Many studies have shown that leisure-time physical activity and exercise have beneficial effects on the risk of cardiovascular disease.1–10 Associations between exercise and known risk factors for cardiovascular disease such as hypertension, adverse blood lipid profile, and elevated plasma fibrinogen have also been observed.11–19 Studies on the association between occupational physical activity and the risk of cardiovascular disease have shown different results, ranging from preventive effects20,21 to null effects2,9 and to indications of increased risk with higher level of activity.22–24 This inconsistency could be the result of differences in composition of occupations and types of physical activities represented in the study populations.
From a public health perspective, it would be valuable to estimate the interaction among exercise, occupational physical activity, and household physical activity, which has rarely been done. For example, do people with a heavy physical load at work benefit from exercise in their leisure time, or is it enough to be active at work?
The aim of the present study was to assess how different aspects of physical activity, through exercise, work, and household activities, are associated with the risk of acute myocardial infarction (MI). A special interest was to analyze the interaction among the various forms of physical activities.
The data come from the Stockholm Heart Epidemiology Program (SHEEP) study, which is a population-based case-control study conducted in Stockholm County, Sweden, 1992–1994. The study base comprised all Swedish citizens living in Stockholm County who were 45–70 years of age and free of clinically diagnosed MI. Male cases were identified during 1992–1993 and female cases during 1992–1994. During the period January to October 1992, the upper age limit for subjects was 65 years; from November 1, 1992, onward it was 70 years.
Myocardial infarction was defined using criteria set up by the Swedish Association of Cardiologists in 1991. These criteria included: 1) certain symptoms according to case history information, 2) specified changes in blood levels of the enzymes serum creatine kinase and serum lactate dehydrogenase, 3) specified electrocardiography changes, and 4) autopsy findings of myocardial necrosis. The diagnosis of acute MI required that 2 of the first 3 criteria be met, or that autopsy findings showed myocardial necrosis of an age compatible with the time of disease onset. Cases were identified from the coronary and intensive-care units at the internal medicine departments at all the emergency hospitals within the Stockholm County area, the hospital discharge register for the Stockholm County area, and through death certificates from the National Register of Death Causes at Statistics Sweden. If the patient died within 28 days of diagnosis, he or she was defined as having experienced a fatal MI.
One control per case was selected randomly from the study base after stratification for sex, age (5-year intervals), and hospital catchment area. If this control was not available, another control, who belonged to the study base at the time of the case occurrence, was randomly chosen. This procedure was repeated at most 4 times. Each control candidate was checked for history of MI. A more detailed description of the SHEEP study can be found in the report of the study by Reuterwall et al.25
Cases and controls were asked to fill in an extensive questionnaire on lifestyle factors. For cases identified by means of death certificates and cases who died before they could be contacted, the questionnaire was sent to a close relative. This was done at least 6 months after the date of death of the case subject. Up to 4 reminders were given in case of nonresponse. If the questionnaire was not completely filled out on receipt, a complementary telephone interview was conducted.
In total, 2246 cases and 3206 controls were invited to participate in the SHEEP study. Of the invited subjects, 1204 male and 550 female cases, and 1538 male and 777 female controls, answered the questionnaire and were included in this substudy. This corresponds to a participation rate of 78% for the cases and 72% for the controls.
In general, men had a higher participation rate compared with women, and nonfatal cases responded to a higher degree than relatives of fatal cases. Among the included cases in this substudy, 968 men and 413 women survived the first 28 days after diagnosis of their MI. The SHEEP study was approved by the Ethics Committee at Karolinska Institutet, Stockholm. All participants gave informed consent to participate.
Respondents were asked to report their average physical activity level during 5- to 10-year age intervals from 15 years of age to 69 years of age regarding physical activity through exercise, work, and household activities. In the analyses, we used the information regarding physical activity in the age interval to which the subject belonged at the time of the case occurrence.
