ORIGINAL ARTICLE. C Pitsavos 1,2, DB Panagiotakos 1,2, C Chrysohoou 1, P Kokkinos 2, A Menotti 3,4, S Singh 2 and A Dontas 3,4 1.

Transcription

1 (2004) 18, & 2004 Nature Publishing Group All rights reserved /04 $ ORIGINAL ARTICLE Physical activity decreases the risk of stroke in middle-age men with left ventricular hypertrophy: 40-year follow-up ( ) of the Seven Countries Study (the Corfu cohort) C Pitsavos 1,2, DB Panagiotakos 1,2, C Chrysohoou 1, P Kokkinos 2, A Menotti 3,4, S Singh 2 and A Dontas 3,4 1 First Department of Cardiology, School of Medicine, University of Athens, Greece; 2 Department of Cardiology, Veterans Affairs Medical Center and Georgetown University, Washington, DC, USA; 3 Association for Cardiac Research, Roma, Italy Left ventricular hypertrophy (LVH) is a dominant characteristic of cardiovascular mortality. We investigated the interaction between physical activity and LVH on stroke mortality, based on a 40-year follow-up of the Corfu cohort from the Seven Countries Study. The population studied consisted of 529 rural men (40 59 years old) enrolled in the Study at LVH was electrographically confirmed according to Minnesota coding. Physical activity levels were assessed by selfreports of habitual, occupational and leisure-time activities. Cox proportional hazard models were used to evaluate exercise levels with respect to stroke mortality in people with and without LVH. During the follow-up, 461 (87%) died, and 74 (16%) of these deaths were due to stroke. LVH was present in 40 (7.5%) men. A total of 362 (68%) men were defined as physically active. Physical activity was associated with a lower risk of stroke (hazard ratio ¼ 0.65, Po0.05). On the other hand, LVH had 5.8-fold the risk of stroke (Po0.001) among sedentary and 4.5-fold the risk (Po0.001) among physically active men, after controlling for several potential confounders. However, moderate physical activity decreased the risk of stroke by 49% in men with LVH as compared to sedentary without LVH (hazard ratio ¼ 0.51, Po0.01), while hard exercise did not confer any significant reduction in stroke risk. We revealed the benefits from moderate physical activity on stroke mortality among men with LVH. Physicians and other health-care professionals should encourage patients with LVH to adopt a physically active lifestyle. (2004) 18, doi: /sj.jhh Published online 26 February 2004 Keywords: physical activity; stroke; left ventricular hypertrophy Introduction Correspondence: Dr. DB Panagiotakos, 46, Paleon Polemiston St., Glyfada , Athens, Greece. d.b.panagiotakos@usa.net 4 On behalf of the Seven Countries Study investigators. Received 30 September 2003; revised/accepted 5 January 2004; published online 26 February 2004 Each year, about people in the Unites States have a stroke and about one-third of them die, making stroke the third most common cause of death after heart disease and cancer. 1 Several factors may increase the risk for stroke events, including high blood pressure, diabetes mellitus, physical inactivity, cigarette smoking and abnormal fibrinogen and homocysteine levels. 1,2 Since treatment options are relatively limited, public health strategies focus on the prevention of stroke by the modification of the traditional risk factors. 1 Increased left ventricular mass is now considered as an independent predictor of coronary heart disease (CHD) and stroke, 3 5 but the mechanisms responsible for this association are not well understood. 4 Recent studies suggest that physical activity may protect against stroke. 1,2 This protection may be due to the favourable effects of exercise on several risk factors associated with the risk of stroke, 2 including reduction in left ventricular mass of hypertensive patients. 6 Recently, a case control study 5 reported that physical activity decreases the risk of stroke in patients with increased left ventricular mass to a level comparable to that of patients without increased left ventricular mass. However, the findings could not provide evidences for causality.

