Structural Brain Changes Associated With Risky Drinking in Late-Life Depression

Abstract

Objectives: As alcohol use is common among older depressives, we assessed structural brain changes over 2 years and examined their association with changes in alcohol consumption.

Design: Longitudinal cohort study.

Setting: Academic health center.

Participants: Adults aged 60 and older who met DSM criteria for a major depressive episode.

Intervention: Participants were offered treatment with sertraline.

Measurements: Participants completed structured interviews for reported alcohol consumption, had a clinical interview with a study psychiatrist, completed a cognitive battery at baseline and every twelve months, and underwent a 3T structural MRI as baseline and at 2-year follow-up. Volumetric brain changes were calculated.

Results: Among 58 participants, 45 were classified as moderate drinkers (≤7 drinks/week) and 13 as risky drinkers (>7 drinks/week). Compared with moderate drinkers, risky drinkers at baseline had significantly thinner cortical thickness and smaller volume in several frontal cortical regions, posterior cingulate, postcentral cortices, right insula, right putamen, and right inferior parietal sulcus. Annualized change in cortical thickness and volume correlated negatively with changes in the average number of drinks per week. Decreased depression severity, increased cognitive function score, and decreased alcohol consumption over the 2-year follow-up were each associated with annualized volumetric changes in specific common regions.

Conclusion: These MRI findings demonstrate the adverse impact of alcohol use in older adults on the fronto-striatal circuit. They highlight the need for careful screening and treatment referral for risky alcohol use among older adults with depression.

Keywords: Alcohol; depression; neuroimaging; risky drinking.

FIGURE 1.

[A] Whole-brain two-sample t-test (corrected using FDR for multiple comparisons) results on cortical thickness at baseline showing significantly reduced cortical thickness in the risky drinking group compared with the moderate drinking group in bilateral medial prefrontal cortex (mPFC), left superior frontal cortex (frontal eye area), superior parietal cortex, and right temporal pole. [B] The scatter plots of brain regions that showed significant differences between risky drinking group and moderate drinking group using region of interests (ROI) analysis on regional brain volume/total intracranial volume ratio. PCC = posterior cingulate cortex, IPC = inferior parietal cortex.

FIGURE 2.

[A] Significant correlations between the change in average number of drinks per week over 2 years and the annualized changes in cortical thickness from whole-brain voxelwise analysis. Blue color = negative correlation, Red color = positive correlation [B] ROI analysis revealing significant correlation between the change in average number (or maximum number) of drinks per week over 2 years and the annualized changes in atlas-based regional volume.

FIGURE 3.

Comparison of brain regions that their annualized changes were significantly correlated with changes in alcohol consumption (measured by the average number of drinks per week over 2 years), depression severity (measured by MADRS), cognitive function (measured CERAD), and age at baseline. Blue color = negative correlation, Red color = positive correlation.

FIGURE 4.

Conjunction analysis reveals the brain regions that their changes in cortical thickness were commonly correlated with changes in the average number of drinks per week, depression severity, and cognitive function [A], as well as with age [B] with p <0.025 without correction for multiple comparisons.