Abstract
Background and purpose
Whether changes in leisure-time physical activity (LTPA) over time are associated with lower risk of stroke is not well established. We examined the association between changes in self-reported LTPA 10 years apart, with risk of incident stroke in the California Teachers Study (CTS). We hypothesized that the risk of stroke would be lowest among those who remained active.
Methods
61,256 CTS participants reported LTPA at two intensity levels (moderate and strenuous activity) at two time points (baseline 1995-96; 10-year follow-up 2005-2006). LTPA at each intensity level was categorized based on American Heart Association (AHA) recommendations (moderate >150 minutes/week, strenuous >75 minutes/week). Changes in LTPA were summarized as follows: (i)not meeting recommendations at both time points; (ii)meeting recommendations only at follow-up; (iii)meeting recommendations only at baseline; (iv)meeting recommendations at both time points. Incident strokes were identified through California state hospitalization records. Using multi-variable Cox models, we examined the associations between changes in LTPA with incident stroke.
Results
987 women were diagnosed with stroke who completed both questionnaires. Meeting AHA recommendations at both time-points was associated with a lower risk of all stroke (adjusted HR 0.84, 95%CI0.72-0.98). The protective effects for stroke were driven by meeting AHA recommendations for moderate activity and largely observed for ischemic strokes (adjusted HR 0.70, 95%CI 0.55-0.88).
Conclusion
Meeting AHA recommendations for moderate activity had a protective effect for reducing ischemic stroke risk. Participants who met AHA recommendations at baseline but not at follow-up, however, were not afforded reduced stroke risk.
Introduction
Engaging in leisure-time physical activity (LTPA) has been consistently associated with a lower risk of stroke and cardiovascular disease1. As part of the American Heart Association's Life's Simple 7, engaging in at least 150 minutes of moderate intensity or 75 minutes per week of vigorous activity is recommended2. Several investigators who have examined the protective effect of LTPA on stroke have focused on single measurements of self-reported LTPA usually during mid-adulthood with follow-up times ranging from 1 to 30 years3-9. There are several limitations of those prior analyses however, including single measures of LTPA and incomplete ascertainment of stroke cases. A measurement of an exposure over several time epochs may provide more complete information about how protective or deleterious that exposure is10. For example, the cardiovascular effect of tobacco use is influenced in studies by whether participants are current, former, or never smokers, and their pack-year history exposure11. Similarly, the patterns of LTPA that individuals engage in are not static but may change over time – some may engage in more activity as they age while most tend to decline in the intensity and time spent on LTPA12. The degree to which changes in LTPA over the lifetime, such as sustained activity over a lifetime versus increasing activity in adulthood, influence the risk of stroke has not been well explored.
Here, we examined the association of self-reported LTPA measured at two time intervals, 10 years apart, with risk of stroke in the California Teachers Study cohort, a large prospective cohort of female teachers and administrators in the state of California. We hypothesized that the risk of stroke would be highest among those who did not exercise in either time interval, and that the effect would thus be smaller in magnitude but still present in those who transitioned from active to inactive over two time points.
Methods
Study population
Details on recruitment of the California Teachers Study have been previously published13. In brief, the California Teachers Study is a prospective study established in 1995, comprising 133,479 current and retired female teachers and administrators who were active members of the California State Teachers Retirement System. Women were excluded if they were not residents of California at baseline (n=8,866), had a prior history of stroke by self-report (n=1,587) or California Office of Statewide Health Planning and Development (OSHPD) diagnosis (97), were <26 years or >99 years old (n=1,145), or had missing or invalid physical activity data (n=970). Of the 120,814 eligible CTS participants identified at baseline, 61,256 participants had complete physical activity data at both baseline (1995-96) and follow-up (2005-06) and comprised our final study population (supplemental figure I).
Data collection
Baseline leisure-time physical activity was obtained using a questionnaire outlining time spent and intensity of exercise (categorized as moderate or strenuous) performed during a typical week. Examples were provided regarding activities which could be categorized as moderate (brisk walking, golf, among others) or strenuous (swimming, aerobics, running) intensity. The questionnaire was first mailed out at enrollment in 1995 and the LTPA section was then repeated in the 2005-2006 follow-up questionnaire. Participants reported their mean hours per week and months per year (1-3, 4-6, 7-9, and 10-12 months) of participation at each level of activity in the past 3 years. For each intensity level, we created mean annual hours using per week for each period by multiplying the hours per week by the portion of the year engaged in the activity, and calculated hours per week in which moderate and/or strenuous activity was engaged14. We dichotomized these measures by the American Heart Association recommendations for ideal cardiovascular health (either 150 minutes per week for moderate activity or 75 minutes per week for strenuous activity). These categorical variables were used as the LTPA definition in our multivariate regression analyses.
