Change in Physical Activity and Sitting Time After Myocardial Infarction and Mortality Among Postmenopausal Women in the Women's Health Initiative‐Observational Study
Background How physical activity (PA) and sitting time may change after first myocardial infarction (MI) and the association with mortality in postmenopausal women is unknown.
Methods and Results Participants included postmenopausal women in the Women's Health Initiative‐Observational Study, aged 50 to 79 years who experienced a clinical MI during the study. This analysis included 856 women who had adequate data on PA exposure and 533 women for sitting time exposures. Sitting time was self‐reported at baseline, year 3, and year 6. Self‐reported PA was reported at baseline through year 8. Change in PA and sitting time were calculated as the difference between the cumulative average immediately following MI and the cumulative average immediately preceding MI. The 4 categories of change were: maintained low, decreased, increased, and maintained high. The cut points were ≥7.5 metabolic equivalent of task hours/week versus <7.5 metabolic equivalent of task hours/week for PA and ≥8 h/day versus <8 h/day for sitting time. Cox proportional hazard models estimated hazard ratios and 95% CIs for all‐cause, coronary heart disease, and cardiovascular disease mortality. Compared with women who maintained low PA (referent), the risk of all‐cause mortality was: 0.54 (0.34–0.86) for increased PA and 0.52 (0.36–0.73) for maintained high PA. Women who had pre‐MI levels of sitting time <8 h/day, every 1 h/day increase in sitting time was associated with a 9% increased risk (hazard ratio=1.09, 95% CI: 1.01, 1.19) of all‐cause mortality.
Conclusions Meeting the recommended PA guidelines pre‐ and post‐MI may have a protective role against mortality in postmenopausal women.
Cardiovascular disease (CVD) is responsible for 1 of every 3 deaths in the United States, with coronary heart disease (CHD) causing ≈1 of every 7 deaths.1 The estimated annual incidence of myocardial infarction (MI) is 660 000 new attacks, with an average age at first MI of 65.1 years for men and 72.0 years for women.1 Approximately 15% of those who experience an MI will die of it.1 Smoking cessation, regular physical activity (PA), and dietary changes may reduce mortality by 20% to 35% in MI survivors.2 Survival after an initial MI is increasing because of the success of secondary prevention programs and promotion of cardiac rehabilitation programs.3, 4, 5 Thus, MI survivors are a growing group and may be motivated to change their behavior to prevent future morbidity and mortality.
An estimated 5.8% of global CHD mortality is attributable to physical inactivity after accounting for other CVD risk factors.6, 7 The role of PA, sitting time, and incident CVD risk has been previously assessed in the WHI‐OS (Women's Health Initiative‐Observational Study) by Manson et al8 and Chomistek et al.9 These studies found that higher levels of PA (metabolic equivalent of task hours [MET‐h]/week) and lower amounts of prolonged sitting time were associated with a decreased risk of incident coronary events, but their analyses focused on women without past CVD. One previous study investigated change in PA after first MI in a population of 406 men and women from the Corpus Christi Heart Project cohort10 and found a reduction in all‐cause mortality for those who remained active, increased activity, or decreased activity post‐MI compared with those who were sedentary pre‐ and post‐MI. However, this study was relatively small and did not examine sitting time, which may be a distinct construct from lack of PA.11
While there is substantial evidence that physical activity lowers risk of incident CVD, less is known on the benefits of PA for decreasing mortality post‐MI. Additionally, more research is needed on the potential effect of changing other lifestyle habits, such as sitting time, following an MI. Therefore, the purpose of the current study was to examine change in PA and sitting time among survivors of a first MI on all‐cause, CHD, and CVD mortality in the WHI‐OS.
Women's Health Initiative
The goal of the WHI was to investigate major causes of morbidity and mortality in a nationally representative cohort of postmenopausal women. The WHI‐OS cohort included 93 676 postmenopausal women, aged 50 to 79 years at study entry, from 40 US clinical centers and enrolled between 1993 and 1998.12 Details on the reliability and measurement of baseline characteristics in the WHI have been previously published.13 All study sites obtained informed consent and institutional review board approval.