2 496 The Seven Countries Study 6 has become a classic in science for its pioneering effort in cardiovascular disease epidemiology. The study was started in the late 1950 s by Ancel Keys and his colleagues. They enrolled middle-aged men from 16 cohorts of seven participating countries (USA, Italy, Japan, The Netherlands, Finland, former Yugoslavia and Greece). During the 2000s, the study has completed, approximately, four decades of prospective investigation. Among several lifestyle, clinical and biological factors, the investigators assessed the effect of physical activity on the risk of coronary heart disease and stroke. In this work, based on 40 years of prospective investigation from the Corfu cohort of the Seven Countries Study, we evaluated the effect of physical activity on stroke mortality in middleaged men with left ventricular hypertrophy (LVH). Methods Study population The population studied in this work was the Corfu cohort (Greece) from the Seven Countries Study. 6 From September to October 1961, an international team of doctors, nurses and technicians examined 529 men from Corfu Island. This cohort was made up of almost all the men who lived in the above area (95.3%) and were years old in the early fall of These were mainly small-scale, nonmechanized farmers (55%) at entry. 6 Follow-up Since 1961, periodic visits, every 5 years, were made by the Seven Countries Study investigators in the Corfu Island in order to define the vital status of the participants. The causes of death were obtained from the previous clinical records filled out by the study s research group, or by hospital records, or by necroscopy records (when it was available), or by information from family or hospital doctors, other specialists, family or relatives, friends and any other witnesses. 6 Information from the police, in case of violent causes or death occurred suddenly in public places or un-witnessed, was also obtained. 6 In July 2001, collection of data on vital status and causes of death was complete for 40 years and no subject was lost to follow-up. A single reviewer following defined criteria, employing the eighth Revision of the WHO-ICD, determined the causes of death. In the presence of multiple causes, a hierarchical preference was adopted with violence, cancer in advanced stages, coronary heart disease and stroke in that order. The ICD codes were then transformed into a more compact classification called Laboratory of Physiological Hygiene (LPH). The primary end point of this analysis was stroke mortality, which occurred in men free from cardiovascular disease at entry examination. Stroke deaths were those coded in the WHO-ICD. They corresponded to the LPH code We also assessed coronary heart disease deaths including myocardial infarction, angina pectoris and other identified forms of ischaemia (LPH codes and 740, or WHO- ICD coding , 427.2, 427.6, A , , 427.9) and cancer (LPH codes 751 and 752 or WHO-ICD coding , ). Physical activity assessment Physical activity classification in the Seven Countries Study was based on responses to questions about occupation and usual leisure-time activities, including part-time jobs and other nonoccupational exercise. 7. A list of almost 100 occupations was used for final classification and has been published by Keys et al. 7 Moreover, the majority of men in the Corfu cohort lived in rural areas where heavy physical activity related to work was usual. Thus, all men in the study were classified at entry into three classes: Class 1 men were sedentary, engaging in little exercise. Class 2 men were moderately active during a substantial part of the day. Class 3 men performed hard physical work much of the time. A rough estimate of energy expenditure corresponding to these three physical activity classes was made on the basis of ergonometric procedures carried out by the investigators of the study. Thus, for class 1 the energy expenditure was less than 2400 kcal/day, for class 2 it was between 2400 and 3000 kcal/day and for class 3 it was more than 3000 kcal/day. 10 Evaluation of LVH All participants had an electrocardiogram (ECG) using a three-channel machine. Standard posture for resting ECG was supine. Smoking, meals and heavy physical exercise were avoided for at least 3 h before the recording. Electrocardiography abnormalities according to Minnesota Code 1968 were defined by the presence of LVH. In particular, the presence of high QRS voltage (codes 3.1, 3.3, and 4-1 to 4-3 or 5-1 to 5-3) was considered evidence for LVH. 8 Other investigated measurements The following entry information has also been considered: age in years, supine resting systolic and diastolic blood pressure of the right arm in mmhg (mean of two consecutive measurements), daily cigarette smoking (identified by a positive response on a standardized questionnaire), total serum cholesterol in mmol/l measured in a casual

