Physical Activity and Coronary Heart Disease Risk in Men
Does the Duration of Exercise Episodes Predict Risk?
Background—Physical activity is associated with a decreased risk of coronary heart disease (CHD). However, it is unclear whether the duration of exercise episodes is important: Are accumulated shorter sessions as predictive of decreased risk as longer sessions if the same amount of energy is expended?
Methods and Results—In the Harvard Alumni Health Study, we prospectively followed 7307 Harvard University alumni (mean age 66.1 years) from 1988 through 1993. At baseline, men reported their walking, stair climbing, and participation in sports or recreational activities. For each of the latter activities, they also reported the frequency and average duration per episode. During follow-up, 482 men developed CHD. In age-adjusted analysis, a longer duration of exercise episodes predicted lower CHD risk (P trend=0.04). However, after total energy expended on physical activity and potential confounders was accounted for, duration no longer had an independent effect on CHD risk (P trend=0.25); that is, longer sessions of exercise did not have a different effect on risk compared with shorter sessions, as long as the total energy expended was similar. In contrast, higher levels of total energy expenditure significantly predicted decreased CHD risk in both age-adjusted (P trend=0.009) and multivariate (P trend=0.046) analyses.
Conclusions—These data clearly indicate that physical activity is associated with decreased CHD risk. Furthermore, they lend some support to recent recommendations that allow for the accumulation of shorter sessions of physical activity, as opposed to requiring 1 longer, continuous session of exercise. This may provide some impetus for those sedentary to become more active.
Physical activity is clearly associated with a decreased risk of coronary heart disease.1 2 Unfortunately, most adults are physically inactive: >60% of US adults are not physically active on a regular basis, whereas 25% are not active at all.2 In an attempt to foster greater physical activity, the Centers for Disease Control and Prevention and the American College of Sports Medicine recently issued a new recommendation: “Every US adult should accumulate 30 minutes or more of moderate-intensity physical activity on most, preferably all, days of the week.”3 This recommendation also was developed to promote physical activity among those who do not enjoy or are unable to participate in vigorous activity.
The new prescription makes lesser demands than previous recommendations.4 The major differences are a concession to moderate intensity activity (as opposed to the prior requirement for vigorous exercise) and an allowance for the accumulation of short sessions of activity (as contrasted to the earlier need for 1 continuous, longer session). Whether intermittent, short sessions of activity are associated with benefit equal to that of 1 continuous, longer session remains controversial. The allowance for accumulated sessions of activity was based in part on the findings from 2 experimental studies in which investigators compared intermittent sessions with 1 continuous session of moderate to vigorous intensity exercise that required the same total energy expenditure.5 6 In both studies, subjects who exercised with the exercise broken up into intermittent sessions also experienced increases in physical fitness and improvements in blood lipid levels.
To date, no study has examined the association between physical activity of different durations and long-term outcomes such as coronary heart disease. It is important to directly assess long-term outcomes because improvements in short-term parameters, such as blood lipid levels, do not necessarily translate into subsequent benefits. For example, in the Heart and Estrogen/Progestin Replacement Study, estrogen-plus-progestin therapy did not decrease the early incidence of subsequent coronary events among women with coronary disease, despite significant improvements in their lipid profile.7
Therefore, we decided to investigate whether different durations of an exercise episode are associated with different risks of coronary heart disease. We specifically wanted to determine whether the duration of an exercise episode predicted risk, after the total energy expended on physical activity was accounted for.
Study subjects were from the Harvard Alumni Health Study, an ongoing study of men who matriculated as undergraduates at Harvard University between 1916 and 1950. This study was approved by the institutional review committee of the Harvard School of Public Health. Beginning in 1962, alumni have been sent periodic surveys that request information on health habits and medical history. For the present study, the 12 805 men who responded to a health survey in 1988 were eligible. We then excluded men who reported a history of cardiovascular disease or cancer (n=3543) because this would likely influence their physical activity levels, as well as those with missing physical activity information on the survey (n=623). Of the remaining 8639 men, we successfully followed 7373 (85%) through 1993 (ie, they returned another health survey in 1993 or were known to have died). We further excluded 66 men who did not provide information regarding coronary heart disease on the 1993 survey or who had died but for whom we were unable to obtain death certificates to determine the cause and date of death. This left 7307 men for the present analyses.