In addition, we collected information on relevant stroke risk factors in the baseline questionnaire, including information on race/ethnicity, smoking history, alcohol consumption, hormone replacement therapy use, and history of hypertension and diabetes. Body mass index (BMI) was calculated using self-reported measurements as weight in kilograms divided by height in meters squared. We used ICD-9 codes 272.xx in OSHPD through 1997 to capture hyperlipidemia at baseline. Socioeconomic status was estimated using baseline residential addresses that were geocoded to a 1990 census block group based on occupation, education and income.
Stroke was defined as the following: incident total stroke, ischemic stroke, hemorrhagic stroke and fatal stroke. Participants were identified through linkage with the OSHPD hospital discharge database (http://www.oshpd.ca.gov/), the California state mortality file, and the National Death Index between 1995 and 2012, using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). Stroke was defined as a hospitalization or death with an ICD-9-CM code of 430, 431, 432.x (excluding 432.1), 433.x1, 434.xx (excluding 434.x0), or 436 in the principal diagnosis position only. Ischemic stroke was defined as hospitalization or death with ICD-9-CM codes 433.x1, 434.xx (excluding 434.x0), or 436. Hemorrhagic stroke was defined as hospitalization or death with ICD-9-CM codes 430, 431, or 432. These definitions were previously validated in the CTS based on expert neurologist reviews. (Wang SS, K Lakshminarayan, Elkind MSV, et al. Validation of hospitalization discharge codes for stroke and transient ischemic attacks in the California Teachers Study Cohort (submitted)). The study was approved by the IRBs at the City of Hope, University of Southern California, University of California at Irvine, and the Cancer Prevention Institute of California.
Statistical analysis
Baseline characteristics of the cohort by change in LTPA were compared using chi-square for categorical and t-tests for continuous variables. For each stroke outcome (total, ischemic, hemorrhagic, and fatal separately), we examined the association of LTPA with incident stroke using Cox proportional hazards models excluding event occurring before the second LTPA assessment.. In order to examine how changes in LTPA are associated with incident stroke (total, ischemic, hemorrhagic, and fatal), we first examined the association of the single measure of LTPA obtained at the follow-up questionnaire (2005-2006) with subsequent stroke. Our primary analyses then examined changes in LTPA between the two questionnaires with risk of ischemic stroke. Specifically, we define four LTPA subgroups for each exercise patterns (moderate, strenuous, and moderate/strenuous), namely, not meeting recommendation at both time points (reference group), not meeting recommendation at baseline but meeting recommendation at follow-up (positive change), meeting recommendation at baseline but not at follow up (negative change), and meeting recommendation at both times (high PA). The time scale (in days) was defined from age at enrollment to age at either incident stroke, death, a move out of California (for more than 3 months), or Dec 31, 2012 – whichever came first. Cox proportional hazard models were stratified by age at baseline (in years) to adjust for calendar effects and adjusted for demographic risk factors (race-ethnicity, socioeconomic status), behavioral risk factors (tobacco use, alcohol use, BMI), and clinical risk factors (hypertension, diabetes, dyslipidemia). We conducted sensitivity analyses using multiple imputation to assess the impact of missing physical activity data at follow-up. Analyses were done in SAS using PROC MI and PROC MIANALYZE, using 20 imputed datasets. Imputations were created using covariates (age, smoking status, race, diabetes, hypertension), the outcome (stroke), and all available physical activity measurements.
All statistical analyses were conducted using SAS v9.3 (Cary, NC) and R.
Results
Baseline demographics
Table 1 outlines the baseline characteristics of all eligible participants by change in LTPA status. The mean age was 53 ±14 years, 88% were non-Hispanic White, and 79% had above median socio-economic status. One-third of participants reported meeting AHA recommended physical activity at both time periods, with the majority due to meeting moderate activity guidelines. Characteristics of this study population included a low proportion of cardiovascular disease risk factors (<3% diabetic, 18% hypertensive). However, we note a few differences between those with and without the second LTPA assessment; those with missing data for the second visit were slightly older at study enrollment and had a slightly higher proportion of hypertension and current alcohol use as reported in their baseline questionnaire (Supplemental Table I); changes in LTPA are outlined in supplemental table II.
