On the Basis of Sex: Outcomes After Ischemic Stroke

Sex differences in ischemic stroke outcome have been reported in many published studies¹, and three recent papers in Stroke highlight these differences^2–4. It is clear that women have poorer outcomes than men, but increasing recognition of important confounders, most notably higher rates of pre-stroke disability experienced by women, has led to a more nuanced understanding of what accounts for these poorer outcomes. In large epidemiological data sets, women are, on average, five years older than men when they experience their first stroke^5–7. However, when age is controlled in multivariable models, women still fare worse. But is controlling for age enough? New studies reveal that other factors like pre-stroke functional status and post stroke depression are key contributors to the sex differences seen in stroke outcome. The poorer baseline function in women may also have major implications for women’s access to acute interventional treatments, as they may be “selected out” of clinical trials or therapies based on their higher pre-morbid modified Rankin scale (mRS). This is an especially important factor to consider as the paper by Sheth³ in this issue found that women are as likely, and perhaps more likely, to benefit from endovascular treatment.

The first study by Phan et al investigated sex differences in health related quality of life (HRQoL) in stroke survivors in the International Stroke Outcomes Study (INSTRUCT)². INSTRUCT was an international collaboration that had individual participant data from 13 studies from patients with first ever stroke². The authors analyzed data from four studies that obtained HRQoL at 1 and 5 years post stroke. As has been reported by others⁸, the authors found that women have a poorer HRQoL at both 1 and 5 years after their first stroke (1 year, median difference −0.147; 5 year median difference −0.090) as compared to men. However, when age, stroke severity, pre-stroke dependency and post stroke depression were included in the quantile regression model, the sex difference in outcome was no longer significant (1 year median difference, −.067, 5 year median difference −0.085). This is consistent with previous work reported for stroke mortality in 2017, also using the INSTRUCT cohort. Using Poisson modeling, it was found that women had a higher mortality rate ratio (MRR) at 1 (MRR 1.35, 35% higher) and 5 (MRR1.24, 24% higher) years post stroke. When controlled for age, pre-stroke functional status, stroke severity and history of atrial fibrillation, the sex differences reversed at 1 (MRR, 0.81) and 5 years (MRR, 0.76)⁹. This current study by Phan et al confirms that sex differences in stroke outcome follow the same pattern as those seen in mortality and stroke severity¹⁰. Perhaps, most importantly, this study clearly highlights the importance of capturing an assessment of pre-stroke function in large epidemiological studies. This has been seen in other cohorts, including those enrolled in the Oxford Vascular Study. Poorer outcomes seen after stroke or TIA in women at 1 month, 6 months, 1 year and 5 years were no longer significant after controlling for pre-morbid mRS¹¹. One important limitation of the current study by Phan et al was the high number of patients lost to follow up in the INSTRUCT cohort (72.5% in the Perth study and 61.6% in the Auckland study). Death or discharge to nursing facilities, which may be more common in older women, remain important outcome measures, and loss of these data points may lead to erroneous conclusions. Obtaining follow up data, although costly, should be a priority in these large epidemiological studies to more accurately determine outcomes for both men and women. It is clear that strategies to improve overall functional status would greatly benefit older stroke survivors, many of which are women. However, reversing the detrimental effects of aging is a difficult task, and new approaches are clearly needed.

The second published study by Sheth et al examined the effect of sex on outcomes after endovascular therapy (EVT) for acute ischemic stroke³. Patients from three trials, Solitaire FR With the Intention for thrombectomy (SWIFT), Solitaire FR Thrombectomy for Acute Revascularization (STAR), and Solitaire FR with the Intention for Thrombectomy as Primary Endovascular Treatment (SWIFT-PRIME) were analyzed to determine sex differences in disability adjusted life years (DALYs) after EVT^12–14. Consistent with the study by Phan et al, there was no sex difference in functional outcome after 90 days, despite the fact women were older and had higher rates of atrial fibrillation. However, when adjusted for factors including age, premorbid mRS, and stroke severity, women had higher DALYs after EVT as compared to men (10.6 vs. 8.5 years). This suggests that women are as likely, and perhaps even more likely to benefit from intervention.