Leisure-Time Physical Activity/Exercise
The respondents were asked to report how often in their leisure time they were engaged in physical activities or exercise that lasted for at least 30 minutes and that made them out of breath. They were asked to include walking or biking to and from work. The predefined activity levels were: very little exercise, occasional walks, some exercise every now and then (sometimes), and exercise on a regular basis (at least once per week). The persons who exercised on a regular basis were asked to specify the type of activity, intensity, and how often they were engaged in the different activities.
In the analysis, only the frequency of exercise was used with no differentiation between different activities or intensities. The categories of very little exercise and occasional walks were combined into one category (seldom) in the analyses. Those who answered that they exercised on a regular basis but did not answer the next question about frequency, intensity, and type of activity were added to the category of individuals who exercised once per week. This was done for 19 cases and 25 controls.
Occupational physical activity was evaluated using 4 aspects of physical activity.
Time Spent Sitting at Work
Three levels were predefined: sitting the major part of the working day, sitting approximately half of the working day, and sitting less than half of the working day.
The respondents were asked to report if they lifted or carried more than 5 kg for at least 2 hours per working day.
The respondents were asked to report if they lifted or carried more than 20 kg (women) or 30 kg (men) at least 5 times per working day.
Perceived Occupational Physical Activity
Perceived occupational physical workload was assessed using a 15-category scale (0–14) in which 0–1 was defined as very, very light and 13–14 was defined as very, very demanding. In the analysis, 5 categories were constructed: very light (0–2), light (3–5), moderate (6–8), strenuous (9–11), and very strenuous (12–14). As a result of the small numbers reporting very strenuous workload, the categories of strenuous and very strenuous were combined into one. Relatives of fatal cases were not asked to answer this question.
Physical activity through household work was assessed using one question.
Demanding Household Activities
The respondents were asked to report if their part of the household work included demanding activities such as snow shoveling or grass mowing.
A measure of a combination of exercise, sitting at work, and demanding physical household work was constructed and is labeled “total activity.” Subjects who reported leisure-time physical activity seldom or sometimes, sitting more than 50% of the working day, and not having demanding household work were classified as passive; those who reported regular leisure-time physical activity, sitting approximately 50% or less of the working day, and having demanding household work were classified as active; and subjects who reported that they were active in one or 2 of these aspects, but not in all, were classified as somewhat active.
Potential Confounding Factors
Age (5-year categories) and hospital catchment area were considered as confounding factors in all analyses on account of matching on these factors.
Smoking was defined as current smoking or no smoking. The no smoking category included those who had never smoked and exsmokers who had stopped smoking 2 years before inclusion in the study. Current smoking included subjects who reported smoking at the time of inclusion and those who reported giving up smoking less than 2 years before study inclusion.
Socioeconomic status (SES) was defined using information from the questionnaire about the occupation and educational level of each subject.26 The subjects were then classified as blue-collar or white-collar workers.
Body mass index (BMI) was calculated as weight (kg)/height2 (m). Overweight was defined as having a BMI greater than 25 kg/m2.
Fiber intake was used as an indicator of dietary habits. The fiber intake was derived from information regarding total food consumption, given in the questionnaire, and was energy-adjusted. The 25th percentile among controls was used as the cutoff. Subjects below the 25th percentile were considered as having a low fiber intake.
Alcohol consumption was classified into 6 categories according to grams of alcohol consumed per day: 0, 0.1–4.9, 5–14.9, 15–19.9, 20–29.9 and ≥30 g/day.
Odds ratios (ORs) together with 95% confidence intervals (CIs) were calculated through both conditional and unconditional logistic regression to estimate the relative risk of acute MI between those exposed and unexposed to various forms of physical activity. Some analyses included all cases, and others included only the nonfatal cases. All controls were included in the unconditional analyses, even if only the nonfatal cases were included.