3 blood sample by the method of Abell Kendall, modified by Anderson and Keys 9 and body mass index (weight (kg)/height (m 2 )). During the base-line period (1961), medications designed to lower blood pressure were rarely prescribed, and therefore use of antihypertensive medication was not included in the analysis followed on. Further details regarding the protocol and procedures applied in the Seven Countries Study as well as the Corfu cohort have been presented in detail elsewhere When the Seven Countries Study began, it was not standard practice in clinical research to ask participants for written informed consent or to ask for approval from medical ethics committees. Statistical analysis Continuous variables are presented as mean7one standard deviation, while qualitative variables are presented as absolute and relative frequencies. Contingency tables with calculation of w 2 -test, as well as use of Student s t-test evaluated associations between the categorical (LVH and physical activity groups) and normally distributed continuous variables (age at entry, body mass index, systolic and diastolic blood pressures, total cholesterol levels). Death rates were calculated using the observed person-time, in years. The proportion of surviving persons has been recorded every 6 months. We had complete data from all the individuals. Survival analysis was performed using the Cox proportional hazards model with stroke as end point and the aforementioned risk factors as predictors. The continuous factors (blood pressures and total serum cholesterol) were entered linearly, while age was tested exponentially in the survival model. The assumption of linearity was graphically tested plotting log(survival function(t) vs time. The final model was developed through backward stepwise elimination procedures, for the selection of variables and after controlling for interactions between LVH and physical activity status and various other potential confounders. In particular, all investigated variables were entered in the initial (full) model. Afterwards, the variable with the higher probability (P-value) was removed, unless its P-value was o5%. This procedure was repeated continuously (ie exclusion of one variable each time), till no variables with P-value 45% were in the model. The associations between the investigated factors and the fatal events are presented as exponentials (hazard ratios) of the estimated coefficients. The goodness-of-fit test was based on comparing the observed survival probability with the expected under the assumption of proportional hazards. The assumption of proportionality was graphically assessed through the plots of weighted deviance residuals vs time. We have calculated that the number of enrolled participants and the duration of follow-up (ie 40 years) provides 82% statistical power for assessing two-sided differences on death rates between groups of study equal to 10%, at probability level (P) equal to SPSS (SPSS Inc., Texas, USA) version 11.0 software was used for all the statistical calculations. Results Baseline characteristics The prevalence of LVH at baseline was 8% or 40 out of 529 men. Of the 529 (68%) men, 362 were classified as physically active. Systolic and diastolic blood pressures were significantly higher in the LVH group, while age, serum total cholesterol, body mass index, physical activity status and smoking habits were similar in those with and without LVH (Table 1). Significant differences in various baseline cardiovascular risk factors were noted among the sedentary, moderate and heavy activity level groups. Particularly, physically active men had lower body mass index (hard: vs moderate: Table 1 Baseline characteristics of men in the Corfu cohort (1961), according to the LVH group Presence of LVH (n ¼ 40) Absence of LVH (n ¼ 489) P-value Age at baseline (years) Systolic blood pressure (mmhg) Diastolic blood pressure (mmhg) Serum total cholesterol (mmol/l) Current smoking (total) 30 (76%) 308 (63%) o10 cigarettes/day 14 (36%) 62 (13%) cigarettes/day 8 (20%) 142 (29%) X20 cigarettes/day 8 (20%) 104 (21%) Physical activity Sedentary life 12 (30%) 155 (32%) Moderate physical activity 12 (30%) 185 (38%) Hard physical activity 16 (40%) 149 (30%) Body mass index (kg/m 2 )

