Cognition, Lifestyles, and Environments: Quantifying the Roles of Body Physiology and the Brain

Lifestyle and environmental factors are increasingly promoted as modifiable targets for maintaining cognitive health. Yet, it remains unclear how well they predict cognitive functioning and to what extent their associations with cognition are reflected in brain and body measures. We applied machine learning to multimodal data from over 10,000 UK Biobank participants to evaluate the predictive value of twelve lifestyle and environment domains, spanning physical activity, diet, smoking, alcohol consumption, sleep, sexual behavior, electronic device use, and environmental exposures, for cognition, both individually and in combination, and quantified the extent to which these associations are captured by body and brain. A model integrating lifestyle and environment domains explained 23% of the variance in cognition at r=0.48, comparable to body and brain markers. Physical activity, diet, alcohol consumption, sun exposure, and local environment emerged as the strongest predictors. A composite brain marker integrating three neuroimaging modalities accounted for 57.7% of the lifestyle-cognition association, while a composite body marker spanning nine physiological systems accounted for 47.8%. Jointly, lifestyle, environment, body, and brain captured 92.6% of age-related cognitive variation. Therefore, integrating lifestyle and environmental factors enables robust prediction of cognition within the UK Biobank cohort, and much of this association is reflected in brain and bodily health. However, the generalizability of these findings to broader populations requires further evaluation.