We evaluated interaction between 2 exposures in terms of biologic interaction, ie, departure from additivity of effects.27
All analyses were conducted separately for men and women. Subjects with missing values on one or several variables were excluded from the analyses. Dummy variables were constructed and used in the statistical models when the variables consisted of 2 or more categories.
Table 1 provides selected characteristics of the study participants. Cases had a higher prevalence of several known risk factors for MI, including smoking, low socioeconomic status, BMI greater than 25 kg/m2, and low fiber intake, than the controls.
The frequencies of physical activities are shown in Table 2. Cases were less likely than controls to participate in leisure-time physical activity or exercise on a regular basis. However, a larger proportion of cases ranked their physical workload as strenuous or very strenuous, and they reported more repetitive and heavy lifting than controls. Among women, cases reported more frequently that they were sitting most of their working day. Among both men and women, cases reported less demanding household work than did controls, and men in general reported more demanding household activities, as measured here, than women.
When we analyzed the various combinations of physical activities among the controls, we found a tendency for subjects with more strenuous occupational workloads (moderate or strenuous, with lifting or carrying at work, or sitting less than half of their workday) were less likely to engage in leisure-time physical activity than those with sedentary work. Those who did not have demanding household work were also less likely to engage in leisure-time physical activities (data not shown).
The associations between the various forms of physical activity and risk of MI are shown in Table 3. The conditional and unconditional logistic regression models produced point estimates that were very similar, with a tendency for the estimates from the conditional analyses to be further from the null value. Only results from the unconditional analyses are presented here. All analyses were adjusted for age, hospital catchment area, SES, smoking, fiber intake, and alcohol consumption.
A distinct inverse dose-response pattern between leisure-time physical activity and the risk of acute MI was noted in both men and women. Protective effects were also seen among subjects who were sitting less than half of their working hours, at least in women. Furthermore, a lower risk of MI was observed among those who reported demanding household work. In contrast, men who perceived their physical occupational workload as moderate, strenuous, or very strenuous had an increased risk of a MI compared with those who perceived their workload as light or very light. Repetitive and heavy lifting or carrying at work were also associated with an increased risk of MI among both men and women. Additional adjustment for leisure-time physical activity and household activities further strengthened the association between lifting and carrying at work and risk of MI, with estimated relative risks for heavy lifting and carrying of 1.30 for men (95% CI, 1.02–1.66) and 1.83 for women (1.05–3.18). The measure of total activity showed a distinct pattern of lower relative risk of MI in subjects who were active compared with those who were passive in their leisure time, during work, and through their household work. In general, the associations between the various physical activities and MI were stronger when all cases were included than when only the surviving cases were analyzed.
With the exception of the results regarding lifting or carrying at work, additional adjustment for other types of physical activity did not change the results in any substantial way, nor did additional adjustment for body mass index and hypertension (data not shown).
Because of high levels of missing information on occupational activities, predominantly for subjects older than 65 years (the age of retirement in Sweden), we carried out additional analyses that excluded all individuals older than 65 years. The major results were unchanged.
Interactions between leisure-time physical activity and occupational physical activity, in relation to the risk of acute MI, are shown in Figure 1 for men and Figure 2 for women. The interaction between occupational and household physical activity are shown in Figures 3 and 4, respectively.
Leisure-time physical activity seemed to be beneficial for all persons irrespective of other forms of physical activities at work or home (Figs. 1 and 2). Furthermore, we found a beneficial effect of regular leisure-time physical activity among each category of perceived occupational physical activity (data not shown). Simultaneously, lack of leisure-time physical activity and having a job sitting most of the working day seemed to be particularly harmful, especially among women (Fig. 2). Lifting or carrying heavy things at work in combination with lack of regular leisure-time physical activity also resulted in increased risk for MI among women (Fig. 2), as did not having demanding household work in combination with sitting a lot at work (Fig. 4).
The results for interaction are based on analyzing all cases. The overall pattern was the same when only surviving cases were analyzed.
