Sex differences in the relationship between physical activity and cognition as mediated by leukocyte telomere length: a UK Biobank study

Background Evaluating sex differences in modifiable risk factors for Alzheimer’s disease could provide valuable information regarding the mechanisms by which these factors confer risk. Physical activity is a risk factor that has been shown to positively impact both telomere length, a marker of cellular age, and cognition. The goal of this study was to evaluate whether telomere length mediates the association between physical activity and cognition differently by sex. Method Participants from the UK Biobank were included in the sample based on availability of demographic variables, physical activity questionnaires, leukocyte telomere length, and cognitive measurements (Fluid Intelligence (FI), Numeric Memory (NM), Reaction Time (RT)). Metabolic Equivalent Task (MET) scores were derived from self‐report questionnaires following the International Physical Activity Questionnaire (IPAQ) guidelines. Relative leukocyte telomere length was measured by quantitative PCR and adjusted for technical variation by the UK Biobank. The indirect effect of physical activity on cognition via telomere length was evaluated using causal mediation analyses in R. Models were adjusted for age, body mass index, and Townsend Deprivation Index (Figure 1). A priori sex‐stratified models were used to examine sex differences. Result The sample included 374,761 adults (52.2% women, mean age±SD = 56.3±8.12 years, see Table 1). Linear regression analyses identified negative associations between weekly physical activity measures (i.e., summed MET minutes of all activity p