coronary heart disease in register-based follow-up studies: experience from the Helsinki Heart Study. Ann Med 1997;29: 253 – 9. [27] Ma¨honen M, Salomaa V, Brommels M, Molarius A, Miettinen H, Pyo¨ra¨la¨ K, Tuomilehto J, Arstila M, Kaarsalo E, Ketonen M, Kuulasmaa K, Lehto S, Mustaniemi H, Niemela¨ M, Paloma¨ki P, Torppa J, Vuorenmaa T. The validity of hospital discharge register data on coronary heart disease in Finland. Eur J Epidemiol 1997;13:403 – 15. [28] Rapola JM, Virtamo J, Korhonen P, Haapakoski J, Hartman AM, Edwards BK, Heinonen OP. Validity of diagnoses of major coronary events in national registers of hospital diagnoses and deaths in Finland. Eur J Epidemiol 1997;13:133– 8. [29] Preston DL, Lubin JH, Pierce DA, McConney ME. Epicure, Release 2.0. Seattle, WA: HiroSoft International, 1993. [30] Pe´russe L, Tremblay A, Leblanc C, Bouchard C. Genetic and environmental influences on level of habitual physical activity and exercise participation. Am J Epidemiol 1989;129:1012. [31] EGRET, Statistical Software Manual. Seattle, WA: Statistics and Epidemiology Research, 1990.

200


[32] Marenberg ME, Risch N, Bergman LF, Floderus B, de Faire U. Genetic susceptibility to death from coronary heart disease in a study of twins. N Engl J Med 1994;330:1041–6. [33] Carmelli D, Selby JV, Quiroga J, Reed T, Fabsitz RR, Christian JC. 16-year incidence of ischemic heart disease in the NHLBI twin study. A classification of subjects into high- and low-risk groups. Ann Epidemiol 1994;4:198–204. [34] Christensen K, Vaupel JW, Holm NV, Yashin AI. Mortality among twins after age 6: Fetal origins hypothesis versus twin

.

method. Br Med J 1995;310:432 – 6. [35] Carmelli D, Page WF. Twenty-four year mortality in World War II US male veteran twins discordant for cigarette smoking. Int J Epidemiol 1996;25:554 – 9. [36] Niebauer J, Hambrecht R, Velich T, Hauer K, Marburger C, Kalberer B, Weiss C, von HE, Schlierf G, Schuler G, Zimmerman R, Kubler W. Attenuated progression of coronary artery disease after 6 years of multifactorial risk intervention: role of physical exercise. Circulation 1997;96:2534 – 41.