Skip main navigation

Protective Effect of Time Spent Walking on Risk of Stroke in Older Men

Originally publishedhttps://doi.org/10.1161/STROKEAHA.113.002246Stroke. 2014;45:194–199

Abstract

Background and Purpose—

Older adults have the highest risks of stroke and the lowest physical activity levels. It is important to quantify how walking (the predominant form of physical activity in older age) is associated with stroke.

Methods—

A total of 4252 men from a UK population-based cohort reported usual physical activity (regular walking, cycling, recreational activity, and sport) in 1998 to 2000. Nurses took fasting blood samples and made anthropometric measurements.

Results—

Among 3435 ambulatory men free from cardiovascular disease and heart failure in 1998 to 2000, 195 first strokes occurred during 11-year follow-up. Men walked a median of 7 (interquartile range, 3–12) hours/wk; walking more hours was associated with lower heart rate, D-dimer, and higher forced expiratory volume in 1 second. Compared with men walking 0 to 3 hours/wk, men walking 4 to 7, 8 to 14, 15 to 21, and >22 hours had age- and region-adjusted hazard ratios (95% confidence intervals) for stroke of 0.89 (0.60–1.31), 0.63 (0.40–1.00), 0.68 (0.35–1.32), and 0.36 (0.14–0.91), respectively, P (trend)=0.006. Hazard ratios were somewhat attenuated by adjustment for established and novel risk markers (inflammatory and hemostatic markers and cardiac function [N-terminal pro-brain natriuretic peptide]) and walking pace, but linear trends remained. There was little evidence for a dose–response relationship between walking pace and stroke; comparing average pace or faster to a baseline of slow pace, the hazard ratio for stroke was 0.65 (95% confidence interval, 0.44–0.97), which was fully mediated by time spent walking.

Conclusions—

Time spent walking was associated with reduced risk of onset of stroke in dose–response fashion, independent of walking pace. Walking could form an important part of stroke-prevention strategies in older people.

Introduction

Stroke is a major cause of disability and mortality at older ages, so preventive strategies are important.1 Physical activity (PA) in middle age is protective against stroke; a meta-analysis reported that high PA levels compared with low PA levels were associated with overall 19% lower risk of stroke, although previous findings are mixed, with some inverse, u-shape, and even positive associations.2 Some studies suggest that PA may be more protective against stroke at older ages than in middle age.3,4 Walking is a predominant form of PA in older adults.5 It is therefore important to understand whether and how walking is related to onset of stroke in older adult populations, which have high risks of stroke and low activity levels.

Although slower walking speeds and spending less time walking are associated with elevated total cardiovascular disease mortality risk,6 few prospective studies of stroke examine the relative importance of pace compared with time spent walking or distance walked.7 Some suggest that both faster walking pace and greater time spent walking3,7,8 or MET hours9 of walking are protective against stroke and others have examined only walking speed.10,11 It is not clear what factors may mediate associations between walking and onset of stroke. To our knowledge, existing prospective studies of walking and stroke risk have not systematically addressed the mediating roles of novel cardiovascular markers,2 including C-reactive protein (CRP) a marker of inflammation, D-dimer a marker of coagulation and fibrinolysis, and N-terminal pro-brain natriuretic peptide (NT-proBNP) a marker of cardiac dysfunction, each of which is strongly related to onset of stroke1214 and also to PA level.15,16

We therefore test the hypotheses that (1) walking (pace, time spent, and distance) and (2) total leisure time PA are protective against onset of stroke. We investigate the role of CRP, D-dimer, and NT-proBNP alongside traditional risk factors as intermediate pathways.

Methods

Study Population

The British Regional Heart Study is a prospective cohort of 7735 men recruited from a single primary care center in each of 24 British towns in 1978 to 1980 (age, 40–59 years). Men were followed up for stroke morbidity and all-cause mortality. In 1998 to 2000, 4252 participants (age, 60–79 years) attended for follow-up measurements (77% response rate).17 A total of 811 with pre-existing myocardial infarction, stroke, or heart failure and 6 confined to a wheelchair were excluded to reduce potential for reverse causality, leaving 3435 men.