Table 1.
Baseline characteristics of the California Teachers Study by change in physical activity (PA) status, defined by strenuous or moderate physical activity from 1995-96 to 2005-06. Below American Heart Association (AHA) recommended levels of physical activity for stroke prevention defined as<150 mins/week moderate activity and <75 mins/week strenuous activity; meeting or exceeding AHA recommended levels: ≥150 mins/week moderate or 75 mins/week strenuous activity)
| Participants with physical activity data at follow-up (2005-06) (n = 61,256) N (%) |
Below AHA-recommended PA levels at both time periods (n=18,781) N (%) |
Below AHA-recommended PA levels in 1995-96 to meeting in 2005-06. (n=11,498) N (%) |
Meeting AHA-recommended PA levels in 1995-96 and below in 2005-06. (n=8,609) N (%) |
Meeting AHA-recommended PA levels at both time periods. (n=22,368) N (%) |
|
|---|---|---|---|---|---|
| Age | 52 (26-94) | 51 (26-93) | 51 (26-87) | 52 (26-94) | 52 (26-89) |
| Race/ethnicity | |||||
| White | 54023 (88) | 16140 (86) | 10001 (87) | 7646 (89) | 20236 (90) |
| Black | 1264 (2) | 509 (3) | 251 (2) | 156 (2) | 348 (2) |
| Hispanic | 2280 (4) | 828 (4) | 481 (4) | 323 (4) | 648 (3) |
| Asian/Pacific Islander | 2250 (4) | 886 (5) | 492 (4) | 265 (3) | 607 (3) |
| Other/Missing | 1439 (2) | 418 (2) | 273 (2) | 219 (3) | 529 (2) |
| SES | |||||
| Below median | 11940 (19) | 4306 (23) | 2204 (19) | 1762 (21) | 3668 (17) |
| Above median | 48591 (79) | 14266 (77) | 9163 (81) | 6735 (79) | 18427 (83) |
| Hypertension | 11092 (18) | 4090 (22) | 1936 (17) | 1731 (20) | 3335 (15) |
| Diabetes | 1272 (2) | 571 (3) | 227 (2) | 230 (3) | 325 (1) |
| Hyperlipidemia | 567 (1) | 217 (1) | 82 (1) | 102 (1) | 166 (1) |
| Body mass index (kg/m2) | |||||
| <20 | 6366 (10) | 1626 (9) | 1182 (11) | 829 (10) | 2729 (13) |
| 20-<25 | 30434 (50) | 7750 (43) | 5578 (50) | 4148 (50) | 12958 (60) |
| 25-<30 | 14840 (24) | 5124 (28) | 2828 (25) | 2313 (28) | 4575 (21) |
| 30+ | 7778 (13) | 3687 (20) | 1561 (14) | 1063 (13) | 1467 (7) |
| Tobacco use | |||||
| Current | 2422 (4) | 889 (5) | 473 (4) | 341 (4) | 719 (3) |
| Former | 17667 (29) | 4845 (26) | 3163 (28) | 2479 (29) | 7180 (32) |
| Never | 40924 (67) | 12973 (69) | 7807 (68) | 5750 (67) | 14394 (65) |
| Alcohol use | |||||
| Current | 40152 (66) | 11109 (63) | 7454 (68) | 5660 (70) | 15929 (75) |
| Former | 7125 (12) | 2394 (14) | 1360 (13) | 1041 (13) | 2330 (11) |
| Never | 10723 (18) | 4188 (24) | 2058 (19) | 1442 (18) | 3035 (14) |
| Hormone replacement therapy | |||||
| Current | 23564 (39) | 6911 (37) | 4631 (40) | 3061 (36) | 8961 (40) |
| Former | 6646 (11) | 2125 (11) | 1121 (10) | 1111 (13) | 2289 (10) |
| Never | 30884 (50) | 9699 (52) | 5718 (50) | 4404 (51) | 11063 (50) |
Association between stroke and LTPA measured at 10-year follow-up
The association of LTPA assessed at the second questionnaire with risk of stroke is outlined in Table 2. Engaging in moderate intensity activity was associated with a lower risk of all stroke, a result driven by a reduced risk of ischemic strokes. Specifically, engaging in at least 150 minutes per week of moderate activity versus not was associated with a lower risk of stroke (Model 3: adjusted HR 0.84, 95% CI 0.73-0.96). Engaging in more than 75 minutes of strenuous activity was not statistically significantly associated with a lower risk of stroke (Model 3: adjusted HR 0.91, 95% CI 0.77-1.07). The reduced stroke risk was largely driven by ischemic stroke risk where meeting recommendations for AHA moderate activity resulted in nearly 20% decrease in risk (Model 3: HR 0.82, 95% CI 0.70-0.96). Results are consistent when using baseline LTPA data (Supplemental Table III).