In this large population-based case-control study, we observed strong relations between physical activity and the risk of developing acute MI for both men and women. We found preventive effects from leisure-time exercise, walking or standing most of the working day, and demanding household activities. However, we also found that repetitive and heavy lifting or carrying at work, as well as perceived strenuous physical workload, increased the risk of acute MI. We used a combination of exercise, not sitting at work, and demanding household work as a measure of total activity. This total resulted in considerably lower risk of MI for those who were active compared with the passive individuals. When we evaluated interaction effects, we also found that lack of exercise and sitting a lot at work, as well as not having demanding household activities in combination with sitting a lot at work, was particularly harmful, at least in women.
We observed stronger associations when all cases, fatal as well as nonfatal, were included in the analyses compared with results based on only the surviving cases. One explanation for this pattern could be that the presence or absence of physical activity influenced the probability of surviving an acute myocardial event. This could occur, for example, if cases who exercised in their leisure time had a higher probability of surviving their MI compared with those who did not exercise. Another explanation would be that relatives of dead cases recall the subjects’ physical activity level differently than the cases would have recalled themselves. In a small methodologic study, done in connection with the SHEEP study, relatives of surviving cases tended to report lower level of leisure-time physical activity compared with the cases themselves (Reuterwall C, unpublished data, 1997).
Self-reported level of physical activity could lead to misclassification of exposure. If the misclassification were not dependent of the outcome, it would dilute the estimated odds ratios toward the null value when comparing the group with the highest exposure level with the lowest exposure group. It is likely that sedentary subjects in general tend to overestimate their activity level, whereas very fit individuals could underestimate their physical activity. If so, our results would be underestimating the true relationships between activity level and MI, at least regarding the preventive physical activity factors.
However, we cannot rule out the presence of differential misclassification of exposure as a result of the fact that information about physical activity was collected after the diagnosis of the cases. The knowledge of beneficial effects from physical activity and exercise is widespread in the general population, which might cause an underestimation of the true level of exercise among the cases. This kind of misclassification would result in an estimated relative risk further away from the null value compared with the true relative risk. Indeed, in a metaanalysis done by Eaton,1 case-control studies showed a greater risk reduction from physical activity compared with cohort studies. However, the pooled relative risk of being physically active compared with being sedentary was found to be 0.70 using data from 10 cohort studies and 2 case-control studies. The same relative risk was found in a 20-year follow-up study among middle-aged men in Sweden.2
If we dichotomize men by leisure-time physical activity into a sedentary group (seldom) and an active group (sometimes and regularly), we find that the adjusted relative risks of being active compared with sedentary is 0.67 (all cases) and 0.74 (only surviving cases). The corresponding relative risks among women were 0.53 and 0.64, respectively. The effect of occupational physical activity is not as clear and well-known as the effect of exercise. The fact that our results regarding leisure-time physical activity are in agreement with previous studies, as well as the fact that we observed both increased as well as decreased risk of MI associated with occupational activities, makes us believe that recall bias cannot explain all of our findings. Even so, we think that extra care should be given when interpreting the results from the analysis including all cases as a result of the fact that we had only proxy information regarding the exposure for the dead cases.
One advantage of our study is that it enables us to disentangle the effects of different aspects of physical activity: leisure-time, occupational, and household activities. Regarding occupational activity, we observed that walking or standing was associated with a decreased risk of MI, whereas lifting or carrying, as well as strenuous self-perceived workload, increased the risk of MI. These contrary findings could illuminate the differences found in previous studies on occupational physical activity. For example, Salonen et al.20 defined physical inactivity at work as “mainly sitting and not much walking,” and found that being physically active at work reduced the risk of death from ischemic heart disease. Both Stender et al.22 and Kristal-Boneh et al.24 used subjective ratings of effort level at work when defining occupational activity; they found that higher effort level tended to increase the risk of MI and cardiovascular disease mortality, respectively. Rosengren and Wilhelmsen2 used a combination of walking and lifting at work, and found no clear effect of occupational physical activity. Our results indicate that these 2 factors could work in opposite directions regarding the risk of MI. Overall, previous studies support our findings that different aspects of occupational activity could influence the risk of MI in different ways. Furthermore, these various results show the need for careful definition of what is actually meant by, and measured as, occupational physical activity.
