Abstract
Objectives
We sought to examine pre-stroke lifestyle factorsassociated with post-stroke mortalityand recovery in older women.
Design
Longitudinal prospective cohort study.
Setting
The Women’s Health Initiative (clinical trials and observational study), 40 clinical centers in the U.S.
Participants
WHI participants, women aged 50–79, who were stroke-free at baseline 1993–1998, with incident stroke prior to 2005.
Measurements
Participants were followed for mortality through 2010. Pre-stroke characteristics were from the most proximal examination before the stroke event. Annual follow-up for clinical events ascertained hospitalization for stroke which was subsequently physician-adjudicated with medical records. Multivariable regression models analyzed factors associated with post-stroke mortality and post-stroke recovery at hospital discharge (post-stroke Glasgow score), adjusting for stroke type.
Results
Of 3,173 women with incident stroke, 1,111 (35%) died. Overweight and obese BMI pre-stroke was associated with reduced post-stroke mortality (vs. normal BMI, obese BMI hazards ratio, HR=0.69, 95% confidence interval (CI)=0.53–0.88; overweight BMI: HR=0.72, 95%CI=0.58–0.90); underweight pre-stroke BMI had borderline increased post-stroke mortality (HR=2.02, 95%CI=0.98–4.16).Other pre-stroke factors associated with post-stroke mortality included diabetes (HR=1.28, 95%CI=1.01–1.64), current smoking (vs. nonsmoker, hazard ratio (HR)=2.13, 95%CI=1.53–3.00), physical inactivity (vs. >150 minutes exercise/week, HR=1.39, 95%CI=1.09–1.78) and lowest physical function quartile (vs. highest, HR=1.54, 95%CI=1.18–2.02). Pre-stroke diabetes was associated with reduced odds of good recovery post-stroke (odds ratio (OR)= 0.60, 95%CI= 0.44–0.82). Current hormone use pre-stroke was associated with increased odds of moderate vs. severe disability post-stroke (OR=1.29, 95%CI=1.00–1.66).
Conclusion
Potentially modifiable factors pre-stroke, including smoking, diabetes and underweight BMI, were associated with increased post-stroke mortality in older women. Pre-stroke overweight or obese BMI and physical activity were associated with reduced post-stroke mortality in older women.
Keywords: stroke, mortality, women, BMI, recovery, diabetes
INTRODUCTION
Older age is a significant risk factor for stroke.1 Stroke is a leading cause of disability2,3 and the third leading cause of death in women.4 Older women may have lower functional recovery,5 higher risk of disability2 and higher mortality6 from stroke than older men.For clinicians caring for older women, a clear understanding of the risks and benefits are important when making recommendations for lifestyle modifications to reduce stroke morbidity and mortality.
Lifestyle modifications including weight reduction, physical exercise, and smoking cessation may be beneficial due to effects on diabetes and vascular disease7 and are recommended to reduce stroke incidence.1 Recent studies in non-stroke populations suggest that in older adults body mass index (BMI) may have an inverse relationship with overall mortality.8 Low BMI has been associated with increased overall mortality, while overweight BMI is associated with reduced overall mortality compared to normal BMI.9
Although obesity is a risk factor for stroke incidence,10 data on the relationships of BMI with post-stroke outcomes have been conflicting.11,12 For example, weight loss is recommended to reduce stroke morbidity and mortality,1 but poor nutritional status was associated with worse outcomes for hospitalized stroke patients in the FOOD trial.13 Data is elusive on the relationship between BMI and stroke mortality or recovery in older women, limiting clinicians’ ability to determine whether weight loss should be recommended for older overweight or obese women with stroke risk factors. Since most previous studies of post-stroke mortality were hospital-based, there are virtually no prospective data from large cohorts on the impact of pre-stroke lifestyle on post-stroke mortality or recovery. However, modifying lifestyle factors may help efforts to reduce post-stroke mortality and improve post-stroke recovery among older women.
The Women’s Health Initiative (WHI) provides a unique opportunity to examine lifestyle factors associated with post-stroke mortality and recovery in older women. This large, multi-center study collected data on older women who suffered first-time incident stroke. We sought to identify pre-strokelifestyle factors related to mortality and recovery after incident stroke among WHI participants.
METHODS
Sample population
The WHI study has been described in detail elsewhere.14–17 Data from the WHI were collected prospectively from 40 clinical centers to complete the randomized clinical trials (CTs) and observational study (OS). The 161,808 postmenopausal women participants aged 50–79 were recruited from 1993 through 1998. Women likely to die or relocate within three years were excluded. Each study center obtained informed consent and IRB approval. These analyses included women, stroke-free at baseline examination, with a first-time incident stroke before 2005. We excluded 2221 women with prevalent strokes at baseline, leaving 3,173 WHI participants who suffered first-time stroke between enrollment and 2005. Figure 1 displays the selection process of participants for the analytic sample.