Table 2.
Risk association (hazard ratio and 95% confidence interval) of strenuous and/or moderate physical activity, dichotomized according to AHA recommendations for moderate (>150 minutes/week) and strenuous (>75 minutes/week) activity, reported in 2005-06 with total, ischemic, and hemorrhagic stroke and stroke mortality.
| Physical activity | Total Stroke (n = 987) |
Ischemic stroke (n = 709) |
Hemorrhagic stroke (n = 221) |
Fatal stroke (n = 247) |
|---|---|---|---|---|
| Model 1: Moderate activity | ||||
| ≤150 mins/week | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| >150 mins/week | 0.80 (0.70-0.91) | 0.77 (0.66-0.90) | 0.92 (0.70-1.21) | 0.64 (0.48-0.84) |
| Model 1: Strenuous activity | ||||
| ≤75 mins/week | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| >75 mins/week | 0.88 (0.75-1.04) | 0.85 (0.70-1.04) | 0.95 (0.68-1.31) | 0.94 (0.67-1.31) |
| Model 1: Moderate/strenuous | ||||
| Neither recommendation met | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| Either recommendation met | 0.78 (0.69-0.89) | 0.77 (0.66-0.89) | 0.85 (0.65-1.11) | 0.71 (0.55-0.93) |
| Model 2: Moderate activity | ||||
| ≤150 mins/week | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| >150 mins/week | 0.82 (0.72-0.94) | 0.80 (0.68-0.94) | 0.93 (0.70-1.22) | 0.64 (0.48-0.85) |
| Model 2: Strenuous activity | ||||
| ≤75 mins/week | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| >75 mins/week | 0.90 (0.76-1.06) | 0.87 (0.72-1.07) | 0.96 (0.69-1.33) | 0.93 (0.66-1.31) |
| Model 2: Moderate/strenuous | ||||
| Neither recommendation met | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| Either recommendation met | 0.81 (0.71-0.93) | 0.80 (0.69-0.94) | 0.86 (0.65-1.13) | 0.71 (0.55-0.93) |
| Model 3: Moderate activity | ||||
| ≤150 mins/week | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| >150 mins/week | 0.84 (0.73-0.96) | 0.82 (0.70-0.96) | 0.93 (0.70-1.23) | 0.65 (0.49-0.87) |
| Model 3: Strenuous activity | ||||
| ≤75 mins/week | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| >75 mins/week | 0.91 (0.77-1.07) | 0.89 (0.73-1.08) | 0.96 (0.69-1.33) | 0.95 (0.68-1.33) |
| Model 3: Moderate/strenuous | ||||
| Neither recommendation met | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| Either recommendation met | 0.83 (0.73-0.94) | 0.82 (0.70-0.96) | 0.86 (0.65-1.13) | 0.73 (0.56-0.95) |
Model 1: Adjusted for age;
Model 2: Model 1 + Adjusted for race, socio-economic status, tobacco use, alcohol use, body-mass index
Model 3: Model 2 + adjusted for hypertension, diabetes, hyperlipidemia.