Clinical Data, Ethical Approval

Men completed questionnaires and nurses measured height, weight, blood pressure (BP), and forced expiratory volume in 1 second (FEV1)17 and recorded an ECG12 (including resting heart rate), and Minnesota coding criteria were used to define atrial fibrillation (8.3.1 and 8.3.3) and definite and possible left ventricular hypertrophy (3.1 and 3.3). Fasting venous blood samples were collected and analyzed for total and high-density lipoprotein-cholesterol, triglycerides,18 and vitamin C.19 CRP was assayed by ultrasensitive nephelometry (Dade Behring). Plasma levels of D-dimer were measured with ELISA (Biopool AB), as was von Willebrand factor antigen (DAKO). NT-proBNP was measured using the Elecsys 2010 electrochemiluminescence method (Roche Diagnostics, Burgess Hill, United Kingdom).13 All relevant local research ethics committees provided ethical approval and all men provided informed written consent to the investigation.

Assessment of PA

In 1998 to 2000, men self-reported usual pattern of PA under the headings of regular daily walking or cycling, recreational activity, and sporting (vigorous) activity. Men reported usual walking during an average week (1) duration: number of hours spent on all forms of walking, (2) distance: number of miles walked in total; (3) pace: slow, steady average, fairly brisk, and fast (≥4 mph). Men reporting recreational and sporting activity were classified according to whether they participated in vigorous activities at least once a month. Recreational activity included, for example, gardening, pleasure walking (hiking), and do-it-yourself jobs. Sporting activity included, for example, running, golf, swimming, and tennis. A PA score (validated in relation to heart rate and FEV1)16,20 was derived for each man on the basis of frequency and type (intensity) of each activity, scored based on intensity and energy demands using Minnesota intensity codes. The score included duration of walking but not distance walked or walking pace. Questionnaire data were available for 3357/3435 (98%) about walking pace, 2995/3435 (87%) for hours walked per week, and 3231 (94%) for distance walked per week; men with missing data for pace, distance, or time spent walking were dropped from relevant analyses.

Case Ascertainment and Follow-up

The outcome was first fatal or nonfatal stroke occurring after the 1998 to 2000 survey up to June 2010. Fatal cases were ascertained through the National Health Services Central Registers (death certificates with International Classification of Diseases-Ninth Revision codes 430–438 for stroke, indicating deaths with cerebrovascular disease as the underlying cause). Nonfatal stroke events were those that produced a neurological deficit that was present for >24 hours; data about nonfatal events was obtained from 2-yearly reviews of patient primary care notes (including all hospital and clinic correspondence). Supplementary information from computed tomography/MRI scans to confirm diagnoses was available in a subset of men.

Statistical Methods

Mean, medians, or proportions of behavioral and demographic factors selected a priori were calculated according to usual walking (1) pace, (2) duration, and (3) distance reported at Q20. Linear regression analyses were used to test trends across the walking categories. Skewed variables were natural log-transformed. Variables exhibiting significant diurnal variation were adjusted for time of measurement. BP and body mass index was adjusted for intraobserver variation.

Cox proportional hazards regression models were used to estimate associations between Q20 walking measures and risk of stroke. Survival times were censored at date of stroke, death from any cause, or end of follow-up period, whichever occurred first. Date of entry into the study in 1998 to 2000 was used as the time origin. The proportional hazards assumption was examined using time varying covariates, calculating interactions of predictor variables and a function of survival time and including them in the models. Examination of time varying covariates indicated that proportionality assumptions were not violated. The hazard ratios (HRs) for categories of walking in 1998 to 2000 were estimated and the overall association was tested with the continuous association among walking (1) time, (2) distance, (3) pace, and stroke risk, adjusted for sex, age (continuous variable), and region of residence. Time and duration of walking were adjusted for pace, and pace adjusted for time. Models were also adjusted for participation in vigorous recreational or sporting activities. Models were next adjusted for covariates associated with both stroke and walking; established risk factors (alcohol and tobacco use, social class [based on occupational group manual {skilled manual, unskilled, or partly skilled}, or nonmanual {professional, managerial, technical, skilled nonmanual occupations}]), and then biological risk factors (as continuous variables): first body mass index and then systolic BP, total and high-density lipoprotein-cholesterol and triglycerides, next FEV1 was added, finally, novel risk markers, CRP, D-dimer, and NT-proBNP were added. Interactions were tested using likelihood ratio tests. Analyses were repeated excluding the first 2 years of follow-up.

