Considering sex in mild‐to‐moderate Alzheimer’s disease clinical trials: A meta‐analysis of ADCS trials

Abstract

AbstractBackgroundNearly two thirds (62%) of Alzheimer’s disease (AD) patients in the United States are women (Nebel 2018; Rajan, 2021). Our goal in this project was to assess whether women are proportionately represented in AD clinical trials and whether sex was associated with other baseline patient characteristics.MethodsWe used data from nine mild‐to‐moderate AD trials conducted by the AD Cooperative Study (ADCS): Docosahexaenoic acid (DHA), FYN Kinase, Homocysteine, Huperzine (Hu), IV immunoglobulin (IVIg), nonsteroidal anti‐inflammatory drugs (NSAIDs), Resveratrol (Res), Simvastatin, Valproate. Sample sizes ranged from 120 (Res) to 642 (IVIg) participants. We compared the proportions of females enrolled to disease prevalence estimates. We conducted a random effects meta‐analysis to estimate female participation rates across the nine trials. Specifically, we used a generalized linear mixed effects model with logit link and random intercept to estimate the marginal and trial specific prevalence of female participants. To compare participant characteristics by sex we pooled samples across trials and used unadjusted chi‐square (categorical variables) and t‐tests (continuous variables).ResultsThe proportion of females enrolled exceeded 50% in each trial, with the exception of the FYN Kinase trial (44.7%). In all but the Hu trial, the rate of female enrollment was less than the estimated female disease prevalence. The estimated prevalence of female participants across all trials was estimated to be 56% (95% CI 0.53 to 0.58) (Figure 1). Females were similar to males in terms of age and APOE genotypes. Compared to males, females had lower education, less often enrolled with a spouse, less often lived with their study partner, and more often were from an underrepresented racial or ethnic group.ConclusionsThough these trials recruited more females than males, on average, women were still underrepresented compared to national disease prevalence estimates. Female and male participants exhibited psychosocial differences in education, race, and study partner characteristics that could affect trial outcomes and should be further investigated.