4 498 vs sedentary: kg/m 2, Po0.001), lower systolic and diastolic blood pressure levels (hard: 131/80721/10 vs moderate: 135/82719/11 vs sedentary: 139/85724/12 mm Hg, Po0.01) and total cholesterol values (hard: vs moderate: vs sedentary: mmol/l, P ¼ 0.04) compared to sedentary men. All the previous associations were taken into account for the multivariate analysis that followed on. No associations were found between physical activity status and age (physically active: 4977 vs inactive: 4976 years old, P ¼ 0.635) as well as smoking habits of the participants (physically active: 277 or 77% vs inactive: 116 or 70%, P ¼ 0.162). Follow-up data During the 40-year follow-up, 461 of 529 (87.1%) men died due to several causes. Of these, 74 (14%) deaths were due to stroke, 120 (26%) deaths were attributed to coronary heart disease, 118 (26%) deaths due to cancer and the rest of deaths, that is 149 (34%), due to other reasons. In addition, seven of 40 (18%) men with LVH and 67 (14%) of those without LVH died due to stroke. The stroke mortality rate for men with LVH was seven deaths per 1000 person-years and for men without LVH was five per 1000 person-years (P ¼ 0.31). Moreover, 8 (20%) men with LVH and 112 (23%) of those without LVH died due to coronary heart disease (P ¼ 0.423). There were 47 (13%) strokes among the 362 physically active men and 27 (16%) strokes among the 167 sedentary people. In addition, the stroke mortality rate for the physically active men was five deaths per 1000 person-years and seven deaths per 1000 person-years for the sedentary men (P ¼ 0.16). Figure 1 illustrates the survival probabilities by physical activity status and LVH. As we can see, an association between LVH and exercise status seems to exist, since the presence of LVH in sedentary men has the worst prognosis for stroke deaths. On the other hand, physical activity seems to confer an increase in survival in men with and without LVH (Po0.001). Multivariate survival analysis In order to further clarify the association between the presence of LVH and physical activity status on the risk of stroke, we performed a stratified analysis (Table 2). After adjustments for age, body mass index, total cholesterol, glucose, smoking and blood pressure levels were made, moderate physical activity was associated with 30% lower risk of stroke compared to sedentary men with LVH. Hard physical activity seems to confer less reduction to the risk of stroke as compared to moderate physical Survival probability PA - No LVH No PA - LVH PA - LVH No PA - No LVH Follow up (years) Figure 1 Age-adjusted survival probability of stroke mortality by physical activity (PA) and LVH status. Table 2 Estimates from a backward stepwise Cox model predicting 40-year stroke mortality as a function of baseline risk factors, by LVH level Presence of LVH Absence of LVH HR 95% CI HR 95% CI Physical activity status Sedentary (reference group) Moderate Hard Age (per 1 year) Systolic blood pressure (per 10 mmhg) Total serum cholesterol (per 1 mmol/l) LVH ¼ left ventricular hypertrophy; HR ¼ hazard ratio; CI ¼ confidence interval. Variables that were also entered in the models but were insignificant are age, smoking habits, body mass index, diastolic blood pressure and glucose levels.

5 activity (hazard ratio ¼ 0.75, P ¼ 0.2). Similar results were observed in the group without LVH since the risk reduction was 36% for those in moderate physical activity and 28% for those in hard physical activity (hazard ratio ¼ 0.72, P ¼ 0.11). The results from the Cox proportional hazard models that evaluated the effect of physical activity on stroke risk in people with and without LVH are presented in Table 2. Thereafter, we assessed the effect of the interaction between the presence of LVH and physical activity. Initially, we found that compared to men without hypertrophy the presence of LVH was 5.8-fold (95% CI ) the relative risk of stroke among sedentary men, after adjustment for several baseline characteristics like age, systolic and diastolic blood pressure, total cholesterol, glucose levels and smoking habits. Similarly, LVH was 4.5-fold (95% CI ) shows the risk of stroke among physically active men. Then we observed a significant interaction between physical activity and the presence of LVH on the risk of stroke (hazard ratio for the interaction term ¼ 0.51; CI ). The latter means that physical activity decreased the risk of stroke by 49% in men with LVH as compared to sedentary men without LVH, after controlling for the previously mentioned set of covariates. Furthermore, we found that age (hazard ratio per 1-year increase ¼ 1.10, Po0.001), smoking (hazard ratio ¼ 1.79, Po0.001) and body mass index (hazard ratio per 5 kg/m 2 increase ¼ 1.05, P ¼ 0.09) were significant determinants for CHD mortality. However, physical activity was not associated with coronary disease mortality (hazard ratio ¼ 1.24, P ¼ 0.15), after adjusting for the aforementioned factors. Moreover, no differences were observed with regard to the effect of physical activity or the other factors on CHD risk when we stratified our analysis by LVH status. Discussion In this work, we assessed the effect of physical activity on stroke mortality in middle-aged men from the Seven Countries Study (the Corfu cohort), after taking into account the presence of LVH. We found that physical activity reduced the risk of stroke in men with and without LVH. Particularly, we noted that physical activity decreased by 49% the risk of stroke in men with LVH as compared to sedentary men without LVH. According to the 1996 US Surgeon General s Report on Physical Activity and Health, the relationship between exercise and stroke is inconclusive. 