Figure 1.
Flow of participants through study process.
Stroke Surveillance
Study neurologists reviewed and adjudicated all potential strokes using standardized criteria. We excluded events not requiring hospitalization and transient ischemic events. Stroke type was classified as ischemic, hemorrhagic or undefined, using brain imaging. Ischemic strokes were rapid onset of a persistent (>24 hours) neurological deficit from an occlusion of cerebral or precerebral arteries with infarction without evidence for other causes.18 Hemorrhagic strokes had blood in the subarachnoid space or intraparenchymal hemorrhage by computed tomography or magnetic resonance imaging. Undefined strokes satisfied criteria for stroke but had inadequate information for categorization as hemorrhagic or ischemic.19 Post-stroke Glasgow Outcomes Scale scores were determined at hospital discharge from hospital records with post-stroke recovery categorized as good recovery, moderate disability, severe disability and vegetative survival.20
Data collection
Pre-stroke baseline data were collected from questionnaires and physical examinations. The most proximal measure before the stroke was used. Covariates included cardiovascular risk factors and mortality risk factors.
Pre-stroke lifestyle factors considered potentially modifiable included BMI (underweight <18.5 kg/m2, normal 18.5–24.9 kg/m2, overweight 25.0–29.9 kg/m2, and obese ≥30 kg/m2);21 functional status (determined by Rand-36 physical function subscale); physical activity (minutes of moderate/vigorous activity per week: none, 1–74 minutes/week, 75–149 minutes/week and >150 minutes/week);smoking (current, past and never smoker); depressive symptoms (assessed using the Center for Epidemiologic Studies Depression Scale);22,23 hormone status (clinical trial vs. observational study enrollment or randomization were not controlled for) and prevalent diabetes and hypertension.
Other pre-stroke factors included prevalent diseases, income, marital status and social support. Pre-stroke prevalent diseases examined included CHD and cancer. Income levels pre-stroke were dichotomized by median $35,000/year. Marital status pre-stroke was dichotomized as married/like relationship vs. not married, divorced, widowed. Social support pre-stroke was assessed using the Medical Outcomes Study Social support questionnaire score. Date and cause of death were adjudicated through 2010.
Statistical methods
The primary endpoint was survival after incident stroke, time from incident stroke to death from any cause or end of follow up in the absence of death. Participants without a verified death date were censored at the date last known alive. Cox proportional hazards models assessed univariate associations between potential risk factors for the outcome of post-stroke good mortality. A secondary analysis explored factors related to post-stroke recovery through multinomial logistic regression models. We modeled post-stroke recovery as measured by the Glasgow outcomes scale comparing the outcomes of good recovery vs. severe disability or vegetative survival and moderate disability vs. severe disability or vegetative survival.20 In both sets of analyses, multivariable models included all covariates strongly related upon univariate analysis. The proportional hazards assumption was assessed, with appropriate adjustments made when non-proportionality was detected. All reported p-values are two-sided. Analyses used SAS (version 9.2, SAS Institute Inc., Cary, NC) and Stata (version 10.1, Stata Corp) software.
RESULTS
Of the 3,173 WHI participants who suffered first-time stroke between enrollment and 2005, 35% (n=1,111) died by 2010. The mean age at stroke was 72.6 years. The majority of women (71%) suffered ischemic strokes, 17% suffered hemorrhagic stroke and 12% suffered undefined strokes. Glasgow Outcome Scale at post-stroke hospital discharge showed good recovery in 32%, moderate disability in 36%, severe disability in 31% and vegetative survival in less than 1% in patients where it was available. The cause of death was cerebrovascular for 44% of women, cancer for 16% of women, and other heart disease for 17% of women. Table 1 displays the participants’ baseline characteristics.
Table 1.
Baseline Characteristics of Postmenopausal Women with Incident Stroke.