Association between stroke and changes in LTPA between baseline and 10-year follow-up
The primary outcome of interest in this analysis was incident strokes occurring after the follow-up questionnaire in 2005-06 (Table 3). In all, there were 987 strokes (mean age of onset 81), with 709 ischemic (mean age at onset 81) and 221 hemorrhagic strokes(mean age at onset 78). Of these, 247 were fatal strokes. The mean follow-up time after the second questionnaire was 6.5 years. In adjusted analyses we found that participants who did not meet AHA recommendations for LTPA at baseline but did meet AHA recommendations in follow-up had reduced risk of stroke (adjusted HR 0.79, 95% CI 0.65-0.97). Interestingly, this magnitude of risk was similar to that of participants who met AHA recommendations at both baseline and follow-up (adjusted HR 0.84, 95% CI 0.72-0.98). Notably, the reduced risks were significant only for moderate activity and ischemic stroke (adjusted HR 0.78, 95% CI 0.69-1.00). Participants who met AHA recommendations at baseline but did not meet recommendations in follow-up did not benefit from reduced stroke risk. Further, evaluation of LTPA in early adulthood (teenage years and early 20s) also did not affect stroke risk (data not shown), further supporting the notion that more recent moderate activity measured in adulthood was the relevant activity that offered benefits for reduced stroke risk. Due to the loss to follow we carried out imputation for missing data to examine whether the associations between change in LTPA was still associated with risk of stroke. Our results were similar indicating a low likelihood that loss to follow up biased our results (Supplemental table IV).
Table 3.
Risk estimates for total, ischemic, hemorrhagic, and fatal strokes by exercise patterns in moderate, strenuous, and strenuous/moderate physical activity among 61,256 women with physical activity measurements in 1995-96 and 2005-06.
| All Stroke HR (95% CI) |
Ischemic HR (95% CI) |
Hemorrhagic HR (95% CI) |
Fatal HR (95% CI) |
|
|---|---|---|---|---|
| Exercise pattern* - Moderate | ||||
| Recommendations not met at both time-points | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| Not meeting recommendations at baseline, but meeting recommendations in follow-up. | 0.73 (0.60-0.89) | 0.70 (0.55-0.88) | 0.76 (0.50-1.14) | 0.65 (0.43-0.99) |
| Meeting recommendations at baseline, but not meeting recommendations in follow-up | 0.96 (0.80-1.14) | 0.94 (0.77-1.16) | 1.00 (0.68-1.48) | 0.97 (0.69-1.35) |
| Meeting recommendations at both time-points | 0.88 (0.74-1.05) | 0.87 (0.71-1.07) | 1.06 (0.75-1.50) | 0.62 (0.43-0.90) |
| Exercise pattern* - Strenuous | ||||
| Recommendations not met at both time-points | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| Not meeting recommendations at baseline, but meeting recommendations in follow-up. | 0.91 (0.72-1.16) | 0.89 (0.67-1.19) | 0.94 (0.58-1.53) | 0.74 (0.42-1.30) |
| Meeting recommendations at baseline, but not meeting recommendations in follow-up | 1.12 (0.94-1.33) | 1.11 (0.91-1.36) | 1.10 (0.76-1.58) | 1.17 (0.84-1.64) |
| Meeting recommendations at both time-points | 0.95 (0.76-1.17) | 0.91 (0.70-1.18) | 1.00 (0.65-1.52) | 1.17 (0.78-1.76) |
| Exercise pattern*– Strenuous and moderate | ||||
| Recommendations not met at both time-points | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| Not meeting recommendations at baseline, but meeting recommendations in follow-up. | 0.79 (0.65-0.97) | 0.78 (0.69-1.00) | 0.75 (0.49-1.15) | 0.73 (0.47-1.13) |
| Meeting recommendations at baseline, but not meeting recommendations in follow-up | 1.00 (0.83-1.19) | 1.01 (0.82-1.24) | 0.93 (0.62-1.37) | 1.08 (0.77-1.51) |
| Meeting recommendations at both time-points | 0.84 (0.72-0.98) | 0.84 (0.70-1.01) | 0.88 (0.63-1.22) | 0.76 (0.55-1.05) |
Adjusted for age, race, socio-economic status, tobacco use, alcohol use, body-mass index, hypertension, diabetes, hyperlipidemia.
Discussion
In the California Teachers Study, a prospective cohort designed to study risk factors for cancer but with a validated stroke adjudication process, we found that meeting AHA guidelines of moderate physical activity was associated with a lower risk of stroke, specifically ischemic stroke. Conversely, we did not detect an association between meeting AHA guidelines for strenuous activity and risk of risk. Several investigators have demonstrated a consistent association of leisure-time physical activity with reduced risk of stroke, with the magnitude of the effect differing by mean age of the cohort, intensity or total calories15. For example, in the Northern Manhattan Study, the highest levels of total leisure-time physical activity, factoring in intensity and time, were associated with a lower risk of ischemic stroke9. In the Nurse's Health Study, moderate intensity activities, such as walking for exercise, were also associated with a lower risk of stroke16. Few analyses, however, have examined the association of changes in physical activity with risk of stroke. In considering more than one time-point of LTPA we found that more recent activity better reflects reduction in stroke risk. Participants who met AHA guidelines at baseline but not in follow-up did not benefit from the same reduction in stroke risk.