Results

Analyses are based on 3357 ambulatory men (mean age, 68.3 years) who were free from CHD, stroke, and heart failure at entry. Men walked a median of 7 hours per week (interquartile range, 3–12 hours). The correlation between categories of hours walked and walking pace was r=0.08 (P<0.001) and between categories of hours walked and walking distance was r=0.22 (P<0.001). Men who walked for more hours per week were younger, more likely to be of manual social class, reported higher levels of usual PA, had faster walking pace, and reported walking more miles per week than those who walked for fewer hours. The men who walked for more hours had lower heart rate, D-dimer, and higher FEV1 (Table 1).

Table 1. Characteristics of Men According to Time Spent Walking per Week (N=2995 Men)

0–3 h/wk (n=803)4–7 h/wk (n=921)8–14 h/wk (n=785)15–21 h/wk (n=270)≥22 h/wk (n=216)P (Trend)*
Age, y68.2168.5668.5367.9966.630.005
Social class, % (n)<0.001
 Manual45.6 (365)45.3 (416)48.7 (381)54.3 (146)64.4 (139)
Alcohol, % (n)0.081
 >15 U/wk13.6 (107)17.0 (153)19.8 (150)16.2 (43)16.5 (35)
Tobacco, % (n)0.411
 Current smoker13.0 (104)11.6 (107)13.4 (105)13.4 (36)17.1 (37)
Physical activity score<0.001
 None22.6 (177)4.2 (38)1.9 (15)1.5 (4)2.3 (5)
 Occasional33.2 (260)26.6 (242)12.8 (99)7.5 (20)10.7 (23)
 Light14.6 (114)24.2 (220)20.3 (157)19.5 (52)21.9 (47)
 Moderate8.2 (64)12.8 (116)24.1 (187)21.7 (58)27.4 (59)
 Moderately vigorous11.5 (90)18.9 (172)20.6 (160)22.5 (60)18.1 (39)
 Vigorous9.8 (77)13.3 (121)20.3 (157)27.3 (73)19.5 (42)
Walking pace<0.001
 Slow20.8 (166)8.7 (80)7.4 (58)6.3 (17)8.9 (19)
 Steady average57.1 (456)65.1 (596)68.5 (535)66.9 (180)69.6 (149)
 Brisk/fast22.1 (176)26.2 (240)24.1 (188)26.8 (72)21.5 (46)
Miles walked/wk§3.96.99.710.610.1<0.001
Prevalent diabetes mellitus11.8 (95)11.3 (104)9 (71)7 (19)8.8 (19)0.091
BMI, kg/m227.0126.6126.6926.6026.870.274
Waist circumference, cm97.6296.2596.3995.9496.780.051
Plasma vitamin C, µmol/L§22.7123.3123.1521.7619.260.052
Total cholesterol, mmol/L6.036.106.076.125.950.892
HDL-Cholesterol, mmol/L1.341.331.351.361.320.697
Triglycerides, mmol/L§1.601.621.601.611.500.250
Systolic blood pressure, mm Hg150.07150.51149.88147.82149.380.297
Diastolic blood pressure, mm Hg85.4986.1685.6985.4085.090.558
Blood pressure–lowering medication, % (n)28.2 (224)28.6 (260)26.2 (203)26.3 (70)24.9 (52)0.674
FEV1, L2.212.302.292.312.300.016
Heart rate, bpm66.1466.3165.4264.7064.100.007
Atrial fibrillation, % (n)2.5 (20)2.6 (24)2.9 (23)2.6 (7)1.9 (4)0.931
Left ventricular hypertrophy, % (n)7.7 (62)6.6 (61)6.6 (52)8.6 (23)7.4 (16)0.754
C-Reactive protein, mg/L§1.781.511.641.441.740.248
Von Willebrand factor, IU/dL138.07136.98137.56134.58132.270.129
D-dimer, ng/mL§81.7576.8579.9274.1769.390.016
NT-proBNP, pg/mL§85.6682.5291.0079.7877.620.538

BMI indicates body mass index; FEV1, forced expiratory volume in 1 second; HDL, high-density lipoprotein; and NT-proBNP, N-terminal pro-brain natriuretic peptide.

