Sex differences across age in cognitive tests in the Rhineland study

Abstract

Introduction Neurodegeneration and memory decline are of growing concern due to population aging. It is well known that both age and educational history impact cognitive function. Furthermore, a growing body of evidence suggests that biological sex may also play a role. Various studies have reported that women tend to outperform men in episodic memory tasks, whereas men tend to outperform women in numerical working memory tasks. However, sex differences are less clear for tests that measure executive function or processing speed, with previous work reporting either no sex differences or a marginally better performance of men. Spatial working memory and premorbid IQ seem not to be influenced by sex. Only limited and inconclusive data exist as to whether the associations of sex and cognition change with aging. We aimed to examine the effects of sex over a wide adult age range and across different cognitive domains in a population-based cohort. Methods The study is based on cross-sectional data from the first 1306 participants of the Rhineland Study. The Rhineland Study is a population-based cohort study that includes participants from the age of 30, living in and around the city of Bonn. We excluded participants > 90 years of age (n = 4) which left n = 1302 to be included in the current analyses. In addition to advanced biological phenotyping, the study protocol includes a cognitive test battery. Episodic memory was assessed using the Verbal Learning and Memory Test (VLMT) immediate and delayed recall, working memory through the Corsi block-tapping test (forward and backward) and the digit span test (forward and backward), processing speed and executive function through the trail-making test A and B (TMT-A, TMT-B), and crystallized intelligence (as a measure of premorbid IQ) through a multiple choice vocabulary test (MWT-B). We performed multi-factor analysis of variance on whether sex or age (in 10-year age groups: 30–39, 40–49, 50–59, 60–69, 70–79, 80–89) were associated with cognitive test performance (z-standardized), either independently or through interaction. All analyses were adjusted for educational attainment. We corrected for multiple testing using the Holm-Bonferroni method. Tukey's test was administered post-hoc to assess the direction of group differences. Results Mean age in our sample was 56 years (SD = 14 years) and 56% were women. Age group was significantly associated with performance for all tests. Compared to younger participants older participants performed better in crystalline intelligence, but worse on all other tests. Men performed significantly better than women on tests of executive function (TMT-B: Mdiff = −0.17; 95% CI: −0.26, −0.08) and working memory (Corsi: forward Mdiff = 0.28; 95% CI: 0.17, 0.38; backward Mdiff = 0.23; 95% CI: 0.12, 0.33; digit span: forward Mdiff = 0.20; 95%CI 0.10, 0.31; backward Mdiff = 0.16; 95%CI 0.05, 0.26). On the other hand, women performed significantly better than men on the episodic memory test (VLMT: immediate Mdiff = 0.45; 95% CI: 0.36, 0.54; delayed Mdiff = 0.42; 95% CI: 0.33, 0.52). We observed no interaction between age and sex across all tests, indicating no sex differences in how cognitive performance changes over the adult age range. Conclusion Our results confirm previously observed sex differences in episodic memory (women outperform) and working memory tasks (men outperform). We also found suggestive evidence for men outperforming women in measures of executive function. Notably, we found no evidence for any variations in these associations across the adult lifespan.