Reduced Cognitive-affective Network Connectivity in Patients at Risk for Opioid Use Disorder

Introduction: Chronic pain, depression, and polysubstance use often co-exist and elevate risk for opioid use disorder (OUD). Previous electrophysiological studies report group differences in gamma-band power in substance use disorder (SUD), but whether gamma activity supports communication within cognitive–affective circuits remain unclear. We therefore compared gamma-band functional connectivity between individuals at high risk for OUD (HR) and low-risk controls (LR), hypothesizing reduced connectivity in HR.

Methods: We analyzed intracranial local field potentials (200 Hz) from 33 epilepsy patients with depth electrodes covering the amygdala (n=21), hippocampus (n=31), dorsolateral prefrontal cortex (DLPFC, n=12), anterior cingulate cortex (ACC, n=12), orbitofrontal cortex (OFC, n=20), insula (n=9), medial prefrontal cortex (MPFC, n=1), ventrolateral prefrontal cortex (VLPFC, n=17), and parietal lobule (n=8). Fifteen-minute pre/post drug (opioid or acetaminophen) segments were band-pass filtered in the gamma band (30–45 Hz). The HR group (n=23) was compared with the LR group (n=10). Functional connectivity was estimated via cross-correlation of gamma-band amplitude envelopes between region pairs. Differences in pairwise connectivity between groups were assessed with Kruskal-Wallis H-tests (Bonferroni-corrected). Graph density (proportion of suprathreshold edges) was computed and compared between groups using the same permutation procedure, with prespecified subgroup analyses contrasting chronic pain (CP), substance-use (SU), depression (D) subgroups against LR group.

Results: Pairwise analyses showed reduced connectivity in the high-risk (HR) group across cognitive–affective regions (Fig. A–B). Network density was also lower in HR than in the low-risk (LR) group (p < 0.01; Fig. C). Stratifying HR revealed heterogeneity: the chronic pain (CP) subgroup did not differ from LR, whereas the depression (D) and substance-use (SU) subgroups showed lower connectivity (p < 0.01), with SU exhibiting the lowest connectivity relative to LR.

Conclusion: These results indicate that individuals at high risk for OUD have weaker gamma-band connectivity within cognitive–affective circuits. The reduction is observed mainly by the substance-use (SU) subgroup, whereas the chronic pain (CP) subgroup resembles low-risk controls. Consistent pairwise connectivity and network-density findings support using connectivity profiles to characterize OUD risk and to guide targeted neuromodulation and pain-management strategies tailored to subgroup-specific network patterns.