We identified 3129 women who experienced a first MI during follow‐up out of the original 93 676 women in the WHI‐OS. Nonfatal MI was confirmed based on standard diagnostic criteria for electrocardiography changes, elevated cardiac enzymes, or both.14 We then excluded 979 women with baseline prevalent CHD, stroke, or cancer, 61 women missing information on sitting or PA at baseline, and 89 women who reported they were unable to walk one block at baseline. Furthermore, 1144 women were excluded from the PA analysis because of MI occurring after the last measurement of PA and 1467 women were excluded from the sitting time analysis because of MI occurring after the last measurement of sitting time. Thus, 856 women were included in the PA analysis and 533 in the sitting‐time analysis.
Baseline characteristics were self‐reported on questionnaires, including demographics, medical history, diet, smoking, and other lifestyle factors. Participants completed follow‐up assessments periodically after enrollment. The study protocol was approved by the institutional review board of each site, and all women provided written informed consent.
Recreational PA was self‐reported at baseline and annually through year 8 by detailed questionnaire. Information on walking frequency, duration, and pace, specifically outside of the home, was asked as well as information on other types of activity (mild, moderate, and strenuous). Examples of mild activity included slow dancing, bowling, and golf. Examples of moderate activity were biking outdoors, using an exercise machine, calisthenics, easy swimming, and popular or folk dancing. Examples of strenuous activity were aerobics, aerobic dancing, jogging, tennis, and swimming laps. Each type of activity was assigned an MET score on the basis of energy cost,15 and PA‐related energy expenditure (MET‐h/week) was calculated as the summed product of the frequency, duration, and intensity of the activity. A random sample of 536 participants had second measurements conducted on all PA ≈10 weeks after baseline. Test‐retest reliability for the PA variables was 0.53 to 0.72 with the intraclass correlation 0.77 for total PA.16
Sitting time was assessed by questionnaire at baseline and twice during follow‐up (year 3 and year 6) with the following question: “During a usual day and night, about how many hours do you spend sitting? Be sure to include the time you spend sitting at work, sitting at the table eating, driving or riding in a car or bus, and sitting up watching TV or talking.” Eight categories were provided for the response, ranging from less than 4 hours to 16 or more hours.
To represent long‐term exposure,17 the cumulative average for PA, sitting time, and walking was calculated from questionnaires at baseline, year 3 to year 8 for PA and walking and baseline, year 3, and year 6 for sitting time. Change in PA, walking, and sitting time were calculated as the difference between the cumulative average immediately following MI and the cumulative average immediately preceding MI. Thus, pre‐MI PA is the average of all measurements before the MI, and post‐MI PA is the average of all measurements after the MI. For physical activity, we used 7.5 MET‐h/week as the cut point, which is equivalent to accumulating 150 min/week of moderate‐intensity exercise, consistent with the current PA guidelines.18 Walking categories were determined based on the median of the data, with 3.5 MET‐h/week being equivalent to 1 h/day of brisk walking. Sitting‐time categories were determined based on previous research in the aging population19 using 8 h/day as the cut point. The 4 categories of change were: maintained low, decreased, increased, and maintained high. The cut points were as follows: PA, ≥7.5 MET‐h/week (high) versus <7.5 MET‐h/week (low); walking, ≥3.5 MET‐h/week (high) versus <3.5 MET‐h/week (low); and sitting time, ≥8 h/day (high) versus <8 h/day (low).
Ascertainment of End Points
The primary end points for this analysis were all‐cause, CHD, and CVD mortality. Underlying cause of death was identified through medical records, death certificates, and autopsy reports.14 Physicians blinded to the exposure information adjudicated the event. CHD and CVD mortality were confirmed through documentation on hospital or autopsy records of coronary disease or CVD was listed as the cause of death on death certificates.
