Functional hyperconnectivity between corticocerebellar networks and altered decision making in young adult cannabis users: Evidence from 7T and multivariate pattern analysis.

Cannabis users show deficits in decision-making that are linked to hyperconnectivity in corticocerebellar brain networks, particularly between the posterior cerebellum and regions involved in reward and self-control. In a study comparing 92 cannabis users with 92 non-users using high-resolution 7 Tesla MRI, users scored higher on the Iowa Gambling Task percentage of larger rewards and lower on percentage of smaller rewards, and had larger left nucleus accumbens volume. Resting-state functional connectivity was enhanced between the left cerebellum region 9 and the right nucleus accumbens, left pallidum, and left putamen, as well as between the posterior cerebellum and default mode network nodes. White matter microstructure showed no significant differences. These findings suggest that altered cerebellar connectivity may underlie decision-making problems in cannabis users, potentially contributing to difficulty quitting or relapse.

Study at a glance

Abstract

Decision-making (DM) impairments are important predictors of cannabis initiation and continued use. In cannabis users, how decision-making abnormalities related to structural and functional connectivity in the brain are not fully understood. We employed a three-method multimodal image analysis and multivariate pattern analysis (MVPA) on high dimensional 7 tesla MRI images examining functional connectivity, white matter microstructure and gray matter volume in a group of cannabis users and non-users. Neuroimaging and cognitive analyses were performed on 92 CU and 92 age- matched NU from a total of 187 7T scans. CU were selected on the basis of their scores on the Semi-Structured Assessment for the Genetics of Alcoholism. The groups were first compared on a decision-making test and then on ICA based functional connectivity between corticocerebellar networks. An MVPA was done as a confirmatory analysis. The anatomy of these networks was then assessed using Diffusion Tensor imaging (DTI) and cortical volume analyses. Cannabis Users had significantly higher scores on the Iowa Gambling task (IGT) [Gambling task Percentage larger] and significantly lower scores on the [Gambling task reward Percentage smaller]. Left accumbens (L NAc) volume was significantly larger in cannabis users. DTI analysis between the groups yielded no significant (FWE corrected) differences. Resting state FC analysis of the left Cerebellum region 9 showed enhanced functional connectivity with the right nucleus accumbens and left pallidum and left putamen in CU. In addition, posterior cerebellum showed enhanced functional connectivity (FWE corrected) with 2 nodes of the DMN and left and right paracingulate (sub genual ACC) and the sub callosal cortex in CU. IGT percentage larger scores correlated with posterior cerebellar functional connectivity in non-user women. A multivariate pattern analysis confirmed this cerebellar hyperconnectivity in both groups. Our results demonstrate for the first time that deficits in DM observed in cannabis users are mirrored in hyper connectivity in corticocerebellar networks. Cortical volumes of some of the nodes of these networks showed increases in users. However, the underlying white matter was largely intact in CU. The observed DM deficits and hyper connectivity in resting networks may contribute to difficulties in quitting and/or facilitating relapse.