Assessment of Physical Activity
On the 1988 survey, we asked men to estimate the number of city blocks walked daily and the number of flights of stairs climbed daily, as well as to list any sports or recreational activities (up to 5) in which they had engaged during the past week.8 For each activity listed, we further inquired about the frequency and average duration per episode during the past week. This assessment of physical activity has been shown to be reliable and valid for large population studies.9 10 11 For example, for energy expenditure, the test-retest correlation coefficient during 1 month was 0.72, whereas questionnaire estimates compared with estimates from physical activity records yielded a correlation coefficient of 0.65.11
We estimated the total energy expended on physical activity in the following manner: walking 1 block daily required 235 kJ/wk (4.2 kJ=1 kcal), whereas climbing up and down 1 flight of stairs daily required 59 kJ/wk.12 The energy expended on each sport or recreation, in kilojoules per week, was estimated by taking into account the energy cost of the activity,13 14 as well as the frequency and duration of participation during the past week. We then summed kilojoules per week for walking, stair climbing, and all sports or recreation to estimate the total energy expenditure. In analyses, we classified men into 1 of 5 categories of total energy expenditure (Table 1⇓): <4200, 4200 to 8399, 8400 to 12 599, 12 600 to 16 799, and ≥16 800 kJ/wk.
We were primarily interested in the average time spent on each episode of physical activity for every sport or recreational activity reported. We hypothesized a priori that if a subject engaged in different activities (of like intensity), with each having a different average duration per episode, maximum benefit should accrue from the activity with the longest average duration per episode. Therefore, for each subject, we selected the activity with the longest average duration per episode and classified men according to this maximum average duration per episode (Table 1⇑): no sports or recreational activities reported, 1 to 15, 16 to 30, 31 to 45, 46 to 60, or >60 minutes.
Determination of Coronary Heart Disease
We sent another health survey to subjects in 1993 that included questions on whether a physician had ever diagnosed angina pectoris or myocardial infarction and whether the subject had ever undergone CABG or PTCA. For those who responded in the affirmative, we further enquired about the year of onset or procedure. Self-reported coronary heart disease is believed to be valid in this population.15
In addition, the Harvard Alumni Office maintains listings of deceased alumni. Using this information, we obtained copies of official death certificates. To determine the completeness of mortality follow-up, we searched the National Death Index, a national compilation of decedents,16 from 1988 through 1992 for 500 randomly chosen subjects who were deemed alive according to our records. Only 2 were positively identified as dead, for a mortality follow-up rate of 99.6%.
Using these sources, we determined whether subjects had developed coronary heart disease after the return of the 1988 health survey through 1993. The year of occurrence was the earliest year reported for any of the conditions given or the year of death for men noted to have coronary heart disease as either the underlying or a contributing cause of death on their death certificates.
We used proportional hazards regression to estimate the relative risks of coronary heart disease associated with the maximum average duration (hereafter referred to as “duration”) per episode of sport or recreational activity.17 There was no evidence that proportional hazards assumptions were violated. In initial analyses, we adjusted the relative risks for differences in age. Because longer durations of activity yield higher energy expenditures, which are associated with lower coronary heart disease risk,2 3 we then also adjusted for total energy expenditure. That is, we sought to determine whether different durations per episode of activity differentially predicted coronary heart disease risk among subjects with the same total energy expenditure. Next, we further controlled for differences in these potential confounders: cigarette smoking; hypertension; diabetes; early parental death; intake of vitamin or mineral supplements; alcohol, red meat, and vegetable consumption; and participation in vigorous activities (≥6 METs). In analyses, variables were categorized as shown in Table 1⇑ with the exceptions of age (in continuous years) and body mass index (grouped in approximate fifths: <22.5, 22.5 to <23.5, 23.5 to <24.5, 24.5 to <26.0, and ≥26.0 kg/m2). We calculated 95% CIs for estimated relative risks and used 2-tailed tests of significance.
