Support for verbal memory reserve in women using an epigenetic marker of age acceleration

Abstract

Background To determine whether characterizing men and women based on rates of biological aging indexed by DNA methylation “GrimAge” reflect differences on measures of verbal and non‐verbal memory. Method Cross‐sectional data from 1,930 adults aged 65 or older (mean age = 75, SD = 7.4, Male = 43%, non‐Hispanic White = 74.8%, non‐Hispanic Black = 13.5%, Hispanic = 9.6%, other race/ethnicity = 2.1%) were included. The standardized residual from regressing chronological age on the DNA methylation GrimAge clock as the measure of age acceleration. Slow and accelerated “DNAmGrimAgers” were defined as individuals below or above 1 SD of their sex‐specific GrimAge acceleration Mean, respectively. Individuals between these cutoffs were categorized as “normal agers” resulting in six different groups (i.e., 3 female and 3 male). Performances on measures of verbal and non‐verbal memory and processing speed were compared across these six different groups, controlling for education, depressive symptoms and chronological age. We reran analysis adjusting for processing speed. Result When comparing men and women’s performances based on their matching DNAmGrimAge category, there was no difference in processing speed or visual memory (e.g., accelerated male DNAmGrimAgers performed similarly to accelerated female DNAmGrimAgers on speed). In contrast, women, regardless of their DNAmGrimAge category, outperformed men on verbal memory measures (in speed and non‐speed adjusted models). Additionally, while accelerated female DNAmGrimAgers had significantly lower memory performances compared to slow female DNAmGrimAgers, this group difference was no longer significant in the speed‐adjusted models. Conclusion Women maintained a verbal memory advantage over men, despite different rates of age acceleration, which was not accounted for by faster processing speed. These findings strengthen prior support for verbal memory reserve that is specific to women. Thus, relying on verbal memory assessments in women at risk of Alzheimer’s disease (AD) may not be sensitive enough to detect hippocampal impairment in the earliest stages. However, performances on visual memory measures may be a more accurate representation of hippocampal function in women and should be included in assessments where AD is suspected.