| Participant characteristic | Overall sample n= 3,173 (%a) |
Survivors n= 2,062 (65.0%) |
Decedents n=1,111 (35.0%) |
|---|---|---|---|
| Age at strokeb | 72.6 ± 6.8 | 71.8 ± 6.8 | 74.0 ± 6.4 |
| Race/ Ethnicity | |||
| White | 2616 (82.8) | 1679 (64.2) | 937 (35.8) |
| Black | 350 (11.1) | 244 (69.7) | 106 (30.3) |
| Hispanic | 76 (2.4) | 51 (67.1) | 25 (32.9) |
| Asian | 67 (2.1) | 48 (71.6) | 19 (28.4) |
| Other | 51 (1.6) | 32 (62.8) | 19 (37.3) |
| Marital status | |||
| Never/ not married | 1500 (47.5) | 924 (61.6) | 576 (38.4) |
| Currently married | 1656 (52.5) | 1129 (68.2) | 527 (31.8) |
| Social support scorec | 34.9 ± 8.1 | 35.1 ± 8.1 | 34.6 ± 8.2 |
| Education | |||
| Less than high school | 245 (7.8) | 151 (61.6) | 94 (38.4) |
| High school graduate | 589 (18.7) | 402 (68.3) | 187 (31.8) |
| > High school | 1263 (40.2) | 826 (65.4) | 437 (34.6) |
| ≥ College degree | 1048 (33.3) | 663 (63.3) | 385 (36.7) |
| Income | |||
| Income <$35,000 | 1648 (55.8) | 1028 (62.4) | 620 (37.6) |
| Income ≥$35,000 | 1308 (44.3) | 890 (68.0) | 418 (32.0) |
| Body mass index (BMI) | |||
| Underweight BMIa | 32 (1.0) | 15 (46.9) | 17 (53.1) |
| Normal BMI | 936 (29.8) | 557 (59.5) | 379 (40.5) |
| Overweight BMI | 1126 (35.9) | 750 (66.6) | 376 (33.4) |
| Obese BMI | 1044 (33.3) | 720 (69.0) | 324 (31.0) |
| Baseline diabetes | 499 (15.7) | 298 (59.7) | 201 (40.3) |
| Baseline hypertension | 2031 (64.0) | 1287 (63.4) | 744 (36.6) |
| Depressive symptomsb | 761 (24.3) | 497 (65.3) | 264 (34.7) |
| Smoking status | |||
| Never smoker | 1589 (50.8) | 1089 (68.5) | 500 (31.5) |
| Past smoker | 1270 (40.6) | 792 (62.4) | 478 (37.6) |
| Current smoker | 268 (8.6) | 153 (57.1) | 115 (42.9) |
| Pre-stroke physical activityc | |||
| No physical activity | 1386 (46.4) | 862 (62.2) | 524 (37.8) |
| 1–74 minutes/week | 574 (19.2) | 380 (66.2) | 194 (33.8) |
| 75–149 minutes/week | 354 (11.8) | 244 (68.9) | 110 (31.1) |
| >150 minutes /week | 675 (22.6) | 462 (68.4) | 213 (31.6) |
| Pre-stroke functional ability | |||
| Lowest function Q1 | 769 (24.5) | 437 (56.8) | 332 (43.2) |
| Function Q2 | 625 (19.9) | 396 (63.4) | 229 (36.6) |
| Function Q3 | 790 (25.1) | 524 (66.3) | 266 (33.7) |
| Highest function Q4 | 959 (30.5) | 688 (71.7) | 271 (28.3) |
| Hormone status | |||
| Never used hormones | 1106 (36.2) | 718 (64.9) | 388 (35.1) |
| Past hormone user | 838 (27.4) | 535 (63.8) | 303 (36.2) |
| Current hormone user | 1114 (36.4) | 746 (67.0) | 368 (33.0) |
| Prevalent diseases | |||
| TIA | 238 (7.5) | 137 (57.6) | 101 (42.4) |
| Coronary Heart Disease | 125 (3.9) | 63 (50.4) | 62 (49.6) |
| Cancer | 492 (15.5) | 258 (52.4) | 234 (47.6) |
| Stroke type | |||
| Ischemic Stroke type | 2255 (71.1) | 1565 (69.4) | 690 (30.6) |
| Hemorrhagic Stroke type | 540 (17.0) | 262 (48.5) | 278 (51.5) |
| Other Stroke type | 378 (11.9) | 235 (62.2) | 143 (37.8) |
| Post-Stroke Glasgow | |||
| Good recovery | 705 (32.1) | 559 (79.3) | 146 (20.7) |
| Moderately disabled | 782 (35.6) | 618 (79.0) | 164 (21.0) |
| Severely disabled | 692 (31.5) | 428 (61.9) | 264 (38.2) |
| Vegetative survival | 15 (0.7) | 6 (40.0) | 9 (60.0) |
Counts may not add up to overall totals due to missing data.
Age at stroke, range 51.08–87.95 years.
Social Support Score (Medical Outcomes Study): median 36.00, range 9–45, higher=greater support.
Mean BMI 28.50 ±6.13 (range 14.63–67.33), median 27.44.
Depressive symptoms(CESD-11 instrument).
Minutes of moderate to vigorous physical activity per week.
Table 2 displays the unadjusted associations between demographic or stroke characteristics and survival. Increased post-stroke mortality was associated with the following lifestyle factors: underweight BMI, smoking, lack of physical activity, lower pre-stroke physical function, diabetes and hypertension. Other factors associated with post-stroke mortality included increased age at stroke, unmarried status, lower income, previous TIA, heart disease and cancer. Hemorrhagic and other stroke types were associated with higher mortality risk compared to ischemic strokes. Severe post-stroke disability (hospital discharge Glasgow Outcome Scale) was associated with higher mortality risk than good function post-stroke.