*P (trend) tested with linear regression for continuous variables and χ2 for categorical variables.

Adjusted for interobserver variation.

Adjusted for time of day.

§Geometric mean.

Adjusted for interobserver variation and height squared.

Men reported usual walking pace as slow (13%; n=425), steady average (64%; n=2143), and fairly brisk or fast (24%; n=789). Men with faster walking pace compared with slower walking pace were more likely to be younger, nonmanual social class, never smokers, light drinkers, had higher usual PA levels, walked for more hours per week and longer distances, and had lower prevalence of diabetes (Table I in the online-only Data Supplement). Faster walking pace showed graded associations with lower body mass index, waist circumference, plasma vitamin C, high-density lipoprotein-cholesterol, triglycerides, systolic BP, use of BP-lowering medications, heart rate, and inflammatory markers CRP, von Willebrand factor, D-dimer, and NTpro-BNP. Men with a faster walking pace had a higher mean FEV1.

During a median 10.9-year follow-up, 195 first stroke events occurred among 2995 men, 6.7 (95% confidence interval [CI], 5.8, 7.7) per 1000 person-years. The HRs for stroke reduced with increasing usual PA level, but CIs were wide (Table 2) and linear trends not significant. Focusing next on walking, we did not find evidence for an association between usual distance walked and stroke (Table 3) although the HR for men who walked the furthest distance (≥15 miles/wk) was smallest, but CIs were wide. However, a strong inverse dose–response association was observed between time spent walking and risk of stroke (Table 4). Compared with men spending 0 to 3 hours/wk walking, men who walked 4 to 7, 8 to 14, 15 to 21, and ≥22 hours had age- and region-adjusted HRs of 0.89 (0.60–1.31), 0.63 (0.40–1.00), 0.68 (0.35–1.32), 0.36 (0.14–0.91), respectively, P (linear trend)=0.006. Adjustments for multiple established risk factors, including vigorous physical activity level (model 2), walking pace (model 3), and FEV1 (model 4), attenuated associations to a very small degree. CRP, D-dimer, and NT-proBNP each attenuated coefficients to a similar degree when adjusted separately (not presented) or together (model 5), and the linear trend remained significant. We did not observe evidence for an interaction between time spent walking and pace (likelihood ratio test, P=0.1). In sensitivity analyses excluding the first 2 years of follow-up time, the point estimates were similar with slightly wider CIs, but significant trends remained.

Table 2. Association Between Total Physical Activity and Onset of Stroke, up to June 2010

Total Leisure Time Physical ActivityNoneOccasionalLightModerateModerately Vigorous and VigorousTotalP (Trend)
Participants (n)2696905694759922995
Person-years2279645155704768993829 001
Stroke events rates/1000 (n)7.9 (18)8.1 (52)7.5 (42)5.7 (27)5.6 (56)6.7 (195)
HR (95% CI)*
 Model 11.01.08 (0.63–1.85)0.99 (0.57–1.71)0.81 (0.44–1.48)0.79 (0.46–1.34)0.092
 Model 21.01.05 (0.61–1.82)0.93 (0.53–1.63)0.80 (0.43–1.47)0.77 (0.45–1.34)0.100

Model 1=age+region; model 2=model 1+alcohol intake (none/occasional, 1–15 U/wk, >15 U/wk)+vigorous recreational or sporting activity+smoking history (never-smoker, ex-smoker, and current smoker)+social class (nonmanual, manual, and armed forces)+total cholesterol+HDL-C+loge (triglycerides)+SBP+taking blood pressure–lowering medication+BMI+atrial fibrillation+left ventricular hypertrophy. BMI indicates body mass index; CI, confidence interval; HDL-C, high-density lipoprotein-cholesterol; HR, hazard ratio; and SBP, systolic blood pressure.

*From Cox regression models of physical activity level and all stroke.

