Sex Differences in Nonadherence to Secondary Stroke Prevention Medications Among Patients With First-Ever Ischemic Stroke

Abstract

Background: More women than men experience recurrent strokes. Medication adherence is critical to prevent recurrence; however, studies investigating sex differences are limited. We examined sex differences in poststroke medication adherence, overall and by drug class, and identified influencing factors.

Methods and results: Patients with first-ever ischemic stroke were identified from a population-based study (2008-2019). At 90 days after a stroke, self-reported medication adherence was defined as never or rarely missing a dose in a typical week for each secondary stroke prevention medication (antihypertensives, cholesterol-lowering drugs, antiplatelets, anticoagulants). We generated prevalence ratios (PRs) using modified Poisson models to assess sex differences with and without adjustment for potential confounding factors, including demographics, social factors, health system-related, lifestyle-related, health condition-related, prestrokehealth-related, and stroke-related factors. Among 1324 participants (48.4% women, 58.0% Mexican American individuals), women were more likely to report nonadherence to cholesterol-lowering drugs (PR, 1.80 [95% CI, 1.14-2.84]) and antiplatelets (PR, 1.53 [95% CI, 1.003-2.34]). Adjusting for obesity attenuated while adjusting for age, marital status, access to care, smoking, and alcohol consumption accentuated sex differences. Race and ethnicity modified the sex difference in nonadherence to cholesterol-lowering drugs (P_interaction=0.054) such that the sex difference was larger in Mexican American individuals (PR, 3.00 [95% CI, 1.65-5.48]) than in non-Hispanic White individuals (PR, 1.30 [95% CI, 0.52-3.27). No significant sex differences were found for nonadherence to antihypertensives and overall nonadherence.

Conclusions: Poststroke medication nonadherence was more prevalent among women than men. This was partially due to the confounding effects of lifestyle, marital status, and access to care, suggesting potential subgroups for interventions to improve adherence.

Keywords: medication adherence; secondary stroke prevention; sex differences.

Figure 1. Flow diagram for participants with first‐ever ischemic stroke from the BASIC Project, 2008 to 2019.

BASIC indicates Brain Attack Surveillance in Corpus Christi; and IPW, Inverse Probability Weighting.

Figure 2. Individual factors associated with sex differences in medication nonadherence 90 days after stroke in the BASIC, 2008 to 2019.

A shows results for overall nonadherence. B–D show results for nonadherence to antihypertensives, cholesterol‐lowering drugs, and antiplatelets, respectively. Each of the potential contributors was separately added to the unadjusted modified Poisson model. We then quantified the contribution of each factor by calculating the percent change in the coefficient for sex (Δβ_sex) according to the formula 100×([β₀]‐[β_NEW])/(β₀), where β₀ is the coefficient for sex in the unadjusted model (β₀), and (β_NEW) is the coefficient for sex in each of the confounder‐adjusted models. BASIC indicates Brain Attack Surveillance in Corpus Christi.