Table 2.
Factors Associated with Mortality among Women after Incident Stroke (Crude Unadjusted Cox Proportional Hazards Models).
| Participant characteristic | Hazards Ratio (95% Confidence Intervals) | p value |
|---|---|---|
| Age at Stroke | 1.06 (1.05–1.07) | <.001 |
| Race/Ethnicity | White = reference | |
| Black | 0.86 (0.71–1.05) | 0.15 |
| Hispanic | 1.00 (0.67–1.49) | 0.99 |
| Asian | 0.82 (0.52–1.29) | 0.39 |
| Other | 1.09 (0.69–1.71) | 0.73 |
| Marital status | Currently married = reference | |
| Never/ not married | 1.28 (1.14–1.44) | <.001 |
| Social Support score | 0.99 (0.99–1.00) | 0.07 |
| Education | College graduate = reference | |
| More than high school | 0.97 (0.84–1.11) | 0.65 |
| High school graduate | 0.89 (0.75–1.06) | 0.18 |
| Less than high school | 1.14 (0.91–1.43) | 0.27 |
| Income ≥$35,000 | 0.79 (0.70–0.89) | <.001 |
| Body Mass Index (BMI) | Normal BMI =reference | |
| Underweight | 1.62 (1.00–2.64) | 0.05 |
| Overweight | 0.78 (0.67–0.89) | <.001 |
| Obese | 0.72 (0.62–0.83) | <.001 |
| Baseline diabetes | 1.27 (1.09–1.48) | 0.002 |
| Baseline hypertension | 1.27 (1.12–1.44) | <.001 |
| Baseline depressive symptoms | 0.98 (0.85–1.12) | 0.74 |
| Baseline smoking status | Never smoker = Reference | |
| Past smoker | 1.26 (1.11–1.43) | <.001 |
| Current smoker | 1.52 (1.24–1.86) | <.001 |
| Pre-stroke physical activity | >150 minutes /week = Reference | |
| No physical activity | 1.31 (1.11–1.53) | 0.001 |
| 1–74 minutes/week | 1.15 (0.95–1.40) | 0.16 |
| 75–149 minutes/week | 1.02 (0.81–1.28) | 0.90 |
| Pre-stroke functional ability | Highest function quartile 4 = Reference | |
| Lowest function quartile 1 | 1.84 (1.57–2.16) | <.001 |
| Function quartile 2 | 1.42 (1.19–1.69) | <.001 |
| Function quartile 3 | 1.27 (1.07–1.50) | 0.006 |
| Hormone status | Never used hormones = Reference | |
| Past hormone user | 1.03 (0.89–1.20) | 0.70 |
| Current hormone user | 0.92 (0.80–1.06) | 0.25 |
| History of previous TIA | 1.31 (1.07–1.61) | 0.009 |
| Baseline coronary heart disease | 1.78 (1.38–2.30) | <.001 |
| Baseline cancer | 1.75 (1.51–2.02) | <.001 |
| Incident stroke type | Ischemic stroke = Reference | |
| Hemorrhagic stroke | 2.29 (1.99–2.64) | <.001 |
| Other stroke type | 1.34 (1.12–1.61) | 0.001 |
| Glasgow Score Post-stroke | Good recovery = Reference | |
| Moderately disabled | 1.07 (0.85–1.33) | 0.58 |
| Severely disabled | 2.57 (2.10–3.15) | <.001 |
| Vegetative survival | 6.07 (3.09–11.90) | <.001 |
Table 3 displays the final adjusted multivariable Cox regression model for the outcome of post-stroke mortality. Overweight and obese BMI pre-stroke was associated with reduced post-stroke mortality compared to normal BMI. Underweight pre-stroke BMI was borderline significantly associated with increased post-stroke mortality compared to normal BMI. Pre-stroke diabetes was associated with increased odds of post-stroke mortality. Other modifiable factors associated with post-stroke mortality included smoking, physical inactivity and low physical function. Prevalent hypertension demonstrated a borderline significantly association with increased post-stroke mortality. Worsened post-stroke recovery on Glasgow Outcome Scale at the time of discharge from stroke hospitalization was associated with post-stroke mortality, withhighest association at 1-year post-stroke and decreasing associations out to 10-years post-stroke.
Table 3.
