Common Dimensional Reward Deficits Across Mood and Psychotic Disorders: A Connectome-Wide Association Study

Abstract

Objective: Anhedonia is central to multiple psychiatric disorders and causes substantial disability. A dimensional conceptualization posits that anhedonia severity is related to a transdiagnostic continuum of reward deficits in specific neural networks. Previous functional connectivity studies related to anhedonia have focused on case-control comparisons in specific disorders, using region-specific seed-based analyses. Here, the authors explore the entire functional connectome in relation to reward responsivity across a population of adults with heterogeneous psychopathology.

Method: In a sample of 225 adults from five diagnostic groups (major depressive disorder, N=32; bipolar disorder, N=50; schizophrenia, N=51; psychosis risk, N=39; and healthy control subjects, N=53), the authors conducted a connectome-wide analysis examining the relationship between a dimensional measure of reward responsivity (the reward sensitivity subscale of the Behavioral Activation Scale) and resting-state functional connectivity using multivariate distance-based matrix regression.

Results: The authors identified foci of dysconnectivity associated with reward responsivity in the nucleus accumbens, the default mode network, and the cingulo-opercular network. Follow-up analyses revealed dysconnectivity among specific large-scale functional networks and their connectivity with the nucleus accumbens. Reward deficits were associated with decreased connectivity between the nucleus accumbens and the default mode network and increased connectivity between the nucleus accumbens and the cingulo-opercular network. In addition, impaired reward responsivity was associated with default mode network hyperconnectivity and diminished connectivity between the default mode network and the cingulo-opercular network.

Conclusions: These results emphasize the centrality of the nucleus accumbens in the pathophysiology of reward deficits and suggest that dissociable patterns of connectivity among large-scale networks are critical to the neurobiology of reward dysfunction across clinical diagnostic categories.

Keywords: Anhedonia; Psychiatry; Resting State fMRI; Reward.

Figure 1

Connectome-wide analysis using multivariate distance-based matrix regression (MDMR). Template-space functional time series were resampled at 4mm³ for computational feasibility. For each gray matter voxel a connectivity map was created for each subject, which were compared in a pairwise manner to create a distance matrix. MDMR used these distance matrices to evaluate the complex multivariate pattern of connectivity in association with reward responsivity (BAS Reward) across subjects while controlling for clinical group, age, sex and in-scanner motion. This yielded a pseudo-F statistic and a corresponding P value through permutation testing. This procedure was repeated for each gray matter voxel, yielding a voxel-wise significance map. Modified with permission from Shehzad et al. (16).

Figure 2

MDMR identified multiple foci of connectivity related to reward responsivity. Cortical projection displaying clusters identified by multivariate distance-based matrix regression (MDMR) where BAS Reward affected the overall multivariate pattern of functional connectivity. All clusters corrected for multiple comparisons at z>1.64, P<0.01.

Figure 3

Follow-up seed-based connectivity analyses explain patterns of connectivity that drive MDMR results. The multivariate results of the connectome-wide association study (CWAS) identified the nucleus accumbens (A), default mode regions (B) and cingulo-opercular regions (C) where the overall pattern of connectivity is related to reward responsivity, but did not delineate the nature of those patterns. Accordingly, each cluster identified by CWAS (left column) was used as a seed to identify what changes in connectivity led to the significant finding. The middle column displays the mean connectivity across all subjects from each seed. The right column displays the association with BAS Reward for each seed.

Figure 4

Dissociable patterns of within and between-network connectivity dimensionally relate to reward deficits across psychiatric disorders. Layout of mean connectivity within a network of nodes defined by multivariate distance-based matrix regression (MDMR) (Panel A). DMN hyperconnectivity and decreased connectivity between DMN and CON regions scales with reward deficits (Panel B). Cortical projection displaying the NAc (green) along with DMN (blue) and CON regions (red) (Panel C). Reward deficits are related to dissociable patterns of NAc dysconnectivity, including diminished connectivity with the DMN and increased connectivity with the CON (Panel D). DMN=default mode network; CON=cingulo-opercular network; NAc= nucleus accumbens.