Treatment for Alzheimer Disease-Sex and Gender Effects Need to Be Explicitly Analyzed and Reported in Clinical Trials.

Abstract

Martinkova et al1 have described the representation and analysis of sex-specific data from published randomized clinical trials of pharmacologic agents for all stages of Alzheimer disease (AD) that enrolled more than 100 adult participants. They addressed 3 issues: (1) the proportion of women enrolled, (2) the proportion of studies that reported sex-stratified data, and (3) temporal trends in enrollment or reporting by sex. The authors found that women represented 59% of study participants, that this percentage did not change significantly over the past decade, and that there was a lesser chance of enrollment of women in trials in North America compared with the rest of the world. They also report that whereas approximately half of the studies may have included sex in randomization schema, fewer than 15% of the publications described methods for analyzing results by sex or presented analyses of potential sex differences in responses. What conclusions should be drawn from these analyses, and what does this study add to the literature? The work confirms a prior meta-analysis that found higher enrollment of women (63.8%) than men in trials of approved AD therapeutics.2 Women are estimated to represent, on average, 68.2% of patients with AD dementia in Europe and 62.1% of those in the US. There are no mandated inclusion metrics for proportions of women or men in clinical trials; however, a participant-toprevalence ratio (ie, the ratio of the clinical trial participant population to the patient population with the disorder to be treated) of 0.8 to 1.2 is usually considered adequate. The report by Martinkova et al1 describes a participant-to-prevalence ratio between 0.87 and 0.95 for women. Enrollment of women into trials of pharmacologic agents for AD appears to be adequate. The striking omission described by the authors is the absence of data to evaluate potential sex or gender differences in responses to the AD drugs studied, also emphasized in an earlier meta-analysis.2 A considerable body of literature describes sex and gender differences in risks and the course of AD.3,4 Data from the Framingham Heart Study reported greater risk of AD dementia in women at age 45 years (1 in 5) than in men at the same age (1 in 10) and an overall increased lifetime risk in women older than 85 years.5 The mechanisms for higher risk of AD dementia in women than in men are not entirely understood. Biologically plausible explanations include longer lifespans on average in women than in men; the effects of sex hormones, including protective effects of testosterone or protective or deleterious effects of estrogen; differential effects of APOE4 gene alleles in men compared with women; age at menopause or duration of exposure to estrogens; and higher depression rates in women than in men. Sociologically plausible explanations include lower average educational attainment in women due to lack of opportunity and lower socioeconomic status in women compared with men. The interpretation of neuropsychological test results relies on corrections for such variables as level of education, sex, race, and age. However, many instruments lack appropriate, fully demographically corrected norms. Thus, it is reasonable to hypothesize that differences in responses to medication may exist between men and women with AD dementia as a result of factors that may be uncontrolled in study design. Data that could identify or address underlying mechanisms for potential sex-related differences in responses to AD medications were collected during the trials identified and analyzed in the systematic review by Martinkova et al,1 but sex-specific analyses were reported in less than 15% of the AD dementia study results.1 The authors also point out the relative paucity of biomarker availability (in vivo or postmortem) in the studies but do not sufficiently emphasize its importance. + Related article