All analyses were performed using SAS statistical software (version 9.4; SAS Institute Inc, Cary, NC). Eligible participants contributed person‐time from return of the questionnaire immediately following first clinical MI to death from any cause, loss to follow‐up, or August 29, 2014, whichever occurred first. Baseline characteristics were compared according to categories of change in PA with a chi‐squared test for categorical variables and generalized linear model for continuous variables. We examined the association between pre‐ and post‐MI levels of PA, walking and sitting time, and risk of all‐cause, CHD, and CVD mortality using Cox proportional hazards models to estimate hazard ratios (HRs) and 95% CIs. The categories were: >20, 7.5 to 20, 1.8 to 7.4, and ≤1.7 MET‐h/week for physical activity; ≥3.5 and <3.5 MET‐h/week for walking; and ≤5, 5.1 to 7.9, and ≥8 h/day for sitting time.
We also examined the association between categories of change in PA, walking, and sitting time with risk of all‐cause, CHD, and CVD mortality using the 4 categories of change described above: maintained low, decreased, increased, and maintained high. Additionally, to further explore the association between change in PA, walking, and sitting time with all‐cause mortality, we assessed change in each exposure continuously, stratified by pre‐MI levels. Finally, conditional relative PA was calculated to express the relative PA change from pre‐MI to post‐MI, which was uncorrelated with PA pre‐MI, to estimate the independent contribution of PA change on mortality.20 Conditional relative PA was calculated as residuals by regressing post‐MI PA on pre‐MI PA.
Three models were used to evaluate the association between change in PA and sitting time after an MI and mortality. The first model was stratified by age only. The second model additionally included race, education, income, marital status, heart failure, cigarette smoking, family history of MI, alcohol consumption, Healthy Eating Index diet score, hours of sleep, depression, sitting time (for PA model), PA (for sitting time model), and vigorous PA (for walking model). The final model also adjusted for diabetes mellitus, hypertension, hyperlipidemia, body mass index, coronary artery bypass graft, and percutaneous transluminal coronary angioplasty. To account for potential bias attributed to severity of MI, in addition to adjusting for heart failure, we conducted a sensitivity analysis by excluding women who died within 2 years of MI and compared with the original analysis.
Baseline characteristics according to category of change in PA are presented in Table 1. Women who maintained high PA pre‐MI and post‐MI were more likely to have attended college, be married, have a lower body mass index, and have an overall healthier diet at baseline compared with women who maintained low PA pre‐ and post‐MI.
During a median follow‐up of 7.2 years, 265 of the 856 women included in the PA analysis and 225 of the 533 women included in the sitting‐time analysis died. In multivariable‐adjusted analyses, PA pre‐ and post‐MI was inversely associated with all‐cause and CVD mortality (Table 2). Compared with women who reported >20 MET‐h/week of PA pre‐MI, women with ≤1.7 MET‐h/week pre‐MI had an HR of 1.61 (95% CI, 1.00, 2.60; P for trend, <0.01) for all‐cause mortality and 3.29 (95% CI, 2.03–5.33) for post‐MI PA ≤1.7 MET‐h/week. Similarly, for CVD mortality, women with ≤1.7 MET‐h/week pre‐MI had a HR of 2.00 (95% CI, 0.92, 4.34; P for trend=0.02) and 3.69 (95% CI, 1.61–8.47; P for trend, 0.001) for post‐MI PA compared with women reporting >20 MET‐h/week. Walking <3.5 MET‐h/week compared with ≥3.5 MET‐h/week pre‐MI was associated with an increased risk of all‐cause mortality (HR=1.44; 95% CI, 1.02–2.03; P for trend=0.04).
