Beyond brain age: Empirically-derived proxy measures of mental health

Biological aging is revealed by physical measures, e.g., DNA probes or brain scans. Individual differences in personal functioning are instead explained by psychological constructs. Constructs such as intelligence or neuroticism are typically assessed by specialized workforce through tailored questionnaires and tests. Similar to how brain age captures biological aging, intelligence and neuroticism may provide empirical proxies for mental health. Could the combination of brain imaging and sociodemographic information yield measures for these constructs that do not rely on human judgment? Here, we built proxy measures by applying machine learning on multimodal MR images and rich sociodemographic information from the largest brain-imaging cohort to date: the UK Biobank. Objective comparisons revealed that all proxies captured the target constructs and related to health-contributing habits beyond the measures they were derived from. Our results demonstrate that proxies targeting classical psychological constructs reveal facets of mental health complementary to information conveyed by brain age.