Abstract
Background
Higher neighborhood disadvantage, indexed by the Area Deprivation Index (ADI), is linked with lower MRI‐measured brain volume in cognitively unimpaired individuals. We investigated whether ADI is associated with longitudinal antemortem MRI‐derived brain volumes across the cognitive continuum in a multi‐site brain bank cohort.
Method
We analyzed 724 T1‐weighted MRI scans from 436 brain donors (Table 1) across 14 Alzheimer’s Disease Research Centers housed in the National Alzheimer’s Coordinating Center. Images were processed using a longitudinal pipeline. Donor addresses were geocoded and linked to ADI percentiles at death year. ADI was binarized (30% most vs 70% least disadvantaged) and used to predict brain volumes in linear mixed‐effects models. Outcomes included atrophy score (CSF / [GM + WM]), total brain volume, and hippocampal, parahippocampal, middle temporal gyrus, and frontal pole volumes, all normalized to intracranial volume. Covariates included age, age², sex, education, and dementia status (CDR 0–0.5 = no dementia; 1–3 = dementia). An exploratory ADI × dementia interaction was tested. Estimated marginal means and contrasts were extracted to interpret ADI effects within each dementia stratum. Random intercepts for participant and site were included.
Result
A significant ADI × dementia interaction was observed for global atrophy (p = .005) and several other outcomes. Among participants in the 30% most disadvantaged group, dementia was associated with greater brain volume differences compared to no dementia (Table 2). Dementia was linked to greater global atrophy (Δ = +0.0502, 95% CI [0.0276, 0.0727]), lower total brain volume (Δ = −0.0435), middle temporal gyrus (Δ = −0.0009), and frontal pole (Δ = −0.0028), all p < .0001. A marginal trend was observed for parahippocampal volume (p = .074); no difference was found for hippocampus.
Conclusion
Dementia was associated with greater global atrophy and lower brain volumes among disadvantaged individuals. These findings suggest that socioeconomic disadvantage may amplify dementia‐related brain changes. The small number of participants with both dementia and high disadvantage (n = 14) is a limitation, underscoring the need for greater socioeconomic representation in future studies.