Table 3. Association Between Distance Walked per Week and Onset of Stroke, up to June 2010

Distance Walked/Wk0–3 Miles/Wk4–7 Miles/Wk8–14 Miles/Wk>15 Miles/WkTotalP (Trend)
Participants (n)6978057875262815
Person-years650878687714526427 354
Stroke events rates/1000 (n)7.4 (48)6.6 (52)7.8 (60)4.9 (26)6.8 (186)
HR (95% CI)*
 Model 11.00.92 (0.62–1.37)1.12 (0.76–1.64)0.73 (0.45–1.19)0.505
 Model 21.00.89 (0.60–1.33)1.08 (0.73–1.59)0.70 (0.43–1.15)0.403

Model 1=age+region; model 2=model 1+alcohol intake (none/occasional, 1–15 U/wk, >15 U/wk)+vigorous recreational or sporting activity+smoking history (never-smoker, ex-smoker, and current smoker)+social class (nonmanual, manual, and armed forces)+total cholesterol+HDL-C+loge (triglycerides)+SBP+taking blood pressure–lowering medication+BMI+atrial fibrillation+left ventricular hypertrophy. BMI indicates body mass index; CI, confidence interval; HDL-C, high-density lipoprotein-cholesterol; HR, hazard ratio; and SBP, systolic blood pressure.

*From Cox regression models of physical activity level and all stroke.

Table 4. Association Between Time Spent Walking per Week and Onset of Stroke, up to June 2010

Time Spent Walking/Wk0–3 h4–7 h8–14 h15–21 h≥22 hTotalP (Trend)
Participants (n)6337476122101832385
Person-years6024734160242095106023 196
Stroke events rate/1000 (n)8.0 (48)7.7 (56)5.5 (33)5.3 (11)2.7 (5)6.6 (153)
HR (95% CI)*
 Model 11.00.89 (0.60–1.31)0.63 (0.40–1.00)0.68 (0.35–1.32)0.36 (0.14–0.91)0.006
 Model 21.00.88 (0.60–1.31)0.65 (0.41–1.02)0.69 (0.35–1.34)0.34 (0.13–0.87)0.006
 Model 31.00.90 (0.61–1.33)0.66 (0.42–1.04)0.70 (0.36–1.36)0.35 (0.14–0.88)0.008
 Model 41.00.89 (0.60–1.33)0.66 (0.42–1.04)0.70 (0.36–1.36)0.35 (0.14–0.88)0.007
 Model 51.00.91 (0.61–1.34)0.66 (0.42–1.04)0.70 (0.36–1.36)0.35 (0.14–0.89)0.008

Model 1=age+region; model 2=model 1+alcohol intake (none/occasional, 1–15 U/wk, >15 U/wk)+vigorous recreational or sporting activity+smoking history (never-smoker, ex-smoker, and current smoker)+social class (nonmanual, manual, and armed forces)+total cholesterol+HDL-C+loge (triglycerides)+SBP+taking blood pressure–lowering medication+BMI+atrial fibrillation+left ventricular hypertrophy; model 3=model 2+walking pace; model 4=model 3+FEV1; and model 5=model 3+loge (C-reactive protein)+loge (D-dimer)+loge (NT-proBNP). BMI indicates body mass index; CI, confidence interval; FEV1, forced expiratory volume in 1 second; HDL-C, high-density lipoprotein-cholesterol; HR, hazard ratio; NT-proBNP, N-terminal pro-brain natriuretic peptide; and SBP, systolic blood pressure.

*From Cox regression models of physical activity level and all stroke.

The risk of stroke was reduced for average pace compared with slow pace HR 0.66 (0.43–0.99) with no further reduction for fairly brisk HR 0.64 (0.39–1.07), so the average and fairly brisk were combined and compared with slow pace. The HR for stroke was significantly reduced in average or brisk pace compared with slow pace, 0.62 (0.42–0.92) for model 2 (Table 5) but was fully mediated on adjustment for either walking distance (HR, 0.67 [0.44–1.02]) or duration (HR, 0.67 [0.43–1.04]).