Factors Associated with Mortality among Women after Incident Stroke (Multivariable Cox Proportional Hazards Model, N=2,749)
| Participant characteristic | Adjusted Hazards Ratio (95% CI) |
p value |
|---|---|---|
| Age at stroke | 1.07 (1.05–1.08) | <.001 |
| Not marrieda vs. married | 0.95 (0.77–1.16) | 0.59 |
| Social support score | 1.00 (0.98–1.01) | 0.39 |
| Income <$35,000 vs. ≥$35,000 | 1.11 (0.91–1.37) | 0.31 |
| Normal BMI | Reference | |
| Underweight BMI | 2.02 (0.98–4.16) | 0.06 |
| Overweight BMI | 0.72 (0.58–0.90) | 0.004 |
| Obese BMI | 0.69 (0.53–0.88) | 0.003 |
| Baseline diabetes | 1.28 (1.01–1.64) | 0.045 |
| Baseline hypertension | 1.19 (0.98–1.45) | 0.09 |
| Never smoker | Reference | |
| Current smoker | 2.13 (1.53–3.00) | <.001 |
| Past smoker | 1.48 (1.22–1.79) | <.001 |
| Pre-stroke physical activity | ||
| No physical activity | 1.39 (1.09–1.78) | 0.009 |
| 1–74 minutes/week | 1.32 (0.99–1.76) | 0.06 |
| 75–149 minutes/week | 1.14 (0.82–1.58) | 0.44 |
| >150 minutes /week | Reference | |
| Pre-stroke functional ability | ||
| Lowest function quartile 1 | 1.54 (1.18–2.02) | 0.002 |
| Function quartile 2 | 1.30 (0.99–1.72) | 0.06 |
| Function quartile 3 | 1.23 (0.95–1.57) | 0.1 |
| Highest function quartile 4 | Reference | |
| Incident stroke type | ||
| Ischemic stroke | Reference | |
| Hemorrhagic stroke | 1.02 (0.77–1.36) | 0.87 |
| Other stroke type | 1.33 (0.80–2.20) | 0.28 |
| Prevalent conditions pre-stroke | ||
| TIA prior to stroke | 0.89 (0.60–1.33) | 0.57 |
| Coronary heart disease | 1.39 (0.92–2.11) | 0.12 |
| Cancer | 1.73 (1.39–2.15) | <0.0001 |
| Post-stroke Glasgow outcome scale | ||
| Good recovery | Reference | |
| Moderately disabled | 1.05 (0.78–1.40) | 0.75 |
| Severely disabled /vegetative survivala 1 year | 2.58 (1.99–3.33) | <.001 |
| Severely disabled / vegetative survival 5 years | 1.69 (1.26–2.26) | <.001 |
| Severely disabled / vegetative survival 10 years | 1.40 (0.97–2.03) | 0.07 |
Never married, divorced or widowed vs. currently married
Glasgow outcome scale score at stroke hospitalization discharge as a time varying covariate for the outcome of post-stroke mortality, (the hazard of death at year 1 for those “Severe or worse” on the Glasgow scale is approximately 2.58 times that of a woman with “Good” Glasgow score, and this risk decreases to 1.4 times by year 10).
Table 4 displays the final adjusted multivariable multinomial logistic regression models for the outcomes of post-stroke recovery. These models examined the outcomes of good recovery vs. severe disability or vegetative survival and moderate disability vs. severe disability or vegetative survival. Pre-stroke diabetes was associated with reduced odds of good recovery post-stroke. Current hormone use pre-stroke was associated with increased odds of moderate vs. severe disability post-stroke. Older age, unmarried status, and hemorrhagic (vs. ischemic) stroke type were also associated with reduced odds of good post-stroke recovery at hospital discharge. Pre-stroke BMI categories were not associated with likelihood of post-stroke recovery at hospital discharge (data not shown).
Table 4.
Factors Associated with Short-Term Recovery among Women at Hospital Discharge Post-stroke (Multivariable Cox Proportional Hazards Model, N=2,118)
| Participant characteristic | Odds of good recovery vs. severe disability/vegetative state |
Odds of moderate disability vs. severe disability/vegetative state |
p value |
|---|---|---|---|
| Age at stroke (1 year increase) | 0.95 (0.93 – 0.96) | 0.96 (0.95 – 0.98) | <0.001 |
| Not married vs. married | 0.79 (0.64 – 0.99) | 0.72 (0.58 – 0.89) | 0.010 |
| Stroke type | |||
| Ischemic | Reference | Reference | <0.001 |
| Hemorrhagic | 0.29 (0.20 – 0.40) | 0.33 (0.24 – 0.46) | |
| Other stroke type | 3.01 (1.55 – 6.13) | 2.21 (1.09 – 4.44) | |
| Baseline diabetes | 0.60 (0.44 – 0.82) | 0.98 (0.74 – 1.31) | 0.001 |
| Hormone status | |||
| Never used | Reference | Reference | 0.075 |
| Past user | 0.85 (0.65 – 1.13) | 1.20 (0.92 – 1.56) | |
| Current user | 0.99 (0.76 – 1.28) | 1.29 (1.00 – 1.66) |
DISCUSSION
This study examined 3,173 post-menopausal women aged 51 to 88 years at the time of stroke and is the first prospective study toexamine the relationship between BMI and post-stroke mortality and recovery specifically in older women. Overweight and obese older women had decreased likelihood of post-stroke mortality compared to women with normal BMI, and underweight women had a borderline increased likelihood of post-stroke mortality compared to normal BMI. Additional potentially modifiable pre-stroke factors associated with post-stroke mortality included diabetes, smoking, physical activity and physical function. Important and potentially modifiable pre-stroke lifestyle factors associated with post-stroke function at the time of acute hospital dischargeincluded diabetes and pre-stroke current hormone use.