To minimize potential bias due to men who may have decreased their activity level at baseline because of preclinical heart disease, we conducted a separate analysis that excluded the first 2 years of follow-up. Finally, we examined whether the association of exercise duration with coronary heart disease risk varied by age, total energy expenditure, or intensity of the sports or recreational activities in which men engaged. We did so by including an interaction term with exercise duration for each of these variables in 3 separate multivariate models.
Table 1⇑ provides a description of the study subjects. Men were older, with a mean age of 66.1 years at study entry. A sizable proportion of men (35.4%) reported episodes of activity lasting >1 hour in the past week. These men spent their time playing tennis (24.1%), golfing (13.6%), gardening and working in the yard (12.2%), skiing (12.0%), and assorted activities (remainder).
When we examined the baseline characteristics of men according to the duration of their exercise episodes, men who did not report any sports or recreational activities had a worse coronary profile than did those who reported any activities (data not shown). However, there was no clear pattern across categories of duration among men who reported at least 1 activity, with the exception of the physical activity variables. As would be expected, with the longer duration of exercise episodes, total energy expenditure increased, as did the proportion of men who engaged in vigorous activities and the number of leisure-time activities reported.
The association of duration per episode of sport or recreational activity with coronary heart disease risk is shown in Table 2⇓. During the 5 years of follow-up (1988 to 1993), 482 men developed coronary heart disease. The number of men who developed coronary heart disease in each of the 6 categories of exercise duration was 155, 24, 61, 32, 55, and 155, respectively. In age-adjusted analysis, there was a significant trend (P=0.04) of decreasing coronary heart disease risk with increasing time per episode spent on sports or recreational activities.
However, when we took into account the total energy expended on walking, climbing stairs, and participation in sports or recreational activities, the duration per episode no longer predicted coronary heart disease risk (P trend=0.68; Table⇑ 2). That is, among men who expended similar total amounts of energy in physical activity, a longer duration per episode of activity did not further decrease coronary heart disease risk. Furthermore, men who engaged in sports or recreational activities, regardless of the duration per episode, did not experience lower risks of coronary heart disease than did those who only walked and climbed stairs but did not report additional participation in sports or recreational activities, provided the total energy output was similar. (Among the 1874 men who did not report any sports or recreational activities, all except 94 walked or climbed stairs.) With additional adjustment for body mass index, smoking, hypertension, diabetes mellitus, early parental death, vitamin or mineral supplements, alcohol intake, diet, and participation in vigorous activities, the duration per episode continued to show no association with coronary heart disease risk (P trend=0.25). In a sensitivity analysis, we refined the longest duration category by defining 2 groups: 61 to 120 and >120 minutes; this did not change the findings (P trend=0.21; data not shown).
We then analyzed the subgroup of men who reported at least 1 sport or recreational activity. After adjustment for total energy expenditure, as well as the potential confounders listed earlier, the relative risks of developing coronary heart disease associated with duration per exercise episode of 1 to 15, 16 to 30, 31 to 45, 46 to 60, and >60 minutes were 1.00 (referent), 0.88 (95% CI 0.52 to 1.48), 0.95 (0.52 to 1.74), 1.02 (0.59 to 1.75), and 1.06 (0.63 to 1.79), respectively (P trend=0.41).
In contrast, the total energy expended on physical activity consistently predicted a decreased risk of coronary heart disease (Table 3⇓). Of the 482 men who developed coronary heart disease, 182, 98, 74, 47, and 81, respectively, fell into the 5 categories of total energy expenditure. In an age-adjusted analysis, greater amounts of energy expended on walking, climbing stairs, and participation in sports or recreational activities strongly predicted decreased coronary heart disease risk (P trend=0.009). With additional adjustment for duration per episode of activity, there was an inverse association with risk that was of borderline significance (P trend=0.09). However, with further adjustment for the other potential confounders described earlier, the inverse association again attained significance (P trend=0.046). In all regression models, men who expended >16 800 kJ/wk in physical activity experienced a significantly lower (on the order of 30% to 40%) risk of coronary heart disease compared with those expending only <4200 kJ/wk.