Table 5. Association Between Usual Walking Pace and Onset of Stroke, up to June 2010

Walking PaceSlow≥Steady AverageTotalP (Trend)
Participants (n)31924642783
Person-years2626244127 067
Stroke events rate/1000 (n)11.8 (31)6.1 (150)6.7 (181)
HR (95% CI)*
 Model 11.000.62 (0.42–0.92)0.017
 Model 21.000.66 (0.44–1.00)0.038
 Model 31.000.67 (0.44–1.02)0.063
 Model 41.000.67 (0.43–1.04)0.077

Model 1=age+region; model 2=model 1+alcohol intake (none/occasional, 1–15 U/wk, >15 U/wk)+vigorous recreational or sporting activity+smoking history (never-smoker, ex-smoker, and current smoker)+social class (nonmanual, manual, and armed forces)+total cholesterol+HDL-C+loge (triglycerides)+SBP+taking blood pressure–lowering medication+BMI+atrial fibrillation+left ventricular hypertrophy; model 3=model 2+walking distance; and model 4=model 2+walking time. BMI indicates body mass index; CI, confidence interval; HDL-C, high-density lipoprotein-cholesterol; HR, hazard ratio; and SBP, systolic blood pressure.

*From Cox regression models of physical activity level and all stroke.

Supplementary information on computed tomography/MRI scans was available for 75/195 (39%) of cases: 64/75 (85%) were ischemic and 11/75 (15%) were hemorrhagic strokes. In sensitivity analyses of the 64 ischemic cases, the dose–response pattern of lower risk with more hours walked was observed (P [trend]=0.004), no associations were observed for the number of miles walked, and the estimates for walking pace were similar to the main analyses.

Discussion

Among community dwelling older men, we observed a weak nonsignificant inverse association between total leisure time PA and stroke, and a strong inverse dose–response association between time spent walking and risk of stroke, independent of walking pace, vigorous physical activity, established, and novel risk factors. Results suggest that the total volume of walking rather than the intensity is important for stroke prevention. We investigated a range of plausible mechanisms to explain associations between walking and stroke, including lipids, hypertension, markers of inflammation, and endothelial dysfunction, but none fully mediated the associations with time spent walking.

Comparison With Other Studies

Findings about total PA and risk of stroke are mixed; inverse, u-shape, and even positive associations are reported.2 We found weak evidence for an inverse association between total leisure time PA or vigorous activity with onset of stroke, fitting with other null findings for total leisure time activity in men.3,8 However, a meta-analysis concluded that high levels of leisure time PA were associated with a 19% reduction in risk of stroke2; our estimates for moderate and more intense activity are of similar magnitude although our CIs were wide.

There is little epidemiological evidence about walking pace and duration about stroke, most focuses on CHD and total cardiovascular disease.6 Our findings that time spent walking was associated with reduced risk of stroke fits with the few existing studies, which include middle aged and older adults.3,79 Hence, our finding that time spent walking, but not total PA, was associated with stroke fits with other data; this consistency strengthens our speculation that it is a true biological finding, although we acknowledge that it could be because of issues with measurement of other types of PA. We did not find consistent evidence that walking pace was related to stroke onset; time spent walking explained the raised risk in slow-paced walkers, nor did we find that distance walked protected against stroke, although distance may be harder to recall accurately than usual pace or time spent walking. A large volume of weekly walking could be a proxy for low levels of sedentary behavior. A recent study reported that sedentary behavior is associated with stroke independently of PA,21 although we cannot test this in our dataset. Our study is novel in exploring mechanisms; we found that established stroke risk factors (including BP and lipids) were weak mediators of the associations between time spent walking and stroke as were markers of inflammation (CRP), coagulation (D-dimer), and also cardiac injury (NT-proBNP). It is possible that associations between walking volume and ischemic stroke may be mediated through mechanisms related to the progression of atherosclerosis and clot rupture that also act in CHD, whereas effects on hemorrhagic stroke may act through BP-related mechanisms, but we could not test this hypothesis.