13 However, recent studies, including the present, support that physical activity may protect people against strokes, especially due to the modification of several traditional risk factors like high blood pressure levels, insulin sensitivity and cigarette smoking. 2,14 Moreover, left ventricular mass is considered a risk factor of stroke. 5,15 17 However, the effect of the interaction between physical activity and LVH on the risk of stroke has rarely been investigated. In a recent case control study, Rodriguez et al 5 observed a significant association between increased left ventricular mass and physical activity. However, the design of this study did not allow inferences on a cause effect relationship. Our findings expanded the existing knowledge on the relationship between exercise, LVH and stroke. In the 40-year follow-up of one of the cohorts from the Seven Countries Study (Corfu cohort), we observed that electrocardiographically confirmed LVH was associated with a 5.8-fold risk of stroke among sedentary and a 4.5-fold risk among physically active men. Moreover, we noted that the adoption of even a moderate physically active lifestyle was associated with a significant reduction in the risk of stroke among men with and without LVH, respectively. The impact of physical activity on stroke mortality can be appreciated further by our finding that the risk of stroke in physically active men with LVH was 49% lower than the risk observed in sedentary men without LVH. No association between physical activity and CHD mortality was observed in our cohort, even when we stratified our sample by LVH status. It is also of interest that hard physical activity appears to confer less reduction in the risk of stroke, compared to moderate physical activity, in men with and without LVH (Table 2). This can have a significant health impact on a large population. Low-intensity exercise carries a lower risk for cardiovascular complications compared to higher intensity. 18 Physicians and other health-care professionals are more likely to encourage patient participation in a low-intensity rather than high-intensity exercise programme. Participants are also more likely to maintain a lower than a higher intensity exercise programme. 18 The mechanisms by which physical activity moderates the effect of LVH on the risk of stroke are not clear. A probe into the mechanisms is beyond the scope of this study. It is likely that the exercise-induced reduction in risk for stroke is a result of favourable changes in several biological systems. It has been suggested that a pathophysiological pathway between physical activity, LVH and stroke may involve left ventricular mass regression. 5 Left ventricular mass regression after 16 weeks of exercise training has been reported in hypertensive patients. 6 It is also well known that physical activity lowers blood pressure, 6 improves haemodynamics and cardiac performance, and favourably alters lipid and carbohydrate metabolism, adipose tissue distribution and body mass index. 18 There are also beneficial changes in hormonal, metabolic, neurologic, respiratory and psychological factors. 19,20 Thus, the beneficial effects of physical activity on other established stroke risk factors may affect their 499

6 500 interaction on the risk of stroke associated with LVH. 5 Limitations Only mortality data were available from this specific cohort. Thus, we could not provide predictors for stroke incidence. Although the collection of mortality data and causes of death was complete for all men for the subsequent 40 years, through periodic site visits by physicians and other scientists, the exact cause of death in several cases could not always be checked. Thus, the calculated causespecific hazard ratio, after adjusting for several predictors, may be over- or under-estimated. Although the proportion of electrocardiographically confirmed LVH observed in our cohort is similar to the prevalence of hypertrophy observed by other studies, 14,15 the imbalance of the sample sizes between men with and without LVH may limit our findings. The inability to segregate thrombotic from haemorrhagic strokes during the subsequent follow-up constitutes a limitation of the study. Physical activities could change through