It is now recognized that risk factor status in individuals is not a static process, but rather a dynamic process rarely accounted for in analyses of LTPA at a single time-point with stroke risk. Our finding lend further evidence to the recommendation that individuals should increase exercise as an effective lifestyle modification for reducing the risk of stroke1,2. These findings also support the concept that low exercise levels performed at younger ages are not sufficient to counteract the risk of cardiovascular disease in the future, but will need to be modified later in adulthood to gain a survival and cardiovascular mortality benefit17.
The lack of an association of risk of stroke with change in strenuous activity was unexpected. One possibility is that there is no clear dose response relationship with exercise, and similar to other CVD outcomes moderate intensity activity is sufficient to reduce risk1. The effect of moderate intensity activity on cardiometabolic profiles, such as dyslipidemia, diabetes, and obesity, has been well documented12,18-20. Similar to other large cohort studies, an analysis of the National Walkers' Health Study21 showed a lower risk in multiple cardiovascular disease outcomes with even modest levels of activity (1.8 – 3.6 MET-hours/day – equivalent to one hour of light walking per day). In those analyses there was a similar magnitude of an effect in comparison to the higher level of activities (MET-hours/day > 3.6) for reducing risk of stroke, congestive heart failure, and cardiovascular disease mortality. A notable exception was the substantial benefit in higher intensity activities on diabetes related mortality (63.8% reduction for moderate, 90.8% reduction for the highest level). 22
The mechanism by which exercise reduces the risk of stroke is likely multi-factorial. The effect of exercise on improving dyslipidemia, hypertension, weight, and diabetes have been well established; this is likely to be an only partial explanation as in multiple studies adjustment for these risk factors in multi-variable models does not alter the protective effect of LTPA. The independent benefit of exercise is likely from improving vascular and endothelial health, reducing inflammation, improving sympathetic tone, and maintenance of cerebral auto-regulation23.
Our study has several strengths, including a large sample size, multiple measure of LTPA, a validated stroke outcome ascertainment process, long duration of follow-up, and information available on multiple confounders. Several weaknesses, however, require discussion. Though the sample size was large, the overall proportion of participants with stroke, particularly hemorrhagic, is small. Our results may therefore not be generalizable to the general population with a higher baseline risk of stroke. Similarly only 30% of participants met guidelines for strenuous activity and we have been underpowered to detect a difference in that group. On the other hand, individuals in their 4th to 6th decade of life (middle age) have had a continued increase in the incidence of stroke in yearly trends and our finding provide at least one intervention that could ameliorate that. We unfortunately do not have data on changes in other risk factors, notably BMI, which may have acted as a confounder in our analyses. Similarly we did not have further details on important confounders at baseline, such as blood pressure and diabetes control markers. We had a significant proportion of participants who were lost to follow up between the first and second assessment which could be a source of selection bias. The baseline characteristics of those lost to follow up showed small increase in baseline risk factors for stroke such that those without second assessment may have been more likely to have an outcome event. On the other hand, we would expect that these results would bias our results towards the null, and yet still noted an effect with changes in LTPA. Furthermore, we carried out imputation analysis among those lost to follow-up and our results were fundamentally similar. We did not objectively collect LTPA with the use of accelerometers, with several studies reporting significant reporting biases with self-reported questionnaires24,25. Participants may also tend to over-report the amount of activity they perform which may bias our results to the null. Though commercial activity monitors have been increasing in use, most individuals are counseled in clinical practice based on their reported activity. Finer details of LTPA, such as differing intensities and frequencies may provide a more nuanced understanding on how these components in combination may reduce the risk of stroke and cardiovascular disease26,27.
