Medial orbitofrontal cortex gray matter is reduced in abstinent substance-dependent individuals

Abstract

Background: Chronic exposure to drugs of addiction induces cellular adaptations in orbitofrontal cortex (OFC) and associated limbic-prefrontal pathways that might underlie abuse-related behavior. A propensity to make risky decisions in spite of substantial negative consequences might be mediated by medial OFC dysfunction in substance-dependent individuals (SDI). We tested the hypothesis that medial OFC gray matter (GM) volume would be lower in SDI compared with control subjects.

Methods: Nineteen SDI and 20 control subjects participated. The SDI were dependent on two or more substances, most often cocaine, amphetamine, and alcohol, with mean duration of abstinence 4.7, 2.4, and 3.2 years, respectively. High-resolution T1-weighted images were acquired on a 3-T magnetic resonance system. Image processing and analyses were conducted with voxel-based morphometry (VBM) implemented in Statistical Parametric Mapping (SPM) 5. Differences in regional GM volume were tested with an analysis of covariance model, co-varying for global GM and age. Statistical maps were set at p < .05, corrected for multiple comparisons. Medial OFC GM volume was correlated with behavioral performance on a modified gambling task.

Results: There was lower GM volume specifically in bilateral medial OFC in SDI compared with control subjects. There was a small but significant correlation between medial OFC GM and persistence of playing high-risk decks on a modified gambling task.

Conclusions: This is the first study to use VBM with whole brain correction for multiple comparisons in SDI after prolonged abstinence. Reduced medial OFC GM might reflect long-term adaptations within the reward-learning circuit underlying pathological decision-making in substance dependence.

Figure 1

Color map and glass brain showing increased gray matter in OFC in controls compared to substance dependent individuals (SDI), after co-varying for total GM and age (threshold p<.05, cluster-level, corrected for multiple comparisons family-wise error, voxel level, p<.005). Color bar represents t-values. Color map is overlaid on canonical avg152T1 template.

Figure 2

Estimated marginal mean of “bad” cards played over time for SDI and controls, adjusted for education, IQ, and age. Over time controls played fewer “bad” cards than SDI, but the group by time interaction was not significant due to high variability. Mean and SEM are shown.

Figure 3

Scatterplot of medial OFC gray matter volume (at -5, 53, -3), adjusted for total GM and age, and persistence in playing “bad” cards. A significant negative correlation was observed (r=-.39, p=.01 without controlling for IQ and education)(r=-.35, p=.03, after controlling for IQ and education).