Overweight and obese pre-stroke was associated with a 30% decreased likelihood of post-stroke mortality compared to women with normal BMI. Women with underweight pre-stroke BMI had double the mortality risk of women with normal BMI, even after controlling for smoking, cancer, diabetes, cardiovascular risk factors and physical activity. To our knowledge, there are no previous population-based studies of the relationship between prospectively measured pre-stroke BMI and post-stroke mortality among older women in the U.S. In previous studies, BMI has been associated with decreased likelihood of post-stroke mortality among men and women of all ages hospitalized for stroke in Greece11 and Japan.24 The FOOD Trial Collaboration found that male and female hospitalized stroke patients (mean age 73 years) who appeared to be undernourished suffered more complications post-stroke, including pneumonia, infections and gastrointestinal hemorrhage.13 Underweight BMI may indicate frailty,21 a mortality risk factor in the WHI Observational Study.25
Our study identified several additional other modifiable pre-stroke factors associated with post-stroke mortality and recovery. Pre-stroke diabetes was associated with decreased odds of post-stroke mortality. Pre-stroke diabetes was also associated with decreased odds of good recovery post-stroke at the time of hospital discharge. Pre-stroke smoking more than doubled the post-stroke mortality risk in older women, after controlling for cancer, other cardiovascular risk factors, stroke type and Glasgow Outcome Scale. Past smoking increased the risk by 49%. Carter examined male and female stroke patients in four hospitals in England combined and reported that smoking was not a significant predictor of post-stroke mortality in multivariate analyses.26 Lack of physical activity pre-stroke was associated with a 39% increased risk of post-stroke mortality compared to women who exercised more than 150 minutes per week. This risk factor may be modifiable through exercise and activity programs for older women. Although low physical activity is a known stroke risk factor,27 no previous study has examined the relationship between physical activity prior to stroke and post-stroke mortality. The only prior study to examine physical activity and post-stroke mortality was a Norwegian study by Engstad which collected data on the Frenchay Activity Index as a measure of activity level after stroke and examined long-term post-stroke mortality comparing to a group of stroke-free subjects. In that study, the post-stroke activity level may have been a function of the stroke severity, rather than a modifiable lifestyle factor, as our pre-stroke physical activity measure represents.28 Lower pre-stroke function was associated with higher risk of mortality post-stroke. No prior study of post-stroke mortality examined prospectively measured pre-stroke function. Few studies have examinedretrospectively assessed pre-stroke function (as prior difficulty walking obtained during stroke admission evaluation)and post-stroke mortality.29,30 Post-stroke mortality risk progressively increased with decreasing levels of pre-stroke function.
This study is one of the largest studies of post-stroke mortality of women and included pre-stroke comorbid diseases, lifestyle and physical function not available in many stroke database studies. The WHI study cohort includes a very large, ethnically and geographically diverse U.S. population and long follow-up period. The stroke data were carefully measured and prospectively collected using standardized criteria. Loss to follow up was minimal.29
This study had several limitations. Not all strokes were typed due to death before imaging. Only strokes with hospitalization were included. Hospital stroke management, medication use, hyperlipidemia, peripheral artery disease, alcohol use, cognitive function and atrial fibrillation were not included in this study.
Further research is needed to identify potential interventions to reduce post-stroke mortality. Smoking cessation and increasing physical activity may be important areas for future study in older women. Low protein diets have been associated with frailty risk.30 Clinical experience suggests that higher BMI could be protective against post-stroke mortality in older women due to increased BMI reserve in the post-stroke period, when dysphagia and nutritional intake issues may be especially problematic for women with lower BMI. Increased dietary protein in underweight older women may be an important potential intervention to reduce post-stroke mortality. Our research suggests that physical activity may be a beneficial recommendation for older women to reduce stroke morbidity and mortality. More research is needed to understand the implications of lifestyle modification among older women at risk for stroke morbidity and mortality.
Conclusion
This study of 3,173 post-menopausal women in the WHI aged 51 to 88 years at the time of stroke is the first to examine prospectively collected lifestyle factors before stroke and found that overweight and obese BMI is associated with reduced post-stroke mortality compared to normal BMI, while underweight BMI was associated with increased post-stroke mortality. Pre-stroke smoking and low physical function were also associated with increased post-stroke mortality; physical activity pre-stroke was associated with decreased post-stroke mortality. Pre-stroke diabetes was associated with reduced post-stroke recovery at hospital discharge.