Summary/conclusions
In summary, we find that starting and sustaining moderate intensity LTPA in mid to late adulthood was associated with a lower risk of stroke. Our findings emphasize the importance of behavioral interventions centered on life-style, such as prescribing exercise in those who are inactive, for preventing stroke. Future studies are required to understand components and patterns of LTPA that are most likely to reduce the risk of stroke, likely informed by objective measures of LTPA.
Supplementary Material
Footnotes
Disclosures: none
References
- 1.Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, et al. Heart disease and stroke statistics--2015 update: a report from the American Heart Association. Circulation. 2015;131:e29–322. doi: 10.1161/CIR.0000000000000152. [DOI] [PubMed] [Google Scholar]
- 2.Sacco RL. The new American Heart Association 2020 goal: achieving ideal cardiovascular health. J Cardiovasc Med (Hagerstown) 2011;12:255–257. doi: 10.2459/JCM.0b013e328343e986. [DOI] [PubMed] [Google Scholar]
- 3.Evenson KR, Rosamond WD, Cai J, Toole JF, Hutchinson RG, Shahar E, et al. Physical activity and ischemic stroke risk. The atherosclerosis risk in communities study. Stroke; a journal of cerebral circulation. 1999;30:1333–1339. doi: 10.1161/01.str.30.7.1333. [DOI] [PubMed] [Google Scholar]
- 4.Gillum RF, Mussolino ME, Ingram DD. Physical activity and stroke incidence in women and men. The NHANES I Epidemiologic Follow-up Study. American journal of epidemiology. 1996;143:860–869. doi: 10.1093/oxfordjournals.aje.a008829. [DOI] [PubMed] [Google Scholar]
- 5.Lee IM, Paffenbarger RS., Jr Physical activity and stroke incidence: the Harvard Alumni Health Study. Stroke; a journal of cerebral circulation. 1998;29:2049–2054. doi: 10.1161/01.str.29.10.2049. [DOI] [PubMed] [Google Scholar]
- 6.McDonnell MN, Hillier SL, Hooker SP, Le A, Judd SE, Howard VJ. Physical activity frequency and risk of incident stroke in a national US study of blacks and whites. Stroke; a journal of cerebral circulation. 2013;44:2519–2524. doi: 10.1161/STROKEAHA.113.001538. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Paffenbarger RS, Jr, Hyde RT, Wing AL, Steinmetz CH. A natural history of athleticism and cardiovascular health. JAMA. 1984;252:491–495. [PubMed] [Google Scholar]
- 8.Thrift AG, Donnan GA, McNeil JJ. Reduced risk of intracerebral hemorrhage with dynamic recreational exercise but not with heavy work activity. Stroke; a journal of cerebral circulation. 2002;33:559–564. doi: 10.1161/hs0202.102878. [DOI] [PubMed] [Google Scholar]
- 9.Willey JZ, Moon YP, Paik MC, Boden-Albala B, Sacco RL, Elkind MS. Physical activity and risk of ischemic stroke in the Northern Manhattan Study. Neurology. 2009;73:1774–1779. doi: 10.1212/WNL.0b013e3181c34b58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cupples LA, D'Agostino RB, Anderson K, Kannel WB. Comparison of baseline and repeated measure covariate techniques in the Framingham Heart Study. Stat Med J. 1988;7:205–222. doi: 10.1002/sim.4780070122. [DOI] [PubMed] [Google Scholar]
- 11.O'Donnell MJ, Xavier D, Liu L, Zhang H, Chin SL, Rao-Melacini P, et al. Risk factors for ischaemic and intracerebral haemorrhagic stroke in 22 countries (the INTERSTROKE study): a case-control study. Lancet. 2010;376:112–123. doi: 10.1016/S0140-6736(10)60834-3. [DOI] [PubMed] [Google Scholar]
- 12.Bann D, Kuh D, Wills AK, Adams J, Brage S, Cooper R. Physical activity across adulthood in relation to fat and lean body mass in early old age: findings from the medical research council national survey of health and development, 1946-2010. American journal of epidemiology. 2014;179:1197–1207. doi: 10.1093/aje/kwu033. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bernstein L, Allen M, Anton-Culver H, Deapen D, Horn-Ross PL, Peel D, et al. High breast cancer incidence rates among California teachers: results from the California Teachers Study (United States) Cancer Causes Control. 2002;13:625–635. doi: 10.1023/a:1019552126105. [DOI] [PubMed] [Google Scholar]