We used information on physical activity according to the age interval of 5–10 years in which the cases occurred. Activity level might have changed as a result of previous illness, which could lead to lower levels of physical activity among cases. To account for this, we also performed analyses using information on physical activity corresponding to the 10-year age interval before the case occurrence. This yielded results similar to the ones presented here, with the exception of heavy lifting at work among women, in which no increased risk was observed when we looked at the age interval before the case occurrence.
In our analyses, we tried to take into account several potential confounding factors known to be associated both with MI and physical activity. However, we cannot preclude effects of residual confounding. For example, we had only a crude measure of fiber intake as an indicator of dietary habits. Although this measure might be considered to be imprecise, our definition of low fiber intake was associated with an increased risk of MI, as well as with low level of exercise. However, controlling for fiber intake did not affect our estimated relative risks in any major way.
We constructed a measure of total activity, which is a combination of leisure-time exercise, not sitting at work, and demanding household work. We have previously used this combination of activities as a measure of total activity in a cross-sectional study19 in which we found very strong associations with lower prevalence of hypertension, decreased high-density lipoprotein cholesterol, and high plasma fibrinogen. In the present study, we observed a distinct reduction in the risk of MI in the most active group compared with their passive counterparts.
We found signs of an interactive effect between lack of exercise and sitting at work, as well as between demanding household work and sitting at work, on the association with MI. We did not find any systematic biologic interaction between exercise and lifting or carrying at work, or demanding household work and lifting or carrying at work. Exercise, not sitting at work, and demanding household activities can be viewed as mainly aerobic activities, and it is likely that they work through the same causal mechanism regarding MI. In contrast, lifting and carrying heavy weights constitutes physical activity that is mainly anaerobic, and as such, could have their effect through another causal mechanism.
Based on the results from this study, it seems to be important to participate in at least some, and preferably several, aerobic activities to reduce the risk of MI. Manson et al.7 similarly found that “women who engaged in both walking and vigorous exercise had greater reductions in cardiovascular risks than those who did either one alone.” Although it is commonly believed that simply having a physically heavy job could prevent MI, our results indicate that some aspects of occupational physical activity (such as heavy lifting) could in fact be related to an increased risk. However, irrespective of occupational workload, leisure-time physical activity was found to be a strong preventive factor.
The participation rate was 78% for the cases and 72% for the controls. Attempts have been made earlier to evaluate whether nonrespondents differed from subjects who took part in the SHEEP study regarding smoking habits and occupation.25 The results showed that smoking was more common among nonrespondents compared with respondents, but not in the magnitude that it would influence the results to any greater extent. It is likely that those results would apply to physical activity as well.
In conclusion, physical activities that are mainly aerobic seemed to reduce the risk of acute MI, whereas repetitive and heavy lifting at work and self-perceived strenuous workload seemed to increase the risk. Synergistic preventive effects were found between exercise and not sitting at work, as well as between demanding household work and not sitting at work. In general, similar effects were found in both men and women.
We thank Annika Gustavsson for excellent management of the SHEEP database.