Means and SDs of PA pre‐ and post‐MI by categories of change in PA are presented in Table S1. The average weekly PA for those who decreased PA was 15.33±7.87 MET‐h/week pre‐MI and 3.01±2.44 MET‐h/week post‐MI. For those who increased PA, average weekly PA was 3.47±2.34 MET‐h/week pre‐MI and 14.47±7.32 MET‐h/week post‐MI. Age and multivariable‐adjusted HRs for all‐cause, CHD, and CVD mortality by categories of change in PA, sitting time, and walking are presented in Table 3. Using maintained low PA as the reference group, the HRs for all‐cause mortality for women who increased PA and maintained high PA were 0.54 (95% CI, 0.34–0.82) and 0.52 (95% CI, 0.36–0.73), respectively. Again, using maintained low PA as the reference group, the HRs for CHD and CVD mortality for women who maintained high PA were 0.26 (95% CI, 0.11–0.60) and 0.41 (95% CI, 0.24–0.71), respectively. When we investigated walking specifically, the all‐cause mortality HR for women who maintained high walking was 0.65 (95% CI, 0.42–0.99), compared with women who maintained low walking pre‐ and post‐MI. Change in sitting time as a categorical variable was not significantly associated with all‐cause, CHD, or CVD mortality.
We then examined continuous change in PA, sitting time, and walking, stratified by pre‐MI levels (Table 4). Among women who reported PA <7.5 MET‐h/week pre‐MI, there was an 8% decreased risk (HR=0.92; 95% CI, 0.88, 0.96) of all‐cause mortality, 16% decreased risk of CHD mortality (HR=0.84; 95% CI, 0.76, 0.93), and 9% decreased risk of CVD mortality (HR=0.91; 95% CI, 0.86, 0.96) for a 1 MET‐h/week increase in PA. Among those women reporting PA ≥7.5 MET‐h/week pre‐MI, there was an 8% decreased risk of CHD mortality for a 1 MET‐h/week increase in PA. HRs for continuous change in PA were not statistically significant among women who reported PA levels that met the PA guidelines pre‐MI for all‐cause and CVD mortality. Women who had pre‐MI levels of sitting time <8 h/day had a 9% higher risk (HR=1.09; 95% CI: 1.01, 1.19) of all‐cause mortality for every 1 h/day increase in sitting time. The HR for continuous change in sitting time was not statistically significant among women who reported sitting time >8 h/day pre‐MI. Finally, results for walking were similar to that for PA, but stronger. Women who had pre‐MI levels of walking <3.5 MET‐h/week had a 19% decreased risk of all‐cause mortality (HR=0.81; 95% CI, 0.72, 0.91), 35% decreased risk of CHD mortality (HR=0.65; 95% CI, 0.45–0.94), and 34% decreased risk of CVD mortality (HR=0.66; 95% CI, 0.52–0.83) for a 1 MET‐h/week increase in walking.
Similar results were found when conditional relative PA was assessed in relation to all‐cause, CHD, and CVD mortality in both age‐ and multivariate‐adjusted models (Table S2). After excluding women who died within the first 2 years post‐MI, the results were similar.
In this prospective study of postmenopausal MI survivors in the United States, maintaining at least 7.5 MET‐h/week of PA pre‐ to post‐MI was associated with lower risk of all‐cause, CHD, and CVD mortality. Additionally, individuals who increased PA after a first MI had a lower risk of all‐cause mortality compared with women who maintained low PA. Finally, when examined continuously, increased sitting time post‐MI was associated with an increased risk of all‐cause mortality among individuals with low sitting time (<8 h/day) pre‐MI.