Strengths and Limitations

This study benefits from prospective data with high follow-up rates, multiple measures of walking habit (pace, time spent walking, and distance), and other types of self-reported habitual PA combined in a score; both PA score16,20 and time spent walking are related to heart rate and FEV1. It is possible that there is a small degree of overlap between the total PA score (which uses weekend pleasure walking plus other components) and the daily walking variable (designed to capture active transport), but given that walking is protective against stroke, and the total PA score is not, any overlap is unlikely to have biased the score. Future studies with objective PA measures may clarify the shape of the dose–response curve between activity volume and intensity (including sedentary behavior) and stroke risk better than self-report, which may be subject to recall bias, yet self-reports are required for identifying activity types. Although we cannot entirely exclude residual confounding as an explanation for our findings, we adjusted for a wide range of important behavioral and social confounders including other dimensions of walking and other types of physical activities. To reduce the risk of reverse causality, we excluded men with pre-existing physician diagnosed cardiovascular disease or confined to a wheelchair, as these men have increased mortality risks and are likely limit their PA because of their health. We also excluded the first 2 years of follow-up in sensitivity analyses. Unlike other studies of walking pace and duration, we could investigate the mediating role of novel and established biological risk factors, which may be on the causal pathway between walking and stroke risk. We were able to distinguish between subtypes of stroke in a subset of cases, and, as reported elsewhere,12 ischemic strokes were more common than hemorrhagic strokes (a small minority) in older adults. In a sensitivity analysis of men with computed tomography–confirmed ischemic stroke, findings were similar to those for all participants. Our study examines only men, and we cannot generalize results to older women; however, the sample is socioeconomically representative of older men in the United Kingdom and has exceptionally high follow-up rates. One meta-analysis found borderline evidence for sex differences,22 and another reported that more vigorous PA may be required for protection against stroke in women than in men,2 but this may be because population levels of PA are lower in women than in men.

In conclusion, time spent walking was associated with risk of stroke in a dose–response fashion independent of walking pace and moderate to vigorous PA, indicating that among older men, daily walking could beneficially reduce risk of stroke. Therefore, walking for more hours per week could form an important part of stroke-prevention strategies. Future studies using objective measures of habitual walking will shed more light on this question.

Acknowledgments

We acknowledge the British Regional Heart Study team for data collection.

Footnotes

The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.113.002246/-/DC1.

Correspondence to Barbara J. Jefferis, PhD, Department of Primary Care and Population Health, UCL, Rowland Hill St, London NW3 2PF, United Kingdom. E-mail