1. Eaton CB. Relation of physical activity and cardiovascular fitness to coronary heart disease, Part I: a meta-analysis of the independent relation of physical activity and coronary heart disease. J Am Board Fam Pract
2. Rosengren A, Wilhelmsen L. Physical activity protects against coronary death and deaths from all causes in middle-aged men. Evidence from a 20-year follow-up of the Primary Prevention Study in Göteborg. Ann Epidemiol
3. Dorn JP, Cerny FJ, Epstein LH, et al. Work and leisure-time physical activity and mortality in men and women from a general population sample. Ann Epidemiol
4. Blair SN, Brodney S. Effects of physical inactivity and obesity on morbidity and mortality: current evidence and research issues. Med Sci Sports Exerc
5. Sesso HD, Paffenbarger RSJr, Lee I-M. Physical activity and coronary heart disease in men. The Harvard Alumni Health Study. Circulation
6. Williams PT. Physical fitness and activity as separate heart disease risk factors: a meta-analysis. Med Sci Sports Exerc
7. Manson JE, Greenland P, LaCroix AZ, et al. Walking compared with vigorous exercise for the prevention of cardiovascular events in women. N Engl J Med
8. Tanasescu M, Leitzman MF, Rimm EB, et al. Exercise type and intensity in relation to coronary heart disease in men. JAMA
9. Wagner A, Simon C, Evans A, et al. Physical activity and coronary event incidence in Northern Ireland and France. The Prospective Epidemiological Study of Myocardial Infarction (PRIME). Circulation
10. Farahmand BY, Ahlbom A, Ekblom Ö, et al. Mortality amongst participants in Vasaloppet: a classical long distance ski race in Sweden. J Intern Med
11. Lakka TA, Salonen JT. Physical activity and serum lipids: a cross-sectional population study in Eastern Finnish men. Am J Epidemiol
12. Lakka TA, Salonen JT. Moderate to high intensity conditioning leisure-time physical activity and high cardiorespiratory fitness are associated with reduced plasma fibrinogen in eastern Finnish men. J Clin Epidemiol
13. Pols MA, Peeters PHM, Twisk JWR, et al. Physical activity and cardiovascular disease risk profile in women. Am J Epidemiol
14. Assanelli D, Bersatti F, Ferrari R, et al. Effect of leisure-time and working activity on principal risk factors and relative interactions in active middle-aged men. Coron Artery Dis
15. Kokkinos PF, Fernhall B. Physical activity and high density lipoprotein cholesterol levels: what is the relationship? Sports Med
16. Fung TT, Hu FB, Yu J, et al. Leisure-time physical activity, television watching, and plasma biomarkers of obesity and cardiovascular disease risk. Am J Epidemiol
17. Nagaya T, Kondo Y, Shibata T. Effects of sedentary work on physical fitness and serum cholesterol profile in middle-aged male workers. Int Arch Occup Environ Health
18. Ford ES. Does exercise reduce inflammation? Physical activity and C-reactive protein among US adults. Epidemiology
19. Fransson E, Alfredsson L, de Faire U, et al. Leisure time, occupational and household physical activity, and risk factors for cardiovascular disease in working men and women: the WOLF study. Scand J Public Health
20. Salonen JT, Slater JS, Tuomilehto J, et al. Leisure time and occupational physical activity: risk of death from ischemic heart disease. Am J Epidemiol
21. Menotti A, Lanti M. Coronary risk factors predicting early and late coronary deaths. Heart
22. Stender M, Hense HW, Doring A, et al. Physical activity at work and cardiovascular disease risk: result from the MONICA Augsburg study. Int J Epidemiol
23. Tuomi K. Characteristics of work and life predicting coronary heart disease. Finnish research project on aging workers. Soc Sci Med
24. Kristal-Boneh E, Harari G, Melamed S, et al. Association of physical activity at work with mortality in Israeli industrial employees: the CORDIS study. J Occup Environ Med
25. Reuterwall C, Hallqvist J, Ahlbom A, et al. Higher relative, but lower absolute risks of myocardial infarction in women than in men: analysis of some major risk factors in the SHEEP study. J Intern Med
26. Swedish socioeconomic classification. Reports on Statistical Co-ordination 1982:4.
Stockholm: Statistics Sweden; 1984.
© 2004 Lippincott Williams & Wilkins, Inc.
27. Rothman KJ. Epidemiology: An Introduction
. New York: Oxford University Press, Inc; 2002;173–180.