Only 1 study to date has examined the association between change in PA after a first MI.10 Thus, the current analysis, which assesses the association between change in PA, sitting time, and walking and all‐cause, CHD, and CVD mortality provides a significant contribution to the current literature. Our results are consistent with a recent study that found10 lower risk of all‐cause mortality in individuals who maintained high PA, or increased PA after a first MI compared with those who maintained low PA pre‐ and post‐MI. Although the earlier study included both men and women, the number of women was small (150) and the instrument used to measure PA was not validated. Moreover, sitting time was not considered a separate construct from physical inactivity in the analysis and walking was not examined in addition to total PA. We extend this research by examining the association of change in sitting time and PA with mortality as independent entities in postmenopausal women. Additionally, by using the cumulative average of PA, sitting time, and walking pre‐ and post‐MI, we were able to capture long‐term exposure.17 Because of the timing and frequency of the exposure measurements throughout the study, using cumulative average is more representative of long‐term exposure for PA and walking, but not necessarily for sitting time, because it was only assessed 3 times throughout follow‐up.
When examined categorically, change in sitting time was not associated with mortality in the main analysis. However, when examined continuously, we did observe an increased risk of all‐cause mortality for every 1‐h/day increase in sitting time in those women who sat for 8 or less hours/day pre‐MI. Our study is the first population‐based study to investigate change in sitting time after a first MI.
Several mechanisms may explain why our results were stronger for CVD mortality compared to all‐cause mortality. Prolonged sitting has been associated with increased total cholesterol, triglycerides, waist circumference, decreased glucose uptake, and decreased skeletal muscle lipoprotein lipase activity.21, 22, 23 Previous research has shown that repeated bouts of prolonged sitting result in low shear rates leading to endothelial dysfunction, which has also been linked to vascular mortality.24 Breaking up sitting time with light PA has been shown to counteract the adverse effects on the endothelium.24 Last, increased lifestyle PA, including walking, post‐MI has been associated with improvement in cardiorespiratory fitness,25 which has been shown to have prognostic benefit even in patients with reduced left ventricular ejection fraction.26
A potential source of bias could be that individuals did not increase their PA or decrease their sitting time because of the severity of their first MI. Unfortunately, we did not have information on MI severity; however, we did exclude women unable to walk at least 1 block and controlled for congestive heart failure, which was updated throughout the study within our multivariate models in an attempt to adjust for severity. We also conducted a sensitivity analysis excluding women who died within 2 years of their MI, and results were similar to our main analysis. When excluding 47 women who died within 2 years of their MI, the HRs in the fully adjusted model for all‐cause mortality in women who maintained high PA was 0.55 (95% CI, 0.37–0.81), compared with women who maintained low PA. Additionally, the women excluded from the analysis for missing or inadequate data did not differ from those included in the analysis on baseline PA, baseline sitting time, age, smoking status, body mass index, or Healthy Eating Index diet quality.
Our study had several strengths, including its prospective design, the large multiethnic cohort of postmenopausal women, wide variety of information on covariates, and detailed longitudinal information on PA and sitting time. This is the first study to investigate the association of change in sitting time and walking after a first MI and subsequent mortality.
The limitations of the present study include some aspects of exposure assessment. First, sitting time was measured only at baseline, year 3, and year 6. More‐frequent ascertainment of sitting time would have provided estimates of sitting times more proximal to time of MI; we were unable to examine changes in sitting time that occurred more frequently than our measurement timing. Additionally, both PA and sitting time were self‐reported. Nonetheless, the PA measurement has been validated in this cohort,27 and measurement error is likely to be nondifferential because of the prospective nature of the study. Accelerometers could provide more‐accurate measurements of sitting time and PA.28 The results of this study are applicable only to postmenopausal women and may not be generalizable to younger women, populations with a different distribution of race/ethnicity, or men. Additionally, there are a small number of deaths in some of the categories. Finally, as with any observational study, residual confounding by other lifestyle factors should be taken into consideration.
With the improvement in medical care, there are increasing numbers of MI survivors,9 a group of individuals who may be highly motivated to make lifestyle changes. Thus, the present study has important public health implications, particularly given the high prevalence of prolonged sitting time and PA in the population. The results of the current study suggest that women who survive an initial MI should limit sedentary time and increase PA, such as walking, to decrease subsequent mortality. However, we encourage women with a history of MI to consult with their physician before increasing PA to ensure that it is safe for them to do so.