References

  • 1. Goldstein LB, Adams R, Alberts MJ, Appel LJ, Brass LM, Bushnell CD, et al;American Heart Association/American Stroke Association Stroke Council; Atherosclerotic Peripheral Vascular Disease Interdisciplinary Working Group; Cardiovascular Nursing Council; Clinical Cardiology Council; Nutrition, Physical Activity, and Metabolism Council; Quality of Care and Outcomes Research Interdisciplinary Working Group; American Academy of Neurology. Primary prevention of ischemic stroke: a guideline from the American Heart Association/American Stroke Association Stroke Council: cosponsored by the Atherosclerotic Peripheral Vascular Disease Interdisciplinary Working Group; Cardiovascular Nursing Council; Clinical Cardiology Council; Nutrition, Physical Activity, and Metabolism Council; and the Quality of Care and Outcomes Research Interdisciplinary Working Group: the American Academy of Neurology affirms the value of this guideline.Stroke. 2006; 37:1583–1633.LinkGoogle Scholar
  • 2. Diep L, Kwagyan J, Kurantsin-Mills J, Weir R, Jayam-Trouth A. Association of physical activity level and stroke outcomes in men and women: a meta-analysis.J Womens Health (Larchmt). 2010; 19:1815–1822.CrossrefMedlineGoogle Scholar
  • 3. Huerta JM, Chirlaque MD, Tormo MJ, Gavrila D, Arriola L, Moreno-Iribas C, et al. Physical activity and risk of cerebrovascular disease in the European Prospective Investigation into Cancer and Nutrition-Spain study.Stroke. 2013; 44:111–118.LinkGoogle Scholar
  • 4. Abbott RD, Rodriguez BL, Burchfiel CM, Curb JD. Physical activity in older middle-aged men and reduced risk of stroke: the Honolulu Heart Program.Am J Epidemiol. 1994; 139:881–893.CrossrefMedlineGoogle Scholar
  • 5. DiPietro L. Physical activity in aging: changes in patterns and their relationship to health and function.J Gerontol A Biol Sci Med Sci. 2001; 56 Spec No 2:13–22.CrossrefMedlineGoogle Scholar
  • 6. Hamer M, Chida Y. Walking and primary prevention: a meta-analysis of prospective cohort studies.Br J Sports Med. 2008; 42:238–243.CrossrefMedlineGoogle Scholar
  • 7. Sattelmair JR, Kurth T, Buring JE, Lee IM. Physical activity and risk of stroke in women.Stroke. 2010; 41:1243–1250.LinkGoogle Scholar
  • 8. Noda H, Iso H, Toyoshima H, Date C, Yamamoto A, Kikuchi S, et al; JACC Study Group. Walking and sports participation and mortality from coronary heart disease and stroke.J Am Coll Cardiol. 2005; 46:1761–1767.CrossrefMedlineGoogle Scholar
  • 9. Hu FB, Stampfer MJ, Colditz GA, Ascherio A, Rexrode KM, Willett WC, et al. Physical activity and risk of stroke in women.JAMA. 2000; 283:2961–2967.CrossrefMedlineGoogle Scholar
  • 10. Longstreth WT, Bernick C, Fitzpatrick A, Cushman M, Knepper L, Lima J, et al. Frequency and predictors of stroke death in 5,888 participants in the Cardiovascular Health Study.Neurology. 2001; 56:368–375.CrossrefMedlineGoogle Scholar
  • 11. McGinn AP, Kaplan RC, Verghese J, Rosenbaum DM, Psaty BM, Baird AE, et al. Walking speed and risk of incident ischemic stroke among postmenopausal women.Stroke. 2008; 39:1233–1239.LinkGoogle Scholar
  • 12. Wannamethee SG, Whincup PH, Lennon L, Rumley A, Lowe GD. Fibrin D-dimer, tissue-type plasminogen activator, von Willebrand factor, and risk of incident stroke in older men.Stroke. 2012; 43:1206–1211.LinkGoogle Scholar
  • 13. Wannamethee SG, Welsh P, Lowe GD, Gudnason V, Di Angelantonio E, Lennon L, et al. N-terminal pro-brain natriuretic peptide is a more useful predictor of cardiovascular disease risk than C-reactive protein in older men with and without pre-existing cardiovascular disease.J Am Coll Cardiol. 2011; 58:56–64.CrossrefMedlineGoogle Scholar
  • 14. Lindsberg PJ, Grau AJ. Inflammation and infections as risk factors for ischemic stroke.Stroke. 2003; 34:2518–2532.LinkGoogle Scholar
  • 15. Klenk J, Denkinger M, Nikolaus T, Peter R, Rothenbacher D, Koenig W; ActiFE Study Group. Association of objectively measured physical activity with established and novel cardiovascular biomarkers in elderly subjects: every step counts.J Epidemiol Community Health. 2013; 67:194–197.CrossrefMedlineGoogle Scholar
  • 16. Wannamethee SG, Lowe GD, Whincup PH, Rumley A, Walker M, Lennon L. Physical activity and hemostatic and inflammatory variables in elderly men.Circulation. 2002; 105:1785–1790.LinkGoogle Scholar
  • 17. Walker M, Whincup PH, Shaper AG. The British Regional Heart Study 1975-2004.Int J Epidemiol. 2004; 33:1185–1192.CrossrefMedlineGoogle Scholar
  • 18. Emberson JR, Whincup PH, Morris RW, Walker M, Lowe GD, Rumley A. Extent of regression dilution for established and novel coronary risk factors: results from the British Regional Heart Study.Eur J Cardiovasc Prev Rehabil. 2004; 11:125–134.CrossrefMedlineGoogle Scholar
  • 19. Wannamethee SG, Lowe GD, Rumley A, Bruckdorfer KR, Whincup PH. Associations of vitamin C status, fruit and vegetable intakes, and markers of inflammation and hemostasis.Am J Clin Nutr. 2006; 83:567–74, quiz 726.CrossrefMedlineGoogle Scholar
  • 20. Shaper AG, Wannamethee G, Weatherall R. Physical activity and ischaemic heart disease in middle-aged British men.Br Heart J. 1991; 66:384–394.CrossrefMedlineGoogle Scholar
  • 21. Chomistek AK, Manson JE, Stefanick ML, Lu B, Sands-Lincoln M, Going SB, et al. 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.CrossrefMedlineGoogle Scholar
  • 22. Wendel-Vos GC, Schuit AJ, Feskens EJ, Boshuizen HC, Verschuren WM, Saris WH, et al. Physical activity and stroke. A meta-analysis of observational data.Int J Epidemiol. 2004; 33:787–798.CrossrefMedlineGoogle Scholar

eLetters(0)

eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.