When we excluded the first 2 years of follow-up to minimize potential bias resulting from men who may have decreased their physical activity at study entry because of symptoms from as-yet-undiagnosed heart disease, findings were little changed (data not shown). Finally, we investigated whether the association of the duration of an exercise episode with coronary heart disease risk differed among subgroups of men defined on the basis of various characteristics. We did not observe any effect modification among men of different ages (P=0.38), among those with different levels of total energy expenditure (P=0.14), or between those who did and those who did not engage in vigorous activities (P=1.00).
These prospective data indicate that longer durations spent on each episode of exercise are not associated with decreased coronary heart disease risk compared with shorter durations, once total energy expenditure is taken into account. Stated in another way, it appears that the accumulation of shorter sessions of activity is associated with equivalent benefit (at least, with regard to coronary heart disease risk) compared with longer sessions, as long as the total amount of energy expended is similar. Even men who only walked and climbed stairs and did not report additional participation in sports or recreational activities fared as well as those who engaged in sports or recreational activities, provided their total energy output was similar. In analyses, we chose not to adjust for the frequency of physical activity because the combined contribution of frequency and duration of exercise episodes yields total energy expenditure. Thus, by the inclusion of terms for duration and total energy expenditure in regression models, the analyses have indirectly accounted for frequency.
No previous study has assessed the association of duration of exercise episodes, apart from its contribution to total energy expenditure, with coronary heart disease risk. Two trials investigated the effects of short versus long sessions of exercise on cardiovascular risk factors. Ebisu5 reported that young men who ran 6 miles a day, whether in 1, 2, or 3 sessions, experienced similar increases in physical fitness after 10 weeks, whereas HDL-cholesterol levels increased most in the 3-session group.5 In another experiment, DeBusk et al6 observed that physical fitness improved after 8 weeks among middle-aged men who jogged 30 minutes a day, whether in 1 or 2 sessions, although the former group had greater improvement. Although it is unclear whether these findings extend to more moderate-intensity activities (eg, walking), the results from these experimental studies may explain why we found no effect of the duration of an exercise session on coronary heart disease risk, after accounting for total energy expenditure.
In an observational study with data that were collected at a single examination, Mensink et al18 found that among men and women aged 25 to 69 years, the frequency of exercise appeared to better predict cardiovascular risk factors compared with exercise intensity or the total time spent exercising. However, total energy expenditure was not held constant, so this finding may instead reflect the association with the amount of energy expended. Recently, 2 trials reported on changes in cardiovascular risk factors among subjects assigned to an intervention that promoted lifestyle physical activity (ie, use of recent physical activity recommendations) or an intervention that promoted traditional exercise prescriptions.19 20 With both interventions, small but significant increases occurred in both physical activity and fitness. However, data regarding actual durations per episode of physical activity were not provided. Presumably, subjects assigned to the former group would likely accumulate short sessions of activity, whereas those assigned to the latter group would likely incline toward exercising in a continuous, longer session. After 1 to 2 years of follow-up, beneficial changes were observed for fat mass and blood pressure levels in both intervention groups; however, benefits for the lipid profile were less clear.
The interpretation of results from the present study must be tempered by some potential limitations. One alternate explanation for the lack of association between the duration of an exercise episode and coronary heart disease risk might simply be that duration was imprecisely reported, leading to misclassification. Furthermore, each exercise episode reported by a subject may not have involved continuous activity but rather fragmented spurts. For example, men who reported playing tennis or golf for >1 hour each episode might have taken breaks for conversation or other reasons, as is likely in a social, rather than competitive, context. In addition, we determined the duration of exercise episodes on the basis of sports and recreational activities only; we did not do so for walking or climbing stairs. Finally, it would be useful to examine the potential health benefits of initially sedentary men taking up short sessions of physical activity. Unfortunately, we were unable to reliably investigate this because only 5% of men had changed from being initially sedentary before study entry to participating in sports or recreational activities at baseline in 1988.