ACKNOWLEDGMENTS
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.
Conflict of Interest:
Drs. Bell, Masaki, Manini and Wassertheil-Smoller received WHI grant funding from NHLBI.
Dr. Wassertheil-Smoller received an honorarium for a talk at Columbia University.
The WHI is funded by the National Heart, Lung, and Blood Institute, NIH, U.S. Department of Health and Human Services (HHSN268201100046C, HHSN268201100001C, HHSN268201100002C, HHSN268201100003C, HHSN268201100004C). This research was supported by WHl Extension 2010–2015 Western Regional Subcontract through Stanford University from the National Heart, Lung and Blood Institute, National Institutes of Health.
Footnotes
An abstract of this paper was presented at the American Geriatrics Society Annual Meeting Presidential Poster Session, Seattle WA, May 2012.
Author Contributions:
Study concept and design: CB, AL, KM, JDC, SWS
Acquisition of data: AL, SWS (representing the WHI)
Analysis and interpretation of data: CB, KM, EH, TM, JDC, SWS
Preparation of manuscript: CB, KM
Critical revision of manuscript: AL, KM, EH, TM, JM, JDC, SWS
Sponsor’s Role: The funding sources had no role in the analysis and preparation of this paper.
REFERENCES
- 1.Goldstein LB, Bushnell CD, Adams RJ, et al. Guidelines for the primary prevention of stroke: A guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2011;42:517–584. doi: 10.1161/STR.0b013e3181fcb238. [DOI] [PubMed] [Google Scholar]
- 2.Persky RW, Turtzo LC, McCullough LD. Stroke in women: Disparities and outcomes. Curr Cardiol Rep. 2010;12:6–13. doi: 10.1007/s11886-009-0080-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Donnan GA, Fisher M, Macleod M, et al. Stroke. Lancet. 2008;371:1612–1623. doi: 10.1016/S0140-6736(08)60694-7. [DOI] [PubMed] [Google Scholar]
- 4.Wassertheil-Smoller S. Stroke in women. Nutr Metab Cardiovasc Dis. 2010;20:419–425. doi: 10.1016/j.numecd.2010.02.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gargano JW, Reeves MJ. Sex differences in stroke recovery and stroke-specific quality of life: results from a statewide stroke registry. Stroke. 2007;38:2541–2548. doi: 10.1161/STROKEAHA.107.485482. [DOI] [PubMed] [Google Scholar]
- 6.Roger VL, Go AS, Lloyd-Jones DM, et al. Heart Disease and Stroke Statistics-2012 Update: A Report From the American Heart Association. Circulation. 2012;125:188–197. doi: 10.1161/CIR.0b013e3182456d46. [DOI] [PubMed] [Google Scholar]
- 7.Ringelstein EB, Nabavi D. Long-term prevention of ischaemic stroke and stroke recurrence. Thromb Res. 2000;98:83–96. doi: 10.1016/s0049-3848(00)00230-9. [DOI] [PubMed] [Google Scholar]
- 8.Newman AB, Yanez D, Harris T, et al. Weight change in old age and its association with mortality. J Am Geriatr Soc. 2001;49:1309–1318. doi: 10.1046/j.1532-5415.2001.49258.x. [DOI] [PubMed] [Google Scholar]
- 9.Flegal KM, Kit BK, Orpana H, et al. Association of all-cause mortality with overweight and obesity using standard body mass index categories: A systematic review and meta-analysis. JAMA. 2013;309:71–82. doi: 10.1001/jama.2012.113905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kurth T, Gaziano JM, Rexrode KM, et al. Prospective study of body mass index and risk of stroke in apparently healthy women. Circulation. 2005;111:1992–1998. doi: 10.1161/01.CIR.0000161822.83163.B6. [DOI] [PubMed] [Google Scholar]
- 11.Vemmos K, Ntaios G, Spengos K, et al. Association between obesity and mortality after acute first-ever stroke: the obesity-stroke paradox. Stroke. 2011;42:30–36. doi: 10.1161/STROKEAHA.110.593434. [DOI] [PubMed] [Google Scholar]
- 12.Whitlock G, Lewington S, Sherliker P, et al. Body-mass index and cause-specific mortality in 900 000 adults: Collaborative analyses of 57 prospective studies. Lancet. 2009;373:1083–1096. doi: 10.1016/S0140-6736(09)60318-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Poor nutritional status on admission predicts poor outcomes after stroke: Observational data from the FOOD trial. Stroke. 2003;34:1450–1456. doi: 10.1161/01.STR.0000074037.49197.8C. [DOI] [PubMed] [Google Scholar]