The third study by Chalos et al presented pooled results from seven trials within the HERMES collaboration to assess the effect of sex on outcomes after EVT⁴. At the time of stroke, women were older (median 70 years vs. 66 years), had higher collateral scores, lower infarct volumes but had similar functional outcomes; mRS 0–2 was achieved in 39% of men and 39% of women. Five of the seven endovascular clinical trials (ESCAPE, EXTEND-IA, REVASCAT, SWIFT PRIME, THRACE) in the Chalos study and two out of the three trials (SWIFT PRIME, STAR) in the Sheth study excluded patients with poor premorbid function. Therefore, it is possible that older women with co-morbidities may have been excluded. Although the numbers suggest that women were equally represented in these trials (pooled analysis from the HERMES collaboration reports 48% women and 52% men were enrolled in the intervention group), the number of women and men screened for EVT may be a more revealing statistic¹⁵. The women in these trials were relatively young, with a median of 70 years in the HERMES cohort and 69 years in the study by Sheth et al, and presumably well enough to be offered endovascular therapy.

Interestingly, the women included in the Chalos study had better baseline National Institutes of Health Stroke Scale (NIHSS), higher collateral grades, smaller infarct volumes, and equal rates of atrial fibrillation compared to men, suggesting that the oldest, frailest women may have been excluded. The severity of stroke is worse in women in most epidemiological studies, possibly due to the higher rates of cardioembolic strokes which cause more strokes in women as compared to men¹⁶. Due to frailty, less women are prescribed anticoagulation as compared to men and this becomes a vicious cycle of increased stroke risk and disability¹⁷. The women in the Chalos et al paper had a lower stroke severity than men, which may be due to selection of healthier women, and underscores the importance of capturing pre-stroke mRS as a potential confounder to effect size.

So are there sex differences in ischemic stroke outcome? The study by Phan et al confirms that women have poorer HrQoL after stroke as compared to men. Although the sex difference “disappears” after correction for age, stroke severity, and pre-morbid function, the fact remains, women do worse. We cannot lose sight of this fact. There are important factors that lead to sex differences in outcome that we still do not understand, or in most cases, attempt to treat. Frailty and co-morbidities associated with aging are difficult targets, but some factors, most notably post-stroke mood disorders and other social factors such as isolation, may be viable approaches to improve outcomes. Rather than focusing on the difference between the sexes, the work by Phan et al suggests that we should focus our attention on potential targets to improve outcomes in women, such as treatment of depression. Mood disorders after stroke remain under-recognized and under-treated. Women have a higher prevalence of post stroke depression as compared to men (20.05 vs. 16.22% 3 months after stroke)¹⁸. Moreover, the severity is worse in women, 10% of men have long term severe symptoms vs. 20% of women with severe or very severe symptoms¹⁹. It is increasingly realized that social isolation in itself is a risk factor for post stroke depression and worse outcomes²⁰. Due to the survival advantage, elderly women are almost twice as likely to live alone as compared to men. This has been recapitulated at the bench, where socially isolated mice had poorer behavioral outcomes as compared to pair housed mice, and this was worse in females²¹. Targeting behavioral health, social isolation, and post stroke depression is vital if we hope to improve functional outcomes in women.

Another important factor, highlighted in these studies but overlooked in many stroke studies, is how poor functional status in women could affect the interpretation of clinical trial results. Women are older and have higher mRS/disability at the time of their first stroke, many may be using a cane or a walker. Therefore, although their post stroke outcome may appear poor (mRS 3 for example), if this was their baseline function, then this would be an excellent outcome. As suggested by Ganesh et al, a step change in mRS scores that takes into account premorbid status, should be used as an outcome measure in future studies²².

In conclusion, women benefit from EVT with higher DALYs after intervention compared to men, although selection bias is likely a factor. Women have worse premorbid function which seems to account for much of the higher mortality and poorer outcomes seen after stroke. Targeting certain modifiable factors such as aggressive anticoagulation for atrial fibrillation, may decrease cardioembolic stroke and reduce these differences. Treatment of mood disorders, especially post stroke depression, also appears to be a viable potential target to improve stroke outcomes in elderly women. Certainly, it is clear that sex should not be considered a factor in determining eligibility for acute stroke intervention. However elderly women may be excluded from some of our most efficacious interventions solely based on their poor pre-stroke functional status.