Subjects in the present study were of a higher educational and socioeconomic status, with the vast majority being university graduates. However, the biologic effects of physical activity are likely to apply to the general population as well. Physical activity lowers blood pressure levels,21 increases insulin sensitivity,22 and favorably influences lipid profiles.23 24 It also enhances cardiac mechanical and metabolic function25 and improves hemostatic factors, decreasing platelet aggregation26 and increasing fibrinolytic activity.27 These beneficial effects likely explain the decreased coronary heart disease risk with increased physical activity.
In conclusion, these data clearly indicate that physical activity is associated with decreased coronary heart disease risk. Furthermore, they lend some support to current recommendations that allow for the accumulation of shorter sessions of physical activity, as opposed to requiring 1 longer, continuous session of exercise. Physicians should advise patients to be physically active to decrease their coronary heart disease risk. Physical activity does not have to be arduously long to be beneficial; even short sessions lasting 15 minutes appear to be helpful. This may provide some impetus for sedentary individuals to take up physical activity.
This work was supported by research grants HL-34174 from the National Heart, Lung, and Blood Institute and CA-44854 from the National Cancer Institute, US Public Health Service. We are grateful to Stacey DeCaro, Sarah E. Freeman, Tina Y. Ha, Martha J. Higgins, Rita W. Leung, Doris C. Rosoff, and Alvin L. Wing for their help with the College Alumni Health Study.
Reprint requests to I-Min Lee, MBBS, ScD, Department of Epidemiology, Harvard School of Public Health, 677 Huntington Avenue, Boston, MA 02115.
This is report No. LXXVII in a series on chronic disease in former college students (College Alumni Health Study).
- Received January 11, 2000.
- Revision received March 23, 2000.
- Accepted March 27, 2000.
- Copyright © 2000 by American Heart Association
Berlin JA, Colditz GA. A meta-analysis of physical activity in the prevention of coronary heart disease. Am J Epidemiol. 1990;132:612–628.
US Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion; 1996.
American College of Sports Medicine. Guidelines for Graded Exercise Testing and Exercise Prescription. 3rd ed. Philadelphia, Pa: Lea & Febiger; 1985.
Ebisu T. Splitting the distance of endurance running: on cardiovascular endurance and blood lipids. Jpn J Phys Educ. 1985;30:37–43.
Lee I-M, Paffenbarger RS Jr, Hsieh C-C. Time-trends in physical activity among college alumni, 1962–1988. Am J Epidemiol. 1992;135:915–925.
Paffenbarger RS Jr, Wing AL, Hyde RT. Physical activity as an index of heart attack risk in college alumni. Am J Epidemiol. 1978;108:161–175.
Boyle CA, Decoufle P. National sources of vital status information: extent of coverage and possible selectivity in reporting. Am J Epidemiol. 1990;131:160–168.
Cox DR. Regression models and life tables. J R Stat Soc (B). 1972;34:187–220.
Mayer-Davis EJ, D’Agostino R Jr, Karter AJ, et al. Intensity and amount of physical activity in relation to insulin sensitivity: the Insulin Resistance Atherosclerosis Study. JAMA. 1998;279:666–674.
Froelicher VF. Exercise, fitness, and coronary heart disease. In: Bouchard C, Shephard RJ, Stephens T, et al, eds. Exercise, Fitness, and Health: A Consensus of Current Knowledge. Champaign, Ill: Human Kinetics Publishers Inc; 1990;429–450.
Rauramma R, Salonen JT, Seppanen K, et al. Inhibition of platelet aggregability by moderate-intensity physical exercise: a randomized clinical trial in overweight men. Circulation. 1986;74:939–944.
Szymanski LM, Pate RR, Durstine JL. Effects of maximal exercise and venous occlusion on fibrinolytic activity in physically active and inactive men. J Appl Physiol. 1994;77:2305–2310.