Reduction of HIV risk behavior with transcranial magnetic stimulation

National Institute On Drug Abuse

Investigators

Abstract

1. Cocaine Use Impacts HIV-Related Decision-Making Deficits through Orbitofrontal Cortex and Nucleus Accumbens Alterations While HIV infection and cocaine use disorder (CUD) independently affect reward processing circuits, their combined effects on decision-making and underlying neural mechanisms remain unclear. Here, we investigated whether cocaine use impacts HIV-related decision-making impairments through structural alterations in reward processing brain regions. This study is a collaboration with groups at University of Pittsburgh School of Medicine and University of Illinois at Chicago School of Medicine. Ninety-nine participants were enrolled in four groups, including participants with HIV infection and concurrent CUD (n=27), participants with HIV infection only (n=20), participants with CUD only (n=26), and healthy controls (n=26). Mean age was 42.7 years (SD=9.5). HIV-positive participants had been infected for an average of 17 years. CUD was diagnosed based via the Structured Clinical Interview for DSM-IV-TR, and cocaine use history was assessed using the Kreek-McHugh-Schluger-Kellogg (KMSK) scale, with peak use scores ranging 0-16. Participants completed the Iowa Gambling Task (IGT) and structural MRI scanning. Decision-making parameters were derived using the Value Plus Persistence computational model through hierarchical Bayesian estimation. Structural metrics were extracted using FreeSurfer 7.3.2, focusing on orbitofrontal cortex (OFC) and nucleus accumbens (NAcc), regions critically involved in reward and decision-making. A two-way ANOVA revealed significant HIV×CUD interactions for learning rate and outcome sensitivity, as well as significant HIV and CUD main effects for these IGT parameters. Learning rate represents sensitivity to recent outcomes, while outcome sensitivity reflects exploration-exploitation balance. Groups of participants with HIV infection only, with CUD only, and with HIV infection and concurrent CUD showed significantly lower learning rate and lower outcome sensitivity than controls. A two-way ANCOVA controlling for total intracranial volume revealed significant effects in reward-related brain regions. The right lateral OFC surface area showed a significant CUD main effect, while the left lateral OFC showed an HIV×CUD interaction. Additionally, the right lateral OFC folding complexity demonstrated a significant CUD effect. For the NAcc volume, the left NAcc showed both significant HIV and CUD main effects, while the right NAcc showed a significant CUD effect and HIV×CUD interaction. Importantly, mediation analyses revealed three structural pathways through which peak cocaine use impacted decision-making. First, peak cocaine use was associated with lower right lateral OFC surface area, which predicted lower learning rate, accounting for 37.05% mediation. Second, peak cocaine use was associated with lower OFC folding complexity, which was also associated with impaired learning, yielding 49.01% mediation. Third, peak cocaine use was linked to the left NAcc volume, which significantly predicted lower learning rate, explaining 52.15% of the mediation effect.