Abbreviated List of WHI Investigators: Program Office: (National Heart, Lung, and Blood Institute, Bethesda, Maryland) Jacques Rossouw, Shari Ludlam, Dale Burwen, Joan McGowan, Leslie Ford, and Nancy Geller.
Clinical Coordinating Center: Clinical Coordinating Center: (Fred Hutchinson Cancer Research Center, Seattle, WA) Garnet Anderson, Ross Prentice, Andrea LaCroix, and Charles Kooperberg.
Investigators and Academic Centers: (Brigham and Women's Hospital, Harvard Medical School, Boston, MA) JoAnn E. Manson; (MedStar Health Research Institute/Howard University, Washington, DC) Barbara V. Howard; (Stanford Prevention Research Center, Stanford, CA) Marcia L. Stefanick; (The Ohio State University, Columbus, OH) Rebecca Jackson; (University of Arizona, Tucson/Phoenix, AZ) Cynthia A. Thomson; (University at Buffalo, Buffalo, NY) Jean Wactawski‐Wende; (University of Florida, Gainesville/Jacksonville, FL) Marian Limacher; (University of Iowa, Iowa City/Davenport, IA) Robert Wallace; (University of Pittsburgh, Pittsburgh, PA) Lewis Kuller; (Wake Forest University School of Medicine, Winston‐Salem, NC) Sally Shumaker.
Sources of Funding
The WHI program is funded by the National Heart, Lung, and Blood Institute, National Institutes of Health, US Department of Health and Human Services through contracts, HHSN268201100046C, HHSN268201100001C, HHSN268201100002C, HHSN268201100003C, and HHSN268201100004C.
Table S1. Means and SDs of Physical Activity (PA) Pre‐ and Post‐Myocardial Infarction (MI) in MET‐h/week by Categories of Change in PA
Table S2. Hazard Ratios (HR) and 95% CI for All‐Cause, Coronary Heart Disease (CHD), and Cardiovascular Disease (CVD) Mortality in Relation to Conditional Relative Physical Activity
- ↵Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, Das SR, de Ferranti S, Després JP, Fullerton HJ, Howard VJ. Executive summary: Heart disease and stroke statistics—2016 update: A report from the American Heart Association. Circulation. 2016;133:447–454.
- ↵O'Connor GT, Buring JE, Yusuf S, Goldhaber SZ, Olmstead EM, Paffenbarger R, Hennekens CH. An overview of randomized trials of rehabilitation with exercise after myocardial infarction. Circulation. 1989;80:234–244.
- ↵Chomistek AK, Manson JE, Stefanick ML, Lu B, Sands‐Lincoln M, Going SB, Garcia L, Allison MA, Sims ST, LaMonte MJ. Relationship of sedentary behavior and physical activity to incident cardiovascular disease: results from the Women's Health Initiative. J Am Coll Cardiol. 2013;61:2346‐2354.
- ↵Steffen‐Batey L, Nichaman MZ, Goff DC, Frankowski RF, Hanis CL, Ramsey DJ, Labarthe DR. Change in level of physical activity and risk of all‐cause mortality or reinfarction: the Corpus Christi Heart Project. Circulation. 2000;102:2204–2209.
- ↵Haskell WL, Lee I‐M, Pate RR, Powell KE, Blair SN, Franklin BA, Macera CA, Heath GW, Thompson PD, Bauman A. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Circulation. 2007;116:1081.
- ↵Matthews CE, George SM, Moore SC, Bowles HR, Blair A, Park Y, Troiano RP, Hollenbeck A, Schatzkin A. Amount of time spent in sedentary behaviors and cause‐specific mortality in US adults. Am J Clin Nutr. 2012;95:437–445.
- ↵Quell KJ, Porcari JP, Franklin BA, Foster C, Andreuzzi RA, Anthony RM. Is brisk walking an adequate aerobic training stimulus for cardiac patients? Chest J. 2002;122:1852–1856.