HHS Public Access

Author manuscript
Am J Med. Author manuscript; available in PMC 2017 July 01.
Author Manuscript

Published in final edited form as:

Am J Med. 2016 July ; 129(7): 715–724.e2. doi:10.1016/j.amjmed.2016.02.002.

Healthy Lifestyle and Functional Outcomes from Stroke in

Women
Pamela M. Rist, ScD, Julie E. Buring, ScD, Carlos S. Kase, MD, and Tobias Kurth, MD, ScD
Division of Preventive Medicine (P.M.R., J.E.B., T.K.), Department of Medicine, Brigham and
Women’s Hospital, Harvard Medical School, Boston, MA; the Department of Epidemiology
(P.M.R., J.E.B., T.K.), Harvard T.H. Chan School of Public Health, Boston, MA; Department of
Author Manuscript

Neurology (C.S.K), Boston University School of Medicine, Boston, MA; Institute of Public Health
(T.K.), Charité Universitätzmedizin, Berlin, Germany

Abstract
Purpose—While a healthy lifestyle has been associated with reduced risk of developing
ischemic stroke, less is known about its effect on stroke severity.

Methods—We performed a prospective cohort study among 37,634 women without stroke or
missing risk factor data at baseline. The healthy lifestyle index was composed of smoking,
physical activity, body mass index, alcohol consumption, and diet (range 0–20, with 20
representing healthiest lifestyle). Possible functional outcomes were no stroke or stroke with
modified Rankin Scale (mRS) score of 0–1 (mild), 2–3 (moderate), or 4–6 (severe). Multinomial
logistic regression was used to analyze the association between healthy lifestyle and functional
Author Manuscript

outcomes from stroke.

Results—Over 17.2 years of follow-up, 867 total strokes were confirmed. Compared to the
lowest category (0–4), the highest category (17–20) was associated with reductions in risk of total
stroke with mild (OR=0.43; 95% CI: 0.20–0.90), moderate (OR=0.53; 95% CI: 0.27–1.06) and
severe (OR=0.48; 95% CI: 0.20–1.18) functional outcomes. Even a modest healthy lifestyle index
(5–8 points) was associated with significant decreases in total stroke with severe and moderate
functional outcomes. Similar results were seen for ischemic but not hemorrhagic strokes.

Conclusions—Highest versus lowest scores on the healthy lifestyle index were associated
with reductions in risk of total and ischemic strokes with mild, moderate, and severe functional
outcomes among women. The evidence that even modest healthy lifestyle index scores reduced
risks of total and ischemic stroke with moderate and severe functional outcomes suggests modest
Author Manuscript

lifestyle changes may reduce risk of disabling stroke events.

Corresponding Author: Pamela M. Rist, ScD, 900 Commonwealth Avenue, 3rd floor, Boston, MA 02215, Tel: (617) 278-0835, Fax:
(617) 731-3843, prist@mail.harvard.edu.
The authors report no conflicts of interest.
All authors had access to the data and a role in writing the manuscript.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our
customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of
the resulting proof before it is published in its final citable form. Please note that during the production process errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
 Rist et al. Page 2

Keywords
Author Manuscript

Epidemiology; stroke outcomes; lifestyle

Introduction
Stroke is a leading cause of morbidity and mortality worldwide.1,2 There is growing interest
in determining whether a “healthy lifestyle” reduces the risk of stroke events. Previous
studies among women have shown a decreased risk of the development of total and ischemic
stroke with healthier lifestyles.3–6 However, research on healthy lifestyle and functional
outcomes from stroke in initially healthy populations is limited. Most research on functional
stroke outcomes has focused on the effect of single risk factors7–23 and has not considered
how a combination of factors may interact to influence total and ischemic stroke functional
Author Manuscript

outcomes. Given the morbidity and mortality consequences of stroke, it is important to

determine if healthy lifestyles also decrease stroke severity.

Results for the effect of healthy lifestyle on hemorrhagic stroke are mixed. One study
observed a decreased risk with healthier lifestyles6, while other studies have observed no
association5 or suggested a U-shaped association where healthier lifestyle was associated
with a non-significant increase in the risk of hemorrhagic stroke.3 None of these studies
examined the impact of healthy lifestyle on functional outcomes from hemorrhagic stroke.

To determine the associations between healthy lifestyle and the risks of functional outcomes
from total, ischemic, and hemorrhagic stroke, we used data from the Women’s Health Study,
a large prospective cohort of initially healthy women with available information on lifestyle
factors and functional outcomes after incident stroke.
Author Manuscript

Methods
The Women’s Health Study (WHS) was a randomized, placebo controlled trial of the effects
of low-dose aspirin and vitamin E in the primary prevention of cardiovascular disease and
cancer. The design, methods, and main findings have been published.24,25 Briefly, at
baseline (1992–1996), the study randomized 39,876 US female health professionals aged
≥45 years or older without a history of cardiovascular disease, cancer or other major
illnesses to receive low dose aspirin and/or vitamin E (versus placebo) in a 2 by 2 factorial
design. After the end of the trial in March 2004, women continue to be followed on an
observational basis. Twice during the first year and yearly thereafter, women were sent
follow-up questionnaires asking about demographic, lifestyle and health information,
Author Manuscript

including the occurrence of stroke events.

The WHS was approved by the Institutional Review Board at Brigham and Women’s
Hospital; all subjects provided written informed consent.

Exposure
We used a healthy lifestyle index previously developed in the WHS3, which contains similar
components to other lifestyle indices associated with stroke risk4–6, to evaluate the

Am J Med. Author manuscript; available in PMC 2017 July 01.

 Rist et al. Page 3

relationship of healthy lifestyle with incident stroke. The healthy lifestyle index incorporates
Author Manuscript

information from the baseline questionnaire on smoking, physical activity, body mass index
(BMI), alcohol consumption, and diet. For each variable, we created five categories and
assigned scores from 0 to 4 with higher scores indicating healthier behaviors. Smoking was
categorized as: current smoker who smokes ≥15 cigarettes per day (0 points), current smoker
who smokes <15 cigarettes per day (1 point), past smoker who smoked ≥20 pack-years (2
points), past smoker who smoked <20 pack-years (3 points), and never smoker (4 points).
Physical activity was categorized based on frequency of strenuous exercise: rarely or never
(0 points), <1 time per week (1 point), 1 time per week (2 points), 2 to 3 times per week (3
points), or ≥4 times per week (4 points). Body mass index was categorized as ≥35.0 kg/m2
(0 points), 30.0–34.9 kg/m2 (1 point), 25.0–29.9 kg/m2 (2 points), 22.0–24.9 kg/m2 (3
points), and >22.0 kg/m2 (4 points). To reflect the J-shaped relationship between alcohol
consumption and cardiovascular disease risk26, we assigned the least number of points to
Author Manuscript

those who rarely or never drank alcohol and the highest number of points to those who
consume light to moderate amounts of alcohol. Alcohol intake was categorized as never (0
points), <1 drink/week (1 point), 1 to 3 drinks/week (3 points), 4 to less than 10.5 drinks per
week (4 points), and ≥10.5 drinks per week (2 points). The construction of our diet score has
been described in detail previously.3 Briefly, women completed a 161-item standardized
food frequency questionnaire at baseline.27 We examined intake of cereal fiber, folate, ratio
of polyunsaturated to saturated fat, omega-3 fatty acids, trans fats, and glycemic load. Each
item was grouped into deciles and scored from 0 (least healthy) to 9 (healthiest) (trans fat
and glycemic load were scored inversely). The scores were summed to create a total diet
score which was then grouped into quintiles. The lowest quintile (representing the least
healthy diet) received 0 points while the highest quintile (representing the healthiest diet)
received 4 points.
Author Manuscript

The individual scores from the five components were summed to create a total final healthy
lifestyle index ranging from 0 to 20. We categorized the final score into five categories for
our total and ischemic stroke analyses: 0 to 4 (least healthy category and the reference
category), 5 to 8, 9 to 12, 13 to 16, and 17 to 20 (healthiest category). Due to the small
number of hemorrhagic strokes among those with the lowest healthy lifestyle index scores,
our exposure categories for our hemorrhagic stroke analyses were 0 to 8 (least healthy
category and the reference category), 9 to 12, 13 to 16 and 17 to 20 (healthiest group).

Outcome Ascertainment
If a woman reported a stroke on her follow-up questionnaire, we asked for permission to
review her medical records. An Endpoints Committee of physicians, including a board-
certified vascular neurologist (CSK) blinded to randomized treatment assignment, reviewed
Author Manuscript

medical records and determined if the self-reported stroke was confirmed. A nonfatal stroke
was defined as a focal neurologic deficit of sudden or rapid onset and vascular mechanism
that lasted >24 hours. Fatal strokes were confirmed using all available sources, including
death certificates and hospital records, to determine if there was evidence of a
cerebrovascular mechanism. Strokes were classified according to major subtype (ischemic,
hemorrhagic, or unknown) with excellent interobserver agreement (Cohen’s κ =0.96)28 and

Am J Med. Author manuscript; available in PMC 2017 July 01.

 Rist et al. Page 4

assigned a modified Rankin Scale (mRS) score based on the degree of impairment at
Author Manuscript

hospital discharge. Only the first stroke event was used in our analyses.

The mRS is a measure of functional outcome from stroke and ranges from 0 (no symptoms)
to 6 (death). We decided a priori to categorize the mRS score into three categories, similar to
previous studies.29–32 Our possible outcomes are no stroke, stroke with mRS 0 to 1 (no
symptoms or no significant disability), stroke with mRS 2 to 3 (slight to moderate
disability), and stroke with mRS 4 to 6 (moderately severe disability to death). Due to the
small number of hemorrhagic strokes, we collapsed mRS 2 to 3 and mRS 4 to 6 for the
hemorrhagic stroke analyses.

Statistical Analysis
Of the 39,876 participants in the WHS, we excluded six women who reported a stroke
before the baseline questionnaire and 2,236 women missing information on the lifestyle
Author Manuscript

factors that comprise the healthy lifestyle score, leaving 37,634 women eligible for our
analyses. We used multinomial logistic regression to calculate age- and multivariable-
adjusted odds ratios and 95% confidence intervals as a measure of the relative risk of the
association between higher categories of healthy lifestyle index and functional outcomes
from stroke compared with the lowest category of the healthy lifestyle index.

Multivariable-adjusted model 1 adjusted for potential confounders as well as randomized

treatment assignment. These potential confounders included age, postmenopausal hormone
use, ever use of oral contraceptives, family history of myocardial infarction before age 60
years, migraine, annual household income, highest level of education, location of residence,
marital status, and ethnicity. Due to the smaller number of hemorrhagic strokes, our first
multivariable-adjusted model for hemorrhagic stroke did not adjust for race or education
Author Manuscript

level.

Multivariate model 2 adjusted for potential consequences of an unhealthy lifestyle in

addition to all covariates included in the first multivariate model. These include self-report
of history of diabetes mellitus, history of hypertension, antihypertensive treatment, and
history of elevated cholesterol. For our hemorrhagic stroke analyses, we were not able to
adjust for diabetes status because there were too few diabetics.

In sensitivity analyses, we used multinomial logistic regression to explore the association

between functional outcome from stroke and a healthy lifestyle index that did not include
BMI as one of its components, as BMI may be considered a consequence of the other
components of the healthy lifestyle index.
Author Manuscript

All analyses were performed with SAS version 9.3 (SAS Institute, Cary, NC). P-values were
two-tailed with p<0.05 considered statistically significant.

Results
Table 1 shows the baseline characteristics of the WHS participants by healthy lifestyle index
categories. As defined, those in the lowest category of the healthy lifestyle index were most
likely to be current smokers who smoke ≥15 cigarettes per day, rarely or never perform

Am J Med. Author manuscript; available in PMC 2017 July 01.

 Rist et al. Page 5

physical activity, have a BMI ≥35 kg/m2, never consume alcohol, and have a diet score in
Author Manuscript

the lowest quintile. As the healthy lifestyle index increased, there was a decreasing
prevalence of hypertension, antihypertensive medication use, elevated cholesterol, family
history of myocardial infarction, black ethnicity, household income <$50,000 per year, and
having less than a bachelor’s degree, and there was an increasing prevalence of current
postmenopausal hormone use, being married, living in the west, and no history of migraine.
There was no clear pattern with ever use of oral contraceptives.

Over 17.2 years of follow-up, 867 total strokes were self-reported (707 ischemic, 156
hemorrhagic, and 4 unknown subtype). Compared with the lowest category of the healthy
lifestyle index (0–4 points), higher healthy lifestyle index categories were associated with
reductions in the risk of mild, moderate, and severe stroke outcomes (Table 2). Compared
with the lowest category (0–4), the highest category (17–20) was associated with a
significant reduction in risk of total stroke with mild functional outcomes (OR=0.43; 95%
Author Manuscript

CI: 0.20, 0.90) and with non-significant reductions in the risk of total stroke with moderate
(OR=0.53; 95% CI: 0.27, 1.06) and severe (OR=0.48; 95% CI: 0.20, 1.18) functional
outcomes. In other words, compared with those with a low healthy lifestyle index score (0 to
4), those with the highest healthy lifestyle index scores (17 to 20) had a 57% reduced risk of
mild stroke outcomes, a 47% reduced risk of moderate stroke outcomes, and a 52% reduced
risk of severe stroke outcomes. However, even those with a healthy index score of 5 to 8 had
a significantly decreased risk of severe stroke (OR=0.54; 95% CI: 0.30, 0.99) and moderate
stroke (OR=0.57; 95% CI: 0.36. 0.92) and a non-significant decrease in the risk of mild
score (OR=0.71; 95% CI: 0.45, 1.14) compared to women with a health index score of 0 to
4.

Controlling for potential intermediates slightly attenuated the effect estimates and some
Author Manuscript

results were no longer statistically significant (Table 2) suggesting that the health index
impacts risk of functional outcomes from total stroke through pathways involving these
potential intermediates. However, the direction of the effects still indicated that compared to
the lowest category of the healthy lifestyle index, higher categories of the healthy lifestyle
index were associated with reductions in the risk of mild, moderate, and severe functional
outcomes from total stroke.

The overall pattern of results from functional outcomes from ischemic stroke was similar to
total stroke. Compared to the lowest healthy lifestyle index category, higher healthy lifestyle
index categories were associated with a decrease in the risk of mild, moderate, and severe
ischemic stroke (Table 3). These results were slightly attenuated when we controlled for
potential intermediates.
Author Manuscript

Unlike the total and ischemic stroke analyses, those with the highest healthy lifestyle index
scores had an increased risk of hemorrhagic stroke with mild outcome (RR=4.23, 95% CI:
1.20, 14.99) which persisted after adjustment for potential intermediates (RR=4.12, 95% CI:
1.15, 14.72) (Table 4). Although there were a greater number of hemorrhagic strokes with
moderate/severe outcomes, we did not see any significant decrease or increase in risk of
hemorrhagic stroke with moderate/severe outcome for those in higher versus the lowest
healthy lifestyle index categories.

Am J Med. Author manuscript; available in PMC 2017 July 01.

 Rist et al. Page 6

When we excluded BMI from our healthy lifestyle index, the association between the
Author Manuscript

modified healthy lifestyle index categories and functional outcomes from ischemic stroke
followed a similar pattern to the original healthy lifestyle index, but the effects were
attenuated. We did not observe a statistically significant association between the highest
modified healthy lifestyle index category and mild functional outcomes from hemorrhagic
stroke. This suggests that the inclusion of body mass index in our original healthy lifestyle
index may have driven the statistically significant association between the highest category
of the healthy lifestyle index and increased risk of hemorrhagic stroke with mild functional
outcome.

Discussion
Previous research in the WHS has shown that a healthy lifestyle is associated with a reduced
risk of stroke.3 The present study expands upon these findings by demonstrating that higher
Author Manuscript

scores on the healthy lifestyle index are associated with reductions in the risk of total and
ischemic stroke with mild, moderate, and severe functional outcomes. Even modest scores
on the healthy lifestyle index (5 to 8 points) were associated with a decreased risk of all
functional outcomes, but this decrease was only statistically significant for moderate and
severe outcomes. This suggests that maintaining a moderately healthy lifestyle may decrease
the risk of a disabling ischemic stroke event among women. However, our healthy lifestyle
index had little impact on the risk of moderate/severe hemorrhagic strokes.

Although research on composite healthy lifestyle indices and functional outcomes from
stroke is limited, some research has examined the association between the individual
components of the healthy lifestyle index and functional outcomes from stroke. Although
smoking has been associated with the risk of both ischemic33 and hemorrhagic stroke34
Author Manuscript

among women, information on stroke outcomes is mixed.7–12 Some studies have shown
beneficial effects of physical activity on functional outcomes from stroke or stroke
severity13–15 while one study showed no significant effect.16 Previous studies on alcohol
consumption have found either modest or no association between moderate alcohol
consumption and stroke outcomes.17–19 Some studies have suggested that stroke patients
who are overweight or obese may have better outcomes.20–22 However, limited evidence
exists on the effect of BMI on stroke outcomes in initially healthy populations. Information
on diet and stroke outcomes in initially healthy populations is limited. Patients with acute
stroke and poor nutritional status have a higher risk of death than those without poor
nutritional status.23 A prospective cohort study among women did not find significant
associations between quintiles of the Mediterranean Diet score and incidence of ischemic,
hemorrhagic, fatal, or nonfatal stroke.35
Author Manuscript

Despite the mixed or weak evidence that these individual components independently
influence stroke outcomes, our study shows that a combination of lifestyle factors may help
reduce the risk of mild, moderate, and severe ischemic stroke. The majority of the women in
this cohort had scores on the health index from 5 to 8 (26.3% of the cohort) or 9 to 12
(40.2% of the cohort). While these categories do not represent the healthiest lifestyle
according to our index, we still observed reductions in the risk of mild, moderate, and severe
ischemic stroke for these women compared to women in the lowest category. This suggests

Am J Med. Author manuscript; available in PMC 2017 July 01.

 Rist et al. Page 7

that adopting even a modestly healthy lifestyle may reduce the risk of risk of mild, moderate,
Author Manuscript

and severe ischemic stroke. For example, someone could achieve 5 to 8 points on the healthy
lifestyle index by being a past smoker, having a BMI between 25.0–29.9 kg/m2, exercising
once per week, never consuming alcohol, and being in the second lowest quintile of the
dietary score distribution.

We observed a different pattern between the healthy lifestyle index and risk of hemorrhagic
stroke with mild or moderate/severe functional outcomes. Those in the highest category of
the healthy lifestyle index had an increased risk of hemorrhagic stroke with a mild functional
outcome. However, the small number of hemorrhagic strokes with mild outcomes (N=4)
makes it difficult to definitively conclude that healthier lifestyles are associated with an
increased risk of hemorrhagic stroke with mild outcome. In addition, this increase was no
longer significant when BMI was removed from the healthy lifestyle index. This suggests
that some of the association between the healthy lifestyle index and risk of mild hemorrhagic
Author Manuscript

stroke is driven by BMI. Evidence on the association between BMI and hemorrhagic stroke
risk is unclear. Some studies have shown an increased risk of hemorrhagic stroke among
people who are lean36–39 while others found no association40–42 or a positive association
with increasing BMI.43 Information on the association between overweight or obesity and
hemorrhagic stroke outcomes is limited. A study among men observed higher proportion of
mild strokes with BMI ≥30 kg/m2 while the proportion of fatal strokes was greater in men
with BMI <23 kg/m2.43

Strengths of the present study include available information on many lifestyle factors in a
large number of participants, the higher number of outcome events which allowed us to
explore outcomes from ischemic and hemorrhagic stroke separately, and high interobserver
agreement on major stroke subtype classification. We used the mRS to measure of functional
Author Manuscript

outcome from stroke, which is widely accepted for use in clinical trials, accounts for
prestroke disability44, has strong test-retest reliability, interrater reliability and validity45,
and does not exhibit a “ceiling effect”.46

Some limitations should be noted. First, we were unable to update the healthy lifestyle index
over time because not all factors included in the index were assessed at multiple time points.
Second, we did not consider weighting the components of the healthy lifestyle index. Third,
there is the potential for non-differential misclassification of self-reported risk factors which
would bias our results towards the null. Fourth, all participants in WHS were female health
professionals and were primarily white which may limit our generalizability.

Data from this large prospective cohort study of women observed that the highest versus
lowest scores on the healthy lifestyle index were associated with reductions in the risk of
Author Manuscript

total and ischemic strokes with mild, moderate, and severe functional outcomes. The
observation that even modest healthy lifestyle index scores reduced risks of total and
ischemic stroke with moderate and severe functional outcomes suggests that even modest
lifestyle changes may translate into reductions in risk of a disabling stroke event.

Am J Med. Author manuscript; available in PMC 2017 July 01.

 Rist et al. Page 8

Online Table 1. Multivariable-adjusted relative risks of functional outcomes by modified

Author Manuscript

Rankin scale (mRS) score after ISCHEMIC stroke according to modified healthy lifestyle
index* (HI) categories in the Women’s Health Study (N=37,474).

Online Table 2. Multivariable-adjusted relative risks of functional outcomes by modified

Rankin scale (mRS) score after HEMORRHAGIC stroke according to modified healthy
lifestyle index* (HI) categories in the Women’s Health Study (N=36,923).

Acknowledgments
Acknowledgements/Study Funding: The WHS is supported by grants from the National Institutes of Health
(CA047988, HL043851, HL080467, HL099355, UM1 CA182913). The funder played no role in the design and
conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or
approval of the manuscript; and decision to submit the manuscript for publication.
Author Manuscript

References
1. Mackay, J.; Mensah, GA. Atlas of heart disease and stroke. World Health Organization; 2004.
2. Mathers, C.; Fat, DM.; Boerma, JT. The Global Burden of Disease: 2004 Update. World Health
Organization; 2008.
3. Kurth T, Moore SC, Gaziano JM, et al. Healthy lifestyle and the risk of stroke in women. Arch
Intern Med. 2006; 166(13):1403–1409. [PubMed: 16832006]
4. Chiuve SE, Rexrode KM, Spiegelman D, Logroscino G, Manson JE, Rimm EB. Primary prevention
of stroke by healthy lifestyle. Circulation. 2008; 118(9):947–954. [PubMed: 18697819]
5. Larsson SC, Akesson A, Wolk A. Healthy diet and lifestyle and risk of stroke in a prospective cohort
of women. Neurology. 2014; 83(19):1699–1704. [PubMed: 25298305]
6. Zhang Y, Tuomilehto J, Jousilahti P, Wang Y, Antikainen R, Hu G. Lifestyle factors on the risks of
ischemic and hemorrhagic stroke. Arch Intern Med. 2011; 171(20):1811–1818. [PubMed:
21911621]
7. Ovbiagele B, Saver JL. The smoking-thrombolysis paradox and acute ischemic stroke. Neurology.
Author Manuscript

2005; 65(2):293–295. [PubMed: 16043802]

8. Bang OY, Park HY, Lee PH, Kim GM, Chung CS, Lee KH. Improved outcome after atherosclerotic
stroke in male smoker. J Neurol Sci. 2007; 260(1–2):43–48. [PubMed: 17477936]
9. Weng W-C, Huang W-Y, Chien Y-Y, et al. The impact of smoking on the severity of acute ischemic
stroke. J Neurol Sci. 2011; 308(1–2):94–97. [PubMed: 21665225]
10. Kumagai N, Origasa H, Nagao T, Takekawa H, Okuhara Y, Yamaguchi T. Prognostic significance
of smoking in patients with acute ischemic stroke within 3 months of onset. J Stroke Cerebrovasc
Dis. 2012; 22(6):792–798. [PubMed: 22633681]
11. Kumagai N, Okuhara Y, Iiyama T, et al. Effects of smoking on outcomes after acute
atherothrombotic stroke in Japanese men. J Neurol Sci. 2013; 335(1–2):164–168. [PubMed:
24112970]
12. Ovbiagele B, Weir CJ, Saver JL, Muir KW, Lees KR. IMAGES Investigators. Effect of smoking
status on outcome after acute ischemic stroke. Cerebrovasc Dis. 2006; 21(4):260–265. [PubMed:
16446540]
Author Manuscript

13. Deplanque D, Masse I, Lefebvre C, Libersa C, Leys D, Bordet R. Prior TIA, lipid-lowering drug
use, and physical activity decrease ischemic stroke severity. Neurology. 2006; 67(8):1403–1410.
[PubMed: 17060566]
14. Krarup L-H, Truelsen T, Gluud C, et al. Prestroke physical activity is associated with severity and
long-term outcome from first-ever stroke. Neurology. 2008; 71(17):1313–1318. [PubMed:
18936423]
15. Stroud N, Mazwi TML, Case LD, et al. Prestroke physical activity and early functional status after
stroke. J Neurol Neurosurg Psychiatry. 2009; 80(9):1019–1022. [PubMed: 19602474]

Am J Med. Author manuscript; available in PMC 2017 July 01.

 Rist et al. Page 9

16. Rist PM, Lee I-M, Kase CS, Gaziano JM, Kurth T. Physical activity and functional outcomes from
cerebral vascular events in men. Stroke. 2011; 42(12):3352–3356. [PubMed: 21940956]
Author Manuscript

17. Rist PM, Berger K, Buring JE, Kase CS, Gaziano JM, Kurth T. Alcohol consumption and
functional outcome after stroke in men. Stroke. 2009; 41(1):141–146. [PubMed: 19910548]
18. Jørgensen HS, Reith J, Nakayama H, Kammersgaard LP, Raaschou HO, Olsen TS. What
determines good recovery in patients with the most severe strokes? The Copenhagen Stroke Study.
Stroke. 1999; 30(10):2008–2012. [PubMed: 10512899]
19. Sturm JW, Donnan GA, Dewey HM, Macdonell RAL, Gilligan AK, Thrift AG. Determinants of
handicap after stroke: the North East Melbourne Stroke Incidence Study (NEMESIS). Stroke.
2004; 35(3):715–720. [PubMed: 14963272]
20. Doehner W, Schenkel J, Anker SD, Springer J, Audebert HJ. Overweight and obesity are
associated with improved survival, functional outcome, and stroke recurrence after acute stroke or
transient ischaemic attack: observations from the TEMPiS trial. Eur Heart J. 2013; 34(4):268–277.
[PubMed: 23076781]
21. Saini M, Saqqur M, Shuaib A, Collaborators V. Body mass index and acute ischemic stroke
outcomes. Int J Stroke. 2014; 9(5):618–623.
Author Manuscript

22. Zhao L, Du W, Zhao X, et al. Favorable functional recovery in overweight ischemic stroke
survivors: findings from the China National Stroke Registry. J Stroke Cerebrovasc Dis. 2014;
23(3):e201–e206. [PubMed: 24231138]
23. FOOD Trial Collaboration. Poor nutritional status on admission predicts poor outcomes after
stroke: observational data from the FOOD trial. Stroke. 2003; 34(6):1450–1456. [PubMed:
12750536]
24. Lee I-M, Cook NR, Gaziano JM, et al. Vitamin E in the primary prevention of cardiovascular
disease and cancer: the Women’s Health Study: a randomized controlled trial. JAMA. 2005;
294(1):56–65. [PubMed: 15998891]
25. Ridker PM, Cook NR, Lee I-M, et al. A randomized trial of low-dose aspirin in the primary
prevention of cardiovascular disease in women. N Engl J Med. 2005; 352(13):1293–1304.
[PubMed: 15753114]
26. Reynolds K, Lewis B, Nolen JDL, et al. Alcohol consumption and risk of stroke: a meta-analysis.
JAMA. 2003; 289(5):579–588. [PubMed: 12578491]
Author Manuscript

27. Willett WC, Sampson L, Browne ML, et al. The use of a self-administered questionnaire to assess
diet four years in the past. Am J Epidemiol. 1988; 127(1):188–199. [PubMed: 3337073]
28. Atiya M, Kurth T, Berger K, Buring JE, Kase CS. Women’s Health Study. Interobserver agreement
in the classification of stroke in the Women’s Health Study. Stroke. 2003; 34(2):565–567.
[PubMed: 12574576]
29. Rist PM, Buring JE, Kase CS, Schürks M, Kurth T. Migraine and functional outcome from
ischemic cerebral events in women. Circulation. 2010; 122(24):2551–2557. [PubMed: 21126968]
30. Rist PM, Buring JE, Kase CS, Ridker PM, Kurth T. Biomarkers and functional outcomes from
ischaemic cerebral events in women: a prospective cohort study. Eur J Neurol. 2012; 20(2):375–
381. [PubMed: 23034002]
31. Rist PM, Buring JE, Kase CS, Kurth T. Effect of low-dose aspirin on functional outcome from
cerebral vascular events in women. Stroke. 2013; 44(2):432–436. [PubMed: 23306328]
32. Rist PM, Glymour MM, Orav EJ, et al. Non-steroidal anti-inflammatory drug use and functional
outcome from ischemic cerebral events among women. Eur J Intern Med. 2014; 25(3):255–258.
[PubMed: 24525385]
Author Manuscript

33. Kawachi I, Colditz GA, Stampfer MJ, et al. Smoking cessation and decreased risk of stroke in
women. JAMA. 1993; 269(2):232–236. [PubMed: 8417241]
34. Kurth T, Kase CS, Berger K, Gaziano JM, Cook NR, Buring JE. Smoking and risk of hemorrhagic
stroke in women. Stroke. 2003; 34(12):2792–2795. [PubMed: 14615625]
35. Fung TT, Rexrode KM, Mantzoros CS, Manson JE, Willett WC, Hu FB. Mediterranean diet and
incidence of and mortality from coronary heart disease and stroke in women. Circulation. 2009;
119(8):1093–1100. [PubMed: 19221219]
36. Rexrode KM, Hennekens CH, Willett WC, et al. A prospective study of body mass index, weight
change, and risk of stroke in women. JAMA. 1997; 277(19):1539–1545. [PubMed: 9153368]

Am J Med. Author manuscript; available in PMC 2017 July 01.

 Rist et al. Page 10

37. Song Y-M, Sung J, Davey Smith G, Ebrahim S. Body mass index and ischemic and hemorrhagic
stroke: a prospective study in Korean men. Stroke. 2004; 35(4):831–836. [PubMed: 15001798]
Author Manuscript

38. Thrift AG, McNeil JJ, Forbes A, Donnan GA. for the Melbourne Risk Factor Study (MERFS)
Group. Risk Factors for Cerebral Hemorrhage in the Era of Well-Controlled Hypertension. Stroke.
1996; 27(11):2020–2025. [PubMed: 8898809]
39. 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(15):1992–1998. [PubMed: 15837954]
40. Jood K, Jern C, Wilhelmsen L, Rosengren A. Body mass index in mid-life is associated with a first
stroke in men: a prospective population study over 28 years. Stroke. 2004; 35(12):2764–2769.
[PubMed: 15514172]
41. Abbott RD, Behrens GR, Sharp DS, et al. Body mass index and thromboembolic stroke in
nonsmoking men in older middle age. The Honolulu Heart Program. Stroke. 1994; 25(12):2370–
2376. [PubMed: 7974575]
42. Rodriguez BL, Agostino RD’, Abbott RD, et al. Risk of hospitalized stroke in men enrolled in the
Honolulu Heart Program and the Framingham Study: A comparison of incidence and risk factor
effects. Stroke. 2002; 33(1):230–236. [PubMed: 11779915]
Author Manuscript

43. Kurth T, Gaziano JM, Berger K, et al. Body mass index and the risk of stroke in men. Arch Intern
Med. 2002; 162(22):2557–2562. [PubMed: 12456227]
44. Van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J. Interobserver agreement for
the assessment of handicap in stroke patients. Stroke. 1988; 19(5):604–607. [PubMed: 3363593]
45. Banks JL, Marotta CA. Outcomes validity and reliability of the modified Rankin scale:
implications for stroke clinical trials: a literature review and synthesis. Stroke. 2007; 38(3):1091–
1096. [PubMed: 17272767]
46. Weimar C, Kurth T, Kraywinkel K, et al. Assessment of functioning and disability after ischemic
stroke. Stroke. 2002; 33(8):2053–2059. [PubMed: 12154262]
Author Manuscript
Author Manuscript

Am J Med. Author manuscript; available in PMC 2017 July 01.

 Rist et al. Page 11

Clinical Significance
Author Manuscript

• Highest versus lowest scores on the healthy lifestyle index were associated
with reductions in risk of total and ischemic strokes with mild, moderate,
and severe functional outcomes among women.

• The evidence that even a modest healthy lifestyle index scores reduced risks
of total and ischemic stroke with moderate and severe functional outcomes
suggests that modest lifestyle changes may translate to reductions in risk of
a disabling stroke event.
Author Manuscript
Author Manuscript
Author Manuscript

Am J Med. Author manuscript; available in PMC 2017 July 01.

 Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Table 1

Baseline characteristics of participants by healthy lifestyle index score*.

Rist et al.

Healthy Lifestyle Index (N= 37624)

Characteristic 0–4 (n = 1615; 4.3%) 5–8 (n = 9890; 26.3%) 9–12 (n = 15108; 40.15%) 13–16 (n = 9265; 24.6%) 17–20 (n = 1756; 4.7%)

Smoking, %
Current >=15 cig/d 54.3 13.9 4.6 1.0 0
Current < 15 cig/d 12.0 7.5 4.3 2.2 0.2
Past >=20 pack-yrs 13.4 13.1 11.5 9.2 3.9
Past < 20 pack-yrs 11.7 20.8 25.0 30.3 31.4
Never 8.6 44.8 54.6 57.3 64.5

Physical activity, %
Rarely or never 85.8 65.5 35.9 11.4 0
< 1 time/wk 12.1 23.9 23.7 14.3 1.7
1 time/wk 1.6 5.9 12.1 12.5 5.2
2–3 times/wk 0.5 4.2 21.4 39.7 43.1
>=4 times/week 0 0.5 6.9 22.1 50.1

Body mass index, %

>=35.0 31.5 12.1 3.2 0.4 0
30.0–34.9 25.4 22.0 10.9 3.6 0.3
25.0–29.9 24.2 36.5 34.5 24.6 6.7

Am J Med. Author manuscript; available in PMC 2017 July 01.

22.0–24.9 13.8 20.7 32.7 38.5 35.7
<22.0 5.2 8.8 18.7 33.0 57.4

Alcohol consumption, %
Never 86.2 70.7 44.8 17.6 0
< 1 drink/wk 12.0 16.9 17.7 12.5 1.8
1–3 drinks/wk 0.9 5.6 16.4 24.9 21.6
2 to <10.5 drinks/wk 0.2 3.4 15.8 39.1 72.7
>=10.5 drinks/wk 0.7 3.5 5.3 5.9 3.9
Page 12
 Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Healthy Lifestyle Index (N= 37624)

Characteristic 0–4 (n = 1615; 4.3%) 5–8 (n = 9890; 26.3%) 9–12 (n = 15108; 40.15%) 13–16 (n = 9265; 24.6%) 17–20 (n = 1756; 4.7%)
Rist et al.

Diet score quintile, %

1st 65.3 37.1 15.8 4.5 0
2nd 25.8 33.6 26.3 14.2 1.8
3rd 6.8 15.2 20.5 18.0 9.0
4th 1.9 9.6 20.5 27.5 27.8
5th 0.3 4.5 16.9 35.8 61.4

Age, mean (SE), y 53.5 (6.2) 54.4 (6.9) 54.9 (7.2) 54.6 (7.1) 54.5 (7.1)

History of hypertension 37.1 32.4 25.3 19.3 14.5

Antihypertensive medication use, % 18.0 17.7 13.4 10.4 6.9

History of diabetes mellitus, % 5.6 4.0 2.2 1.2 0.9

History of elevated cholesterol 34.2 31.5 30.5 25.6 22.7

Postmenopausal hormone use

Never 55.2 53.1 49.2 47.3 44.9

Past 13.3 10.6 8.8 7.5 6.0
Current 31.3 36.1 41.8 45.1 49.0
Ever used oral contraceptives, % 71.2 67.9 68.0 72.1 75.2

Am J Med. Author manuscript; available in PMC 2017 July 01.

Family history of MI before age 60 years, % 14.3 12.5 11.3 10.8 10.9

Ethnicity, %
White 94.4 93.5 94.0 95.5 96.9
Black 3.4 2.7 2.3 1.4 0.9
Other 2.2 3.8 3.8 3.1 2.2

Annual household income, %

<$50,000 61.9 53.3 43.7 33.9 26.9
$50,000–$100,000 30.3 35.3 39.6 41.9 39.5
Page 13
 Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Healthy Lifestyle Index (N= 37624)

Characteristic 0–4 (n = 1615; 4.3%) 5–8 (n = 9890; 26.3%) 9–12 (n = 15108; 40.15%) 13–16 (n = 9265; 24.6%) 17–20 (n = 1756; 4.7%)
Rist et al.

>=$100,000 3.2 6.4 10.6 18.3 28.3

Highest level of education, %

Less than bachelor’s degree 77.1 65.9 55.8 45.2 37.9
Bachelor’s degree 13.7 19.1 23.4 26.3 28.2
Master’s degree or doctorate 7.4 13.2 19.2 26.8 32.3

Marital status, %
Single 7.0 6.3 5.8 4.8 4.4
Married 62.0 68.8 72.3 73.8 74.9
Other 31.0 24.9 21.9 21.4 20.7

Geographic location, %
Northeast 20.5 19.3 19.0 20.2 17.9
Southeast 24.2 23.8 23.7 21.5 20.8
Midwest 41.1 39.9 36.0 31.5 27.0
West 14.2 16.7 21.0 26.6 34.0

Migraine status, %
No history of migraine 79.4 80.0 81.7 83.1 84.1
Past history of migraine 6.6 5.7 5.3 5.2 5.6
Migraine without aura 8.4 8.5 7.9 6.8 5.0

Am J Med. Author manuscript; available in PMC 2017 July 01.

Migraine with aura 5.4 5.6 5.0 4.8 5.2

Note: Numbers may not add to 100% due to missing data.

*
The healthy lifestyle index includes information on smoking status, physical activity, BMI, alcohol consumption, and diet.
Page 14
 Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Table 2

Age- and multivariable-adjusted relative risks of functional outcomes by modified Rankin scale (mRS) score after TOTAL stroke according to Healthy
Lifestyle Index* categories in the Women’s Health Study (N=37,634).
Rist et al.

No Stroke (n=36,767) mRS 0–1 (n=360) mRS 2–3 (n=303) mRS 4–6 (n=204)

n % n % RR (95% CI) n % RR (95% CI) n % RR (95% CI)

Age-adjusted

HI 0–4 1556 4.2 22 6.1 1.00 23 7.6 1.00 14 6.9 1.00

HI 5–8 9644 26.2 101 28.6 0.69 (0.43, 1.09) 90 29.7 0.56 (0.35, 0.88) 53 26.0 0.52 (0.29, 0.94)

HI 9–12 14759 40.1 150 41.7 0.63 (0.40, 0.98) 115 38.0 0.44 (0.28, 0.69) 84 41.2 0.50 (0.28, 0.88)

HI 13–16 9086 24.7 74 20.6 0.51 (0.32, 0.83) 60 19.8 0.38 (0.24, 0.62) 45 22.1 0.45 (0.25, 0.82)

HI 17–20 1722 4.7 11 3.1 0.41 (0.20, 0.84) 15 5.0 0.50 (0.26, 0.98) 8 3.9 0.43 (0.18, 1.02)

Multivariable-adjusted**

HI 0–4 1556 4.2 22 6.1 1.00 23 7.6 1.00 14 6.9 1.00

HI 5–8 9644 26.2 101 28.6 0.71 (0.45, 1.14) 90 29.7 0.57 (0.36, 0.92) 53 26.0 0.54 (0.30, 0.99)

HI 9–12 14759 40.1 150 41.7 0.66 (0.41, 1.04) 115 38.0 0.47 (0.29, 0.74) 84 41.2 0.54 (0.30, 0.97)

HI 13–16 9086 24.7 74 20.6 0.55 (0.33, 0.89) 60 19.8 0.41 (0.25, 0.68) 45 22.1 0.51 (0.27, 0.95)

HI 17–20 1722 4.7 11 3.1 0.43 (0.20, 0.90) 15 5.0 0.53 (0.27, 1.06) 8 3.9 0.48 (0.20, 1.18)

Multivariable-adjusted (with intermediates)***

HI 0–4 1556 4.2 22 6.1 1.00 23 7.6 1.00 14 6.9 1.00

HI 5–8 9644 26.2 101 28.6 0.74 (0.46, 1.18) 90 29.7 0.59 (0.37, 0.94) 53 26.0 0.56 (0.31, 1.02)

Am J Med. Author manuscript; available in PMC 2017 July 01.

HI 9–12 14759 40.1 150 41.7 0.72 (0.45, 1.14) 115 38.0 0.51 (0.32, 0.81) 84 41.2 0.59 (0.33, 1.06)

HI 13–16 9086 24.7 74 20.6 0.62 (0.38, 1.01) 60 19.8 0.47 (0.28, 0.78) 45 22.1 0.58 (0.31, 1.08)

HI 17–20 1722 4.7 11 3.1 0.49 (0.23, 1.04) 15 5.0 0.64 (0.33, 1.27) 8 3.9 0.56 (0.23, 1.37)

*
Healthy lifestyle index includes smoking, physical activity, body mass index, alcohol consumption, and diet.
**
Adjusted for age (squared term), postmenopausal hormone use (never, past, or current), ever use of oral contraceptives, family history of myocardial infarction before age 60 years, migraine status (no
history, past history, active migraine with aura, or active migraine without aura), annual household income (<$50,000, $50,000 to <$100,000, or ≥$100,000), highest level of education (less than a bachelor’s
degree, bachelor’s degree, or master’s degree or doctorate), location of residence (Northeast, Southeast, Midwest, or West), marital status (single, married, or other), and ethnicity (white, black, or other).
***
Adjusted for all covariates above in addition to history of diabetes mellitus, history of hypertension (defined as systolic blood pressure of ≥140 mm Hg, diastolic blood pressure of ≥90 mm Hg, or self-
reported physician-diagnosed hypertension), antihypertensive treatment, and history of elevated cholesterol of ≥240 mg/dL (6.21 mmol/L).
Page 15
 Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Note: Less than 100 women were missing information on diabetes, history of hypertension, antihypertensive treatment, history of high cholesterol, migraine status, location of residence, postmenopausal
hormone use, history of oral contraceptive use, and were assigned to the most frequent category. More than 100 women were missing information on annual household income, family history of myocardial
infarction, education, and marital status and were assigned to a separate “missing” category. People missing information on ethnicity were included in the “other” category.

HI = healthy lifestyle index score; mRS = modified Rankin scale; RR = relative risk; CI = confidence interval.
Rist et al.

Am J Med. Author manuscript; available in PMC 2017 July 01.

Page 16
 Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Table 3

Age- and multivariable-adjusted relative risks of functional outcomes by modified Rankin scale (mRS) score after ISCHEMIC stroke according to
Healthy Lifestyle Index* (HI) categories in the Women’s Health Study (N=37,474).
Rist et al.

No Stroke (n=36,767) mRS 0–1 (n=329) mRS 2–3 (n=269) mRS 4–6 (n=109)

n % n % RR (95% CI) n % RR (95% CI) n % RR (95% CI)

Age-adjusted

HI 0–4 1556 4.2 22 6.7 1.00 20 7.4 1.00 8 7.3 1.00

HI 5–8 9644 26.2 96 29.2 0.64 (0.40, 1.02) 77 28.6 0.54 (0.33, 0.89) 32 29.4 0.51 (0.23, 1.11)

HI 9–12 14759 40.1 139 42.3 0.58 (0.37, 0.91) 109 40.5 0.47 (0.29, 0.76) 43 39.5 0.40 (0.19, 0.87)

HI 13–16 9086 24.7 66 20.1 0.46 (0.28, 0.74) 49 18.2 0.36 (0.21, 0.60) 22 20.2 0.35 (0.16, 0.79)

HI 17–20 1722 4.7 6 1.8 0.22 (0.09, 0.55) 14 5.2 0.54 (0.27, 1.08) 4 3.7 0.34 (0.10, 1.15)

Multivariable-adjusted**

HI 0–4 1556 4.2 22 6.7 1.00 20 7.4 1.00 8 7.3 1.00

HI 5–8 9644 26.2 96 29.2 0.65 (0.41, 1.04) 77 28.6 0.56 (0.34, 0.92) 32 29.4 0.56 (0.25, 1.23)

HI 9–12 14759 40.1 139 42.3 0.59 (0.37, 0.94) 109 40.5 0.50 (0.31, 0.82) 43 39.5 0.47 (0.21, 1.01)

HI 13–16 9086 24.7 66 20.1 0.47 (0.29, 0.78) 49 18.2 0.38 (0.22, 0.65) 22 20.2 0.42 (0.18, 0.98)

HI 17–20 1722 4.7 6 1.8 0.23 (0.09, 0.56) 14 5.2 0.57 (0.28, 1.16) 4 3.7 0.42 (0.12, 1.45)

Multivariable-adjusted (with intermediates)***

HI 0–4 1556 4.2 22 6.7 1.00 20 7.4 1.00 8 7.3 1.00

HI 5–8 9644 26.2 96 29.2 0.67 (0.42, 1.08) 77 28.6 0.57 (0.35, 0.95) 32 29.4 0.58 (0.26, 1.27)

Am J Med. Author manuscript; available in PMC 2017 July 01.

HI 9–12 14759 40.1 139 42.3 0.65 (0.41, 1.03) 109 40.5 0.55 (0.33, 0.89) 43 39.5 0.51 (0.23, 1.11)

HI 13–16 9086 24.7 66 20.1 0.54 (0.32, 0.89) 49 18.2 0.44 (0.25, 0.75) 22 20.2 0.50 (0.21, 1.16)

HI 17–20 1722 4.7 6 1.8 0.26 (0.10, 0.66) 14 5.2 0.69 (0.34, 1.40) 4 3.7 0.50 (0.15, 1.74)

*
Healthy lifestyle index includes smoking, physical activity, body mass index, alcohol consumption, and diet.
**
Adjusted for age (squared term), postmenopausal hormone use (never, past, or current), ever use of oral contraceptives, family history of myocardial infarction before age 60 years, migraine status (no
history, past history, active migraine with aura, or active migraine without aura), annual household income (<$50,000, $50,000 to <$100,000, or ≥$100,000), highest level of education (less than a bachelor’s
degree, bachelor’s degree, or master’s degree or doctorate), location of residence (Northeast, Southeast, Midwest, or West), marital status (single, married, or other), and ethnicity (white, black, or other).
***
Adjusted for all covariates above in addition to history of diabetes mellitus, history of hypertension (defined as systolic blood pressure of ≥140 mm Hg, diastolic blood pressure of ≥90 mm Hg, or self-
reported physician-diagnosed hypertension), antihypertensive treatment, and history of elevated cholesterol of ≥240 mg/dL (6.21 mmol/L).
Page 17
 Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Note: Less than 100 women were missing information on diabetes, history of hypertension, antihypertensive treatment, history of high cholesterol, migraine status, location of residence, postmenopausal
hormone use, history of oral contraceptive use, and were assigned to the most frequent category. More than 100 women were missing information on annual household income, family history of myocardial
infarction, education, and marital status and were assigned to a separate “missing” category. People missing information on ethnicity were included in the “other” category.

HI = healthy lifestyle index score; mRS = modified Rankin scale; RR = relative risk; CI = confidence interval.
Rist et al.

Am J Med. Author manuscript; available in PMC 2017 July 01.

Page 18
 Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Table 4

Age- and multivariable-adjusted* relative risks of functional outcomes by modified Rankin scale (mRS) score after HEMORRHAGIC stroke according
to Healthy Lifestyle Index categories in the Women’s Health Study (N=36,923).
Rist et al.

No Stroke (n=36,767) mRS 0–1 (n=29) mRS 2–6 (n=127)

n % n % RR (95% CI) n % RR (95% CI)

Age-adjusted

HI 0–8 11200 30.5 7 24.1 1.00 42 33.1 1.00

HI 9–12 14759 40.1 10 34.5 1.08 (0.41, 2.85) 47 37.0 0.81 (0.53, 1.23)

HI 13–16 9086 24.7 8 27.6 1.41 (0.51, 3.89) 33 26.0 0.94 (0.60, 1.49)

HI 17–20 1722 4.7 4 13.8 3.71 (1.09, 12.69) 5 3.9 0.76 (0.30, 1.93)

Multivariable-adjusted*

HI 0–8 11200 30.5 7 24.1 1.00 42 33.1 1.00

HI 9–12 14759 40.1 10 34.5 1.17 (0.44, 3.08) 47 37.0 0.83 (0.55,1.27)

HI 13–16 9086 24.7 8 27.6 1.57 (0.56, 4.41) 33 26.0 1.00 (0.63, 1.61)

HI 17–20 1722 4.7 4 13.8 4.23 (1.20, 14.99) 5 3.9 0.81 (0.32, 2.09)

Multivariable-adjusted (with intermediates)**

HI 0–8 11200 30.5 7 24.1 1.00 42 33.1 1.00

HI 9–12 14759 40.1 10 34.5 1.15 (0.43, 3.04) 47 37.0 0.86 (0.57, 1.32)

HI 13–16 9086 24.7 8 27.6 1.55 (0.55, 4.39) 33 26.0 1.06 (0.66, 1.71)

HI 17–20 1722 4.7 4 13.8 4.12 (1.15, 14.72) 5 3.9 0.88 (0.34, 2.27)

Am J Med. Author manuscript; available in PMC 2017 July 01.

*
Healthy lifestyle index includes smoking, physical activity, body mass index, alcohol consumption, and diet.
**
Adjusted for age (squared term), postmenopausal hormone use (never, past, or current), ever use of oral contraceptives, family history of myocardial infarction before age 60 years, migraine status (no
history, past history, active migraine with aura, or active migraine without aura), annual household income (<$50,000, $50,000 to <$100,000, or ≥$100,000), highest level of education (less than a bachelor’s
degree, bachelor’s degree, or master’s degree or doctorate), location of residence (Northeast, Southeast, Midwest, or West), marital status (single, married, or other), and ethnicity (white, black, or other).
***
Adjusted for all covariates above in addition to history of diabetes mellitus, history of hypertension (defined as systolic blood pressure of ≥140 mm Hg, diastolic blood pressure of ≥90 mm Hg, or self-
reported physician-diagnosed hypertension), antihypertensive treatment, and history of elevated cholesterol of ≥240 mg/dL (6.21 mmol/L).

Note: Less than 100 women were missing information on diabetes, history of hypertension, antihypertensive treatment, history of high cholesterol, migraine status, location of residence, postmenopausal
hormone use, history of oral contraceptive use, and were assigned to the most frequent category. More than 100 women were missing information on annual household income, family history of myocardial
infarction, education, and marital status and were assigned to a separate “missing” category. People missing information on ethnicity were included in the “other” category.

HI = healthy lifestyle index score; mRS = modified Rankin scale; RR = relative risk; CI = confidence interval.
Page 19