Introduction: While men suffer from more strokes in younger and middle age, the incidence of stroke is higher among women in older age. Specifically, 60% of stroke fatalities occur in females and 40% occur in males. Recent findings have attributed these sex-specific differences to a surge in the prevalence of cardiovascular risk factors among women. Traditionally, men have presented more frequently with micro-vascular complications (hypertension, diabetes, hyperlipidemia, and obesity), but now these risks are increasing in women.
Hypothesis: Whether the modifiable risk factors examined, both alone and together, will differ in their ability to predict stroke severity across sex.
Methods: A retrospective analysis of 811 primary stroke cases from two comprehensive and two primary stroke centers in a single region were analyzed for prevalence of modifiable and non-modifiable risk factors. Risk factors such as medical comorbidities, clinical findings and behavioral risk factors were compared across sexes and then assessed for correlation with initial presentation severity and ambulatory status. The relative influence of risk factors on sex differences in stroke severity (NIH Stroke Scale) and symptomatology were tested.
Results: Females were more likely to present as non-Hispanic, with history of migraines and more severe levels of hypertension, with hypertensive crisis status at presentation relative to normal hypertensive state (all p < 0.05). Males were more likely to present with history of coronary artery disease, documentation of drug or alcohol use, higher fasting blood sugar, A1c, and BMI levels at presentation (all p < 0.05). Multivariate models improved the observed association between female sex and severe NIHSS score (OR = 2.1 [1.03-4.35]) through inclusion of hypertension, fasting blood sugar, and diabetes status.
Conclusions: While the sex differences in strokes are well studied, the complex interrelationships between sex, variance in known medical risk factors, and severity of stroke presentation are not so well understood. This study identifies new patterns in risk by building independent models for odds of stroke severity for each sex, which may in turn help to explain sex differences in stroke presentation phenotype.