- 14.Dallal CM, Sullivan-Halley J, Ross RK, Wang Y, Deapen D, Horn-Ross PL, et al. Long-term recreational physical activity and risk of invasive and in situ breast cancer: the California teachers study. Archives of internal medicine. 2007;167:408–415. doi: 10.1001/archinte.167.4.408. [DOI] [PubMed] [Google Scholar]
- 15.Lee CD, Folsom AR, Blair SN. Physical activity and stroke risk: a meta-analysis. Stroke; a journal of cerebral circulation. 2003;34:2475–2481. doi: 10.1161/01.STR.0000091843.02517.9D. [DOI] [PubMed] [Google Scholar]
- 16.Hu FB, Stampfer MJ, Colditz GA, Ascherio A, Rexrode KM, Willett CE, et al. Physical activity and risk of stroke in women. JAMA. 2000;283:2961–2967. doi: 10.1001/jama.283.22.2961. [DOI] [PubMed] [Google Scholar]
- 17.Shah RV, Murthy VL, Colangelo LA, Reis J, Venkatesh BA, Sharma R, et al. Association of Fitness in Young Adulthood With Survival and Cardiovascular Risk: The Coronary Artery Risk Development in Young Adults (CARDIA) Study. JAMA Intern Med. 2016;176:87–95. doi: 10.1001/jamainternmed.2015.6309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bateman LA, Slentz CA, Willis LH, Shields At, Piner LW, Bales CW, et al. Comparison of aerobic versus resistance exercise training effects on metabolic syndrome (from the Studies of a Targeted Risk Reduction Intervention Through Defined Exercise - STRRIDE-AT/RT) Am J Cardiol. 2011;108:838–844. doi: 10.1016/j.amjcard.2011.04.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kujala UM, Makinen VP, Heinonen I, Soininen P, Kangas AJ, Leskinen TH, et al. Long-term leisure-time physical activity and serum metabolome. Circulation. 2013;127:340–348. doi: 10.1161/CIRCULATIONAHA.112.105551. [DOI] [PubMed] [Google Scholar]
- 20.Carnethon MR, Gidding SS, Nehgme R, Sidney S, Jacobs DR, Jr, Liu K. Cardiorespiratory fitness in young adulthood and the development of cardiovascular disease risk factors. JAMA. 2003;290:3092–3100. doi: 10.1001/jama.290.23.3092. [DOI] [PubMed] [Google Scholar]
- 21.Williams PT. Dose-response relationship of physical activity to premature and total all-cause and cardiovascular disease mortality in walkers. PLoS One. 2013;8:e78777. doi: 10.1371/journal.pone.0078777. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Simon HB. Exercise and Health: Dose and Response, Considering Both Ends of the Curve. Am J Med. 2015;128:1171–1177. doi: 10.1016/j.amjmed.2015.05.012. [DOI] [PubMed] [Google Scholar]
- 23.Alevizos A, Lentzas J, Kokkoris S, Mariolis A, Korantzopoulos P. Physical activity and stroke risk. Int J Clin Pract. 2005;59:922–930. doi: 10.1111/j.1742-1241.2005.00536.x. [DOI] [PubMed] [Google Scholar]
- 24.Espana-Romero V, Golubic R, Martin KR, Harty R, Ekelund U, Kuh D, et al. Comparison of the EPIC Physical Activity Questionnaire with combined heart rate and movement sensing in a nationally representative sample of older British adults. PLoS One. 2014;9:e87085. doi: 10.1371/journal.pone.0087085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Innerd P, Catt M, Collerton J, Davies K, Trenell M, Kirkwood TB, et al. A comparison of subjective and objective measures of physical activity from the Newcastle 85+ study. Age Ageing. 2015;44:691–694. doi: 10.1093/ageing/afv062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mooney SJ, Joshi S, Cerda M, Quinn JW, Beard JR, Kennedy GJ, et al. Patterns of Physical Activity Among Older Adults in New York City: A Latent Class Approach. Am J Prev Med. 2015;49:e13–22. doi: 10.1016/j.amepre.2015.02.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Cheung YK, Yu G, Wall MM, Sacco RL, Elkind MS, Willey JZ. Patterns of leisure-time physical activity using multivariate finite mixture modeling and cardiovascular risk factors in the Northern Manhattan Study. Annals of epidemiology. 2015;25:469–474. doi: 10.1016/j.annepidem.2015.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.