Brain Tissue Proportionality and General Cognitive Ability Across Three Adult Cohorts

Abstract General cognitive ability may relate to the proportional allocation of brain tissue beyond overall tissue volume. We examined five FreeSurfer compartments using compositional data analysis: cortical gray matter, cerebral white matter, subcortical gray matter, total cerebellum, and brainstem. Human Connectome Project Young Adult data (HCP-YA; N = 1,101) provided discovery estimates; the Dallas Lifespan Brain Study (DLBS; N = 418) and Amsterdam Open MRI Collection ID1000 (AOMIC; N = 926) provided independent cohort tests. Four isometric log-ratio coordinates represented composition, with age, sex, and log total analyzed tissue volume included separately. Composition explained additional cognitive variance in DLBS (partial R² = 4.20%, p = .00154) and AOMIC (partial R² = 1.35%, p = .014). The total-cerebellum-to-cerebral-white-matter contrast selected in HCP was positive in both replication cohorts: standardized β = 0.162 (heteroskedasticity-robust 95% confidence interval 0.072–0.252) in DLBS and 0.097 (0.024–0.170) in AOMIC. Participant-pairs bootstrap intervals supported both associations. Replacing total tissue volume with estimated intracranial volume attenuated the AOMIC association, whereas including both preserved it. Relative macroanatomical allocation is associated with broad cognitive ability conditional on analyzed tissue volume, with differing effect magnitudes across cohorts. These cross-sectional findings do not establish causal tissue trade-offs.