- 14.Anderson GL, Manson J, Wallace R, et al. Implementation of the Women's Health Initiative study design. Ann Epidemiol. 2003;13:S5–S17. doi: 10.1016/s1047-2797(03)00043-7. [DOI] [PubMed] [Google Scholar]
- 15.Langer RD, White E, Lewis CE, et al. The Women's Health Initiative Observational Study: baseline characteristics of participants and reliability of baseline measures. Ann Epidemiol. 2003;13:S107–S121. doi: 10.1016/s1047-2797(03)00047-4. [DOI] [PubMed] [Google Scholar]
- 16.Ritenbaugh C, Patterson RE, Chlebowski RT, et al. The Women's Health Initiative Dietary Modification trial: overview and baseline characteristics of participants. Ann Epidemiol. 2003;13:S87–S97. doi: 10.1016/s1047-2797(03)00044-9. [DOI] [PubMed] [Google Scholar]
- 17.Stefanick ML, Cochrane BB, Hsia J, et al. The Women's Health Initiative postmenopausal hormone trials: overview and baseline characteristics of participants. Ann Epidemiol. 2003;13:S78–S86. doi: 10.1016/s1047-2797(03)00045-0. [DOI] [PubMed] [Google Scholar]
- 18.McGinn AP, Kaplan RC, Verghese J, et al. Walking speed and risk of incident ischemic stroke among postmenopausal women. Stroke. 2008;39:1233–1239. doi: 10.1161/STROKEAHA.107.500850. [DOI] [PubMed] [Google Scholar]
- 19.Curb JD, McTiernan A, Heckbert SR, et al. Outcomes ascertainment and adjudication methods in the Women's Health Initiative. Ann Epidemiol. 2003;13:S122–S128. doi: 10.1016/s1047-2797(03)00048-6. [DOI] [PubMed] [Google Scholar]
- 20.Albanese MA, Clarke WR, Adams HP, Jr, et al. Ensuring reliability of outcome measures in multicenter clinical trials of treatments for acute ischemic stroke. Theprogram developed for the Trial of Org 10172 in Acute Stroke Treatment (TOAST) Stroke. 1994;25:1746–1751. doi: 10.1161/01.str.25.9.1746. [DOI] [PubMed] [Google Scholar]
- 21.LaCroix AZ, Gray SL, Aragaki A, et al. Statin use and incident frailty in women aged 65 years or older: prospective findings from the Women's Health Initiative Observational Study. J Gerontol A Biol Sci Med Sci. 2008;63:369–375. doi: 10.1093/gerona/63.4.369. [DOI] [PubMed] [Google Scholar]
- 22.LS R. The CES-D scale: A self-report depression scale for research in the general population. Appl Psychol Meas. 1977;1:385–401. [Google Scholar]
- 23.Wassertheil-Smoller S, Shumaker S, Ockene J, et al. Depression and cardiovascular sequelae in postmenopausal women. The Women's Health Initiative (WHI) Arch Intern Med. 2004;164:289–298. doi: 10.1001/archinte.164.3.289. [DOI] [PubMed] [Google Scholar]
- 24.Kiyohara Y, Kubo M, Kato I, et al. Ten-year prognosis of stroke and risk factors for death in a Japanese community: the Hisayama study. Stroke. 2003;34:2343–2347. doi: 10.1161/01.STR.0000091845.14833.43. [DOI] [PubMed] [Google Scholar]
- 25.Woods NF, LaCroix AZ, Gray SL, et al. Frailty: Emergence and consequences in women aged 65 and older in the Women's Health Initiative Observational Study. J Am Geriatr Soc. 2005;53:1321–1330. doi: 10.1111/j.1532-5415.2005.53405.x. [DOI] [PubMed] [Google Scholar]
- 26.Carter AM, Catto AJ, Mansfield MW, et al. Predictive variables for mortality after acute ischemic stroke. Stroke. 2007;38:1873–1880. doi: 10.1161/STROKEAHA.106.474569. [DOI] [PubMed] [Google Scholar]
- 27.Willey JZ, Moon YP, Paik MC, et al. 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]
- 28.Engstad T, Viitanen M, Arnesen E. Predictors of death among long-term stroke survivors. Stroke. 2003;34:2876–2880. doi: 10.1161/01.STR.0000101751.20118.C1. [DOI] [PubMed] [Google Scholar]
- 29.Dallas MI, Rone-Adams S, Echternach JL, et al. Dependence in prestroke mobility predicts adverse outcomes among patients with acute ischemic stroke. Stroke. 2008;39:2298–2303. doi: 10.1161/STROKEAHA.107.506329. [DOI] [PubMed] [Google Scholar]
- 30.Hankey GJ, Jamrozik K, Broadhurst RJ, et al. Five-year survival after first-ever stroke and related prognostic factors in the Perth Community Stroke Study. Stroke. 2000;31:2080–2086. doi: 10.1161/01.str.31.9.2080. [DOI] [PubMed] [Google Scholar]