Moving beyond the longitudinal analyses, the spatial patterns of cross-sectional aging differences in CT across the lifespan were drawn from a previous study (N=17,075) by the ENIGMA Lifespan working group [20]. In that study, inter-scanner variability in the data was adjusted using an empirical Bayesian approach and then the effect of age on regional CT was modeled using higher-order fractional polynomial regression analyses with sex included as a covariate. Variances in regional data explained by age and its fractional polynomial combinations across the lifespan were calculated. Then, these data were divided into early (3–29 years), middle (30–59 years) and late life (60–90 years) periods and Pearson’s correlation coefficients between age and regional CT were calculated for each age group.
