ANCOVA and post hoc Tukey HSD tests were used to detect between-group differences in (A) subcortical volumes, (B, C) cortical volumes, (D) global brain volumes, and (E) cognitive function in two subtypes of institutionalized patients and demographically matched healthy controls. Age, sex, and ICV were included as covariates for brain volumes, and age, sex, and education level were covariates for BACS raw scores. In bar charts, significant patient-control and between-subtype differences, determined by FDR-corrected P-values generated in post hoc pairwise tests, are marked by one and two asterisks, respectively. Shading bars represent Glass’s delta (Δ) effect sizes, which were calculated after removing variance related to corresponding covariates and used to demonstrate patient-control differences for each subtype. Error bars mean 95% confidence interval of Δ. In cortical maps, only regions that survived FDR corrections are colored by t statistics from post hoc tests. CV Cortical volume; GMV Gray matter volume; HC Healthy controls; L the left hemisphere; R the right hemisphere; S1 Subtype 1 of institutionalized patients; S2 Subtype 2 of institutionalized patients; TBV Total brain volume; WMV White matter volume.