lifetime and, consequently, change blood pressures, lipids levels, etc. This might be considered as another limitation of the study. However, due to the nature of the cohort (fixed, rural), lifestyle habits related to work might be moderated, but rarely change through lifetime in this specific population. There is also a stronger probability that the exercise effects on risk for stroke are underestimated for two reasons: first, the physical activity level of the active group at baseline is likely to decline with advancing age. Second, the physically inactive group at baseline could either stay physically inactive or (some of them) may increase their physical activity level. Either condition is likely to obscure the true effect of physical activity on the risk reduction of stroke. Conclusion Stroke has a tremendous impact on public health, and the reduction of stroke burden through preventive interventions targeted at lifestyle factors seems essential. Based on 40 years (or person-years) of observation, we found that even moderate physical activity eliminates the risk of stroke associated with LVH. Our findings add to the wealth of evidence on the health benefits of exercise for a special population of patients. We recommend that health-care professionals encourage low-tomoderate exercise for patients with LVH. Acknowledgements We are grateful to Dr Dia Pitsavos, Mrs Klio Dontas and Dr Anastasia Katinioti for their substantial assistance during the 40 years follow-up. We are grateful particularly to the men of the villages of Corfu whose cooperation made this follow-up study feasible. Also, we thank the mayors of the villages, the register of the city of Corfu, the director and the staff of the San Marcos Medical Center as well as the microbiological laboratory of the Corfu s Prefectorial Hospital for their substantial support at the followup. The Seven Countries Study was funded by Grants HE 04697, HE 6090 and HE00278 from the National Heart, Lung and Blood Institute. References 1 Warlow CP. Epidemiology of stroke. Lancet 1998; 352: Kiely DK et al. Physical activity and stroke risk: the Framingham Study. Am J Epidemiol 1994; 139: Levy D et al. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990; 322: Benjamin EJ, Levy D. Why is left ventricular hypertrophy so predictive of morbidity and mortality? Am J Med Sci 1999; 317: Rodriguez C et al. Physical activity attenuates the effect of increased left ventricular mass on the risk of ischemic stroke. J Am Coll Cardiol 2002; 39: Kokkinos P et al. Effects of exercise on blood pressure and left ventricular hypertrophy in African-Americans with severe hypertension. N Engl J Med 1995; 333: Keys A et al. Epidemiological studies related to coronary heart disease: characteristics of men aged in seven countries. Acta Med Scand 1967; 460: Furberg CD et al. Major electrocardiographic abnormalities in persons aged 65 years and older. Cardiovascular Health Study Collaborative Research Group. Am J Cardiol 1992; 69: Anderson JT, Keys A. Serum cholesterol in serum and lipoprotein fractions: its measurement and stability. Clin Chem 1956; 2: Menotti A, Seccareccia F. Risk factors and mortality patterns in the Seven Countries Study. In: Toshima H, Koga Y, Blackburn H (eds) Lessons for Science from the Seven Countries Study. Springer: Tokyo. 1994, pp Menotti A et al. Twenty-year mortality and prediction of stroke in twelve cohorts of the Seven Countries Study. Int J Epidemiol 1990; 19: Panagiotakos DB et al. Forty-year ( ) of all cause mortality and its determinants: the Corfu cohort from the Seven Countries Study. Int J Cardiol 2003; 90: Summary of the Surgeon General s report addressing physical activity and health. Nutr Rev 1996; 54: Lee IM, Paffenbarger Jr RS. Physical activity and stroke incidence: the Harvard Alumni Health Study. Stroke 1998; 29:

7 15 Thrainsdottir IS et al. Survival and trends of occurrence of LVH, gender differences, ; the Reykjavik study. J Intern Med 2003; 253: Martnez MA et al. Prevalence of left ventricular hypertrophy in patients with mild hypertension in primary care; impact of echocardiography on cardiovascular stratification. Am J Hypertens 2003; 16: Bikkina M et al. Left ventricular mass and risk of stroke in elderly cohort. JAMA 1994; 272: Fletcher GF et al. Benefits and recommendations for physical activity programs for all Americans. A statement for health professionals by the Committee on Exercise and Cardiac Rehabilitation of the Council on Clinical Cardiology, American Heart Association. Circulation 1992; 86: Fletcher GF et al. Statement on exercise: benefits and recommendations for physical activity programs for all Americans. Circulation 1996; 94: Folkins CH, Sime WE. Physical activity training and mental health. Am J Psychol 1981; 36: