Sex differences in the risk of recurrent ischemic stroke after ischemic stroke and transient ischemic attack

Abstract

Background

Sex differences in stroke outcomes have been noted, but whether this extends to stroke recurrence is unclear. We examined sex differences in recurrent stroke using data from the Platelet-Oriented Inhibition in New TIA and Minor Ischemic Stroke (POINT) trial.

Patients and methods

We assessed the risk of recurrent stroke in women compared to men using data from the POINT trial. Adults >18 years old were randomized within 12 hours of onset of minor ischemic stroke or transient ischemic attack (TIA), and followed for up to 90 days for ischemic stroke, our primary outcome. We used Cox proportional hazards model adjusted for demographics and stroke risk factors to evaluate the association between sex and stroke recurrence. We used interaction term testing and prespecified subgroup analyses to determine if the association between sex and recurrent stroke differed by age (<60 versus >60 years old), locale (US versus non-US), and index event type (stroke versus TIA). Last, we evaluated whether sex modified the effect of common stroke risk factors on stroke recurrence.

Results

Of 4,881 POINT trial participants with minor stroke or high-risk TIA, 2,195 (45%) were women. During the 90-day follow-up period, 267 ischemic strokes occurred; 121 were in women and 146 in men. The cumulative risk of recurrent ischemic stroke was not significantly different among women (5.76%; 95% CI, 4.84%–6.85%) compared to men (5.67%; 95% CI, 4.83%–6.63%). Women were not at a different risk of recurrent ischemic stroke compared to men (hazard ratio [HR], 1.02; 95% CI, 0.80–1.30) in unadjusted models or after adjusting for covariates. However, there was a significant interaction of age with sex (P=0.04). Among patients <60 years old, there was a non-significantly lower risk of recurrent stroke in women compared to men (HR 0.66; 95% CI 0.42–1.05). Last, sex did not modify the association between common stroke risk factors and recurrent stroke risk.

Discussion and Conclusion

Among patients with minor stroke or TIA, the risk of recurrent ischemic stroke and the impact of common stroke risk factors did not differ between men and women.

Keywords

stroke/cerebrovascular disease risk factors epidemiology sex/gender

Introduction

One in four strokes in the US is a recurrent stroke, and approximately 13% of patients with minor stroke/transient ischemic attack (TIA) face recurrent stroke, heart attack, or cardiovascular death within 5 years.^1,2 Sex differences in stroke outcomes have been noted, including both post-stroke mortality and patient-reported outcomes.^3,4 Further, several sex-specific stroke risk factors have been identified, such as migraine and oral contraceptive use for women and substance use for men.^5–9 Some conventional stroke risk factors, including hypertension, diabetes, and metabolic syndrome, may also have a stronger association with incident stroke risk in women than in men.^6,9,10 Whether these differences impact stroke recurrence risk has not been definitively assessed. Prior studies, which included a large proportion of stroke mimics¹¹ or used non-adjudicated administrative claims data,^12,13 did not find clear evidence of a sex difference in stroke recurrence.

Several challenges exist in studying sex differences and recurrent stroke risk using observational data. For example, women are less likely to be diagnosed with stroke despite presenting with similar symptoms as men.¹¹ Additionally, recurrent stroke is often early,^2,14 and early recurrent strokes can be misclassified as worsening stroke symptoms rather than new events. To circumvent these challenges, we studied sex differences in stroke recurrence using adjudicated clinical trial data from the Platelet-Oriented Inhibition in New TIA and Minor Ischemic Stroke (POINT) trial, in which substantial efforts were made to capture early stroke recurrence.¹⁵ First, we examined sex differences in recurrent ischemic stroke rates after minor stroke or TIA. Second, we evaluated sex differences in the strength of the association between common conventional stroke risk factors and recurrence.

Patients and Methods

Design

POINT was a randomized clinical trial that evaluated aspirin and clopidogrel (dual anti-platelet therapy) versus aspirin alone for preventing major ischemic events in patients with minor ischemic stroke or TIA.^15,16 In total, 4,881 individuals were enrolled from May 2010 to December 2017, across 269 sites globally. Subsequently, these participants were followed for up to 90 days for the primary composite outcome of cardiovascular events and secondary outcomes including recurrent ischemic stroke. The National Institute of Neurological Disorders and Stroke (NINDS) funded the trial and also made available the deidentified data employed in this analysis. The Weill Cornell Medicine Institutional Review Board granted an exemption for analysis of these data.

Patient population

Adults 18 years and older were randomized into the POINT trial within 12 hours of onset of either minor ischemic stroke (National Institutes of Health Stroke Scale score 0–3) or high-risk TIA (ABCD² score ≥4). Brain imaging was obtained before randomization to confirm ischemic stroke and rule out hemorrhage or other symptom etiologies. Patients who were candidates for thrombectomy, thrombolysis, or on anticoagulant therapy were not included in the study. Women who were pregnant or at risk for pregnancy were excluded from POINT because the safety of clopidogrel in pregnant women was not established.¹⁵

Measurements

Our primary exposure variable was self-reported sex. The primary outcome in our analysis was recurrent ischemic stroke that occurred after randomization in POINT. This was defined as (1) the rapid onset of a new focal neurological deficit with evidence of infarction or (2) rapid worsening of an existing deficit attributed to a new infarction. Outcomes of the POINT trial were adjudicated by a blinded endpoint adjudication committee. Covariates included the index event type (minor stroke versus TIA), randomization arm (clopidogrel versus placebo), demographic characteristics, participant site (US versus non-US), and stroke risk factors. Demographics included age, race, and ethnicity. Stroke risk factors included hypertension, diabetes mellitus, coronary heart disease, congestive heart failure, valvular heart disease, atrial fibrillation, carotid atherosclerosis, and history of tobacco use, based on baseline ascertainment. Participants were categorized as having atherosclerosis if they had at least 50% stenosis in either carotid artery at the time of the index event.

Statistical analysis

The primary statistical analysis entailed use of a Cox proportional hazards model to compare the risk of recurrent stroke in women versus men. We constructed the following stepwise models: (1) unadjusted, (2) adjusted for randomization arm (clopidogrel versus placebo) and event type (stroke versus TIA), (3) additionally adjusted for age, race, ethnicity, and site (US vs non-US), and 4) additionally adjusted for pre-existing stroke risk factors and risk factors ascertained at the time of the index event. Given prior evidence of sex differences in stroke risk factors in younger people with stroke,^6,17 we used formal interaction term testing and prespecified subgroup analyses to determine if the association between sex and recurrent stroke differed by age group (<60 versus >60 years old). We dichotomized at 60 years as this age is within the range of cut offs used to define stroke in young¹⁸ while ensuring adequate sample sizes in the two subgroups. We additionally used interaction term testing and prespecified subgroup analyses to determine if the association between sex and recurrent stroke varied by enrollment site (US versus non-US) and by index event type (ischemic stroke versus TIA). In a post-hoc analysis, we tested for interaction by treatment arm (clopidogrel versus placebo). In subgroup analyses by event type, we included the NIHSS score as a covariate for models of stroke and the ABCD² score as a covariate in models for TIA. Finally, we used interaction term testing in adjusted models to evaluate sex differences in the relationship between conventional stroke risk factors which were common in the POINT trial participants (diabetes, hypertension, carotid atherosclerosis, and tobacco use) and recurrent ischemic stroke risk. Subgroup analyses were performed when significant interactions were identified. The threshold of statistical significance was set at α = .05. All statistical analyses were performed using SAS (version 9.4).

Results

Of the 4,881 POINT trial participants with minor stroke or high-risk TIA, 2,195 (45%) were women and 2,686 (55%) were men. Women and men participants were similar with respect to baseline characteristics apart from differences in racial groups in addition to smoking and coronary heart disease, which were both more common in men (Table 1). Rates of assignment to clopidogrel did not differ by sex.

Table 1.

Baseline characteristics of POINT trial participants, stratified by sex.

Characteristic*	Women	Men
Demographics
Age, mean (SD)	65.9 (13.5)	63.5 (12.7)
Race
White	69.3	75.7
Black	23.8	16.5
Other	6.8	7.8
Ethnicity
Hispanic	8.3	7.6
Region
U.S.	90.6	85.4
Other countries	9.4	14.6
Treatment Assignment
Clopidogrel + aspirin	50.0	49.7
Placebo + aspirin	50.0	50.3
Medical History
Hypertension	72.2	66.6
Diabetes mellitus	28.2	26.8
Coronary heart disease	7.7	12.3
Congestive heart failure	2.6	2.6
Valvular heart disease	2.0	1.5
Atrial fibrillation	1.3	0.8
Carotid atherosclerosis^†	6.8	9.0
History of tobacco use	21.6	31.9
Initial Event Type
TIA	45.4	41.4
ABCD2, median (IQR)	5.0 (4.0–6.0)	5.0 (4.0–5.0)
Ischemic stroke	54.6	58.6
NIHSS, median (IQR)	2.0 (1.0–2.0)	2.0 (1.0–2.0)

Data presented as % unless otherwise stated.

^†Participants were categorized as having atherosclerosis if they had at least 50% stenosis in either carotid artery at the time of the index event.

During the 90-day follow-up period, 267 (5%) ischemic strokes occurred; 121 were in women and 146 in men. The cumulative risk of recurrent ischemic stroke was not significantly different among women (5.76%; 95% CI, 4.84%–6.85%) compared to men (5.67%; 95% CI, 4.83%–6.63%) (Figure 1). In an unadjusted Cox proportional hazards model, women were not at a different risk of recurrent stroke as compared to men (hazard ratio [HR] 1.02, 95% CI 0.80–1.30). After stepwise adjustment for covariates, there was no association between sex and 90-day recurrent ischemic stroke risk, including in the final model, which was adjusted for index event type, study treatment assignment, demographics, and vascular risk factors (HR 0.94; 95% CI 0.72–1.22) (Table 2).

Figure 1.

Title: Kaplan–Meier curves demonstrating risk of recurrence stroke by sex in the Platelet-Oriented Inhibition in New Transient Ischemic Attack or Minor Ischemic Stroke Trial. Caption: The rate of recurrent stroke was similar in men and women enrolled in the Platelet-Oriented Inhibition in New Transient Ischemic Attack or Minor Ischemic Stroke Trial over 90 days.

Table 2.

Association between sex and 90-day ischemic stroke recurrence.

Model	Hazard ratio^* (95% CI)
Unadjusted	1.02 (0.80–1.30)
Adjusted for index event type^† and study arm^‡	1.04 (0.82–1.33)
Additionally, adjusted for age, race, ethnicity, site	0.95 (0.74–1.21)
Additionally, adjusted for cardiovascular risk factors^\| ^\|	0.94 (0.72–1.22)
Stratified by age^#
Age <60 years	0.66 (0.42–1.05)
Age >60 years	1.18 (0.84–1.65)
Stratified by index event type^#
TIA	0.99 (0.62–1.59)
Ischemic stroke	0.91 (0.66–1.26)
Stratified by site^#
US sites	0.90 (0.68–1.19)
Non-US sites	1.62 (0.67–3.93)

^*Hazard ratios compare the risk of recurrence in women versus men.

^†TIA versus ischemic stroke.

^‡Aspirin + clopidogrel versus aspirin + placebo.

^||Hypertension, Diabetes mellitus, Coronary heart disease, Congestive heart failure, Valvular heart disease, Atrial fibrillation, Carotid atherosclerosis, Tobacco use.

^#Adjusted for all above covariates.

Formal interaction testing revealed a significant interaction between age group (<60 years versus >60 years old) and sex in the association with 90-day recurrent ischemic stroke (P=0.04). In participants below 60 years of age, the 90-day recurrent ischemic stroke rate was 6.04% (95% CI 4.75–7.67) in men, while it was 4.49% (95% CI, 3.20–6.69) in women. In participants over 60 years of age, the 90-day recurrent ischemic stroke rate was 5.41% (95% CI, 4.39–6.67) in men and 6.40% (95% CI, 5.33–7.83) in women. In stratified maximally adjusted Cox proportional hazards models, among participants 60 years of age and younger, women appeared to have a non-significantly lower rate of recurrence than men (HR 0.66; 95% CI 0.42–1.05), but this was not the case among participants over 60 years old (HR 1.18; 95% CI 0.84–1.65). Interactions by index event type (P=0.82) and site (P=0.49) were not significant, without meaningful differences between subgroups (Table 2). Additionally, in a post-hoc analysis, an interaction between sex and treatment arm was not identified (P= 0.62).

Last, we assessed sex differences in the relationship between conventional stroke risk factors and recurrent ischemic stroke. Overall, in adjusted models, the associations between conventional risk factors and recurrent ischemic stroke were as follows: diabetes (HR, 1.52; 95% CI, 1.15–2.00), hypertension (HR, 1.29; 95% CI, 0.93–1.79), carotid atherosclerosis (HR, 2.8; 95% CI, 2.01–3.90), and tobacco use (0.76; 95% CI, 0.56–1.03). Significant interactions with sex for diabetes (P=0.11), hypertension (P=0.70), carotid atherosclerosis (P=0.17), and tobacco use (P=0.88) were not seen.

Discussion

Among participants in the POINT randomized controlled trial with minor ischemic stroke or high-risk transient ischemic attack, we did not find an association between sex and 90-day risk of ischemic stroke recurrence. Additionally, sex did not modify the association between common conventional stroke risk factors and recurrent ischemic stroke.

Our findings build on prior work on sex differences in stroke recurrence. A prior analysis of data from a prospective cohort study based in Canada found men and women had a similar risk of 90-day ischemic stroke after presenting with stroke-like symptoms.¹¹ In this study, 452 of the 1,648 patients enrolled were diagnosed with a stroke mimic. In contrast, our use of data from the POINT trial ensured that all patients in our analysis had an adjudicated ischemic stroke or transient ischemic attack as their index event. Further, the POINT trial carefully included and adjudicated early recurrent events, which may have been missed otherwise in an observational study design. These methodological differences may explain the higher 90-day recurrence rate in the POINT trial (∼5% vs ∼2%). An alternate explanation is that the mean age in POINT was relatively young, with a relatively low prevalence of high-risk conditions such as atrial fibrillation.^19–21 Additional prior studies on the association between sex and stroke recurrence include analyses of administrative claims data from a rural region of the United States¹² and for diabetic patients in Ontario, Canada.¹³ Though these data were from narrowly defined populations and lacked granular clinical data, their findings were similar to ours.

Taken together, our findings suggest that the overall rate of recurrent ischemic stroke is similar in women and men, despite the influence of sex-specific risk factors unique to men and women.^8,17,22,23 This has several possible explanations. First, it is possible that the effects of sex-specific risk factors unique to women and men are similar in magnitude, resulting in similar overall recurrence rates at the group level. Second, the effects of sex-specific risk factors may be too modest, relative to the large effect of conventional stroke risk factors,²⁴ to result in an appreciable difference in recurrence rates. Third, trial participation may obscure sex differences in stroke recurrence due to potential healthy participant selection bias and the controlled environment of the trial. Fourth, it is possible that sex differences in recurrence do exist, but among age groups (such as octogenarians and nonagenarians) not well represented in this analysis. Notwithstanding, our findings do not suggest that there exists an unidentified and influential sex-specific risk factor for ischemic stroke recurrence. Additionally, while prior research has suggested that the strength of the relationship between conventional stroke risk factors and incident stroke risk varies by sex,¹⁰ this does not appear to be the case for recurrent ischemic stroke.

We identified a possible interaction between age and sex, with women having a non-significantly lower rate of recurrence than men under age 60. This may be due to a higher burden or greater severity of conventional stroke risk factors in young men than in young women.²⁵ Alternatively, differences in care may also account for our findings. Specifically, some young women with recurrent stroke may have been misdiagnosed with a stroke mimic and not reported to the trial,^7,26 although this is unlikely given the rigorous nature of follow-up in POINT. Third, this finding may be due to chance in this post-hoc analysis. Yet, given the increasing incidence of stroke in younger adults,^27–29 surveillance for sex differences in outcomes including recurrent stroke in this population may be warranted.

The strengths of our analysis include the use of a multicenter clinical trial dataset with adjudicated 90-day ischemic stroke outcomes. Our findings should be interpreted considering several limitations. First, the POINT trial was not designed to address sex differences in stroke recurrence; therefore, sex-specific risk factors were not tabulated in the trial. Additionally, we analyzed data for a clinical trial population, which excluded pregnant women. Pregnancy-related stroke is a unique clinical scenario that is not accounted for by these data, thus highlighting the potential lack of generalizability to the greater public. However, pregnancy-related stroke is rare and including it would have unlikely changed our results.^17,30 Generalizability may be further limited for the additional reason that the POINT trial population does not fully represent stroke across the lifespan; sex differences in high-risk conditions, such as atrial fibrillation,^19,20 among older individuals may not be reflected in our findings. Second, the trial had a 90-day follow-up period; so, it is unclear if our results hold true for long-term recurrence. Third, the POINT trial excluded participants with moderate and severe ischemic stroke, so these results may not be generalizable to all people with ischemic stroke. However, there was no interaction by event type (ischemic stroke versus TIA), so it is unlikely that our results are specific to minor ischemic stroke alone.

Conclusions

Among people with minor ischemic stroke and high-risk transient ischemic attack, men and women had a similar short-term ischemic stroke recurrence risk, and sex did not modify the impact of conventional stroke risk factors on recurrent ischemic stroke risk.

Footnotes

Acknowledgments

This research is based on the National Institute of Neurologic Disease and Stroke’s Archived Clinical Research data (Trial name: Platelet-Oriented Inhibition in New TIA and Minor Ischemic Stroke; Principal Investigator: S. Clairborne Johnston, MD, PhD; grant number: 1U01S062835-01A1) received from the Archived Clinical Research Dataset web site.

Declaration of conflicting interests

EB and SSO declare that there is no conflict of interest. HK serves as a principal investigator for the NIH-funded ARCADIA trial (National Institute of Neurological Disorders and Stroke U01NS095869) which receives in-kind study drug from the BMS-Pfizer Alliance for Eliquis® and ancillary study support from Roche Diagnostics, serves as Deputy Editor for JAMA Neurology, serves as a steering committee member of Medtronic’s Stroke AF trial (uncompensated), and serves on an endpoint adjudication committee for a trial of empagliflozin for Boehringer-Ingelheim. NP has received grants from the Leon Levy Foundation and the New York State Empire Clinical Research Investigator Program unrelated to this work and personal compensation for medicolegal consulting.

Funding

NP was supported by the Florence Gould Endowment for Discovery in Stroke for this work and receives funding from the NIH/NIA (K23AG073524).

Informed consent

Informed consent was not sought for this article because this was a retrospective analysis of existing, deidentified data.

Ethical approval

The Weill Cornell Medicine Institutional Review Board granted an exemption for analysis of these data because this was a retrospective analysis of deidentified data.

Trial registration

Not applicable because this was a retrospective analysis of existing data from a previously published trial ( Identifier: NCT00991029).

Guarantor

NP.

Contributorship

EB and NP researched literature and conceived the study. EB wrote the first draft of the manuscript. NP performed statistical analyses. All authors reviewed and edited the manuscript and approved the final version of the manuscript.

ORCID iD

Neal S. Parikh

References

Virani

Alonso

Benjamin

, et al. Heart disease and stroke statistics-2020 update: a report from the american heart association. Circulation 2020; 141: e139–e596.

Amarenco

Lavallée

Monteiro Tavares

, et al. Five-year risk of stroke after TIA or minor ischemic stroke. New England Journal of Medicine 2018; 378: 2182–2190.

Phan

Blizzard

Reeves

, et al. Sex differences in long-term mortality after stroke in the INSTRUCT (INternational STRoke oUtComes sTudy): a meta-analysis of individual participant data. Circulation: Cardiovascular Quality and Outcomes 2017; 10: e003436.

Gall

Phan

Madsen

, et al. Focused update of sex differences in patient reported outcome measures after stroke. Stroke 2018; 49: 531–535.

Poorthuis

MHF

Algra

, et al. Female- and male-specific risk factors for stroke. JAMA Neurology 2017; 74: 75.

Demel

Kittner

Ley

, et al. Stroke risk factors unique to women. Stroke 2018; 49: 518–523.

Reeves

Bushnell

Howard

, et al. Sex differences in stroke: Epidemiology, clinical presentation, medical care, and outcomes. The Lancet Neurology 2008; 7: 915–926.

Andersen

Olsen

. Age- and gender-specific prevalence of cardiovascular risk factors in 40 102 patients with first-ever ischemic stroke. Stroke 2010; 41: 2768–2774.

Madsen

Howard

Jiménez

, et al. Impact of conventional stroke risk factors on stroke in women. Stroke 2018; 49: 536–542.

10.

Peters

SAE

Carcel

Millett

ERC

, et al. Sex differences in the association between major risk factors and the risk of stroke in the UK biobank cohort study. Neurology 2020; 95: e2715–e2726.

11.

AYX

Penn

Lesperance

, et al. Sex differences in presentation and outcome after an acute transient or minor neurologic event. JAMA Neurol 2019; 76: 962–968.

12.

Lambert

Chaudhary

Olulana

, et al. Sex disparity in long-term stroke recurrence and mortality in a rural population in the United States. Therapeutic Advances in Neurological Disorders 2020; 13: 1756286420971895.

13.

Dhamoon

Liang

Zhou

, et al. Sex differences in outcomes after stroke in patients with diabetes in Ontario, Canada. Journal of Stroke and Cerebrovascular Diseases 2018; 27: 210–220.

14.

Johnston

Gress

Browner

, et al. Short-term prognosis after emergency department diagnosis of TIA. JAMA 2000; 284: 2901–2906.

15.

Johnston

Easton

Farrant

, et al. Clopidogrel and aspirin in acute ischemic stroke and high-risk TIA. New England Journal of Medicine 2018; 379: 215–225.

16.

Kamel

Zhang

Kleindorfer

, et al. Association of Black Race With Early Recurrence After Minor Ischemic Stroke or Transient Ischemic Attack. JAMA Neurology 2020; 77: 601–605.

17.

George

. Risk factors for ischemic stroke in younger adults. Stroke 2020; 51: 729–735.

18.

Ekker

Verhoeven

Vaartjes

, et al. Stroke incidence in young adults according to age, subtype, sex, and time trends. Neurology 2019; 92: e2444–e2454.

19.

Friberg

Benson

Rosenqvist

, et al. Assessment of female sex as a risk factor in atrial fibrillation in Sweden: Nationwide retrospective cohort study. BMJ 2012; 344: e3522.

20.

Mikkelsen

Lindhardsen

Lip

GYH

, et al. Female sex as a risk factor for stroke in atrial fibrillation: A nationwide cohort study. Journal of Thrombosis and Haemostasis 2012; 10: 1745–1751.

21.

Marini

De Santis

Sacco

, et al. Contribution of Atrial Fibrillation to Incidence and Outcome of Ischemic Stroke. Stroke 2005; 36: 1115–1119.

22.

Ekker

de Leeuw

F-E

. Higher Incidence of Ischemic Stroke in Young Women Than in Young Men. Stroke 2020; 51: 3195–3196.

23.

Leppert

Burke

, et al. Young women had more strokes than young men in a large, United States claims sample. Stroke 2020; 51: 3352–3355.

24.

O'Donnell

Chin

Rangarajan

, et al. Global and regional effects of potentially modifiable risk factors associated with acute stroke in 32 countries (INTERSTROKE): A case-control study. The Lancet 2016; 388: 761–775.

25.

Aigner

Grittner

Rolfs

, et al. Contribution of established stroke risk factors to the burden of stroke in young adults. Stroke 2017; 48: 1744–1751.

26.

Bushnell

Chaturvedi

Gage

, et al. Sex differences in stroke: challenges and opportunities. Journal of Cerebral Blood Flow & Metabolism 2018; 38: 2179–2191.

27.

Edwards

Kapral

Lindsay

, et al. Young stroke survivors with no early recurrence at high long-term risk of adverse outcomes. Journal of the American Heart Association 2019; 8: e010370.

28.

Kissela

Khoury

Alwell

, et al. Age at stroke: temporal trends in stroke incidence in a large, biracial population. Neurology 2012; 79: 1781–1787.

29.

George

Tong

Bowman

. Prevalence of cardiovascular risk factors and strokes in younger adults. JAMA Neurology 2017; 74: 695–703.

30.

Elgendy

Gad

Mahmoud

, et al. Acute stroke during pregnancy and puerperium. Journal of the American College of Cardiology 2020; 75: 180–190.