Neurophysiological affective reactivity classification and its stability in individuals with cannabis use disorder.

Schermitzler B.; Preston T.; Cannon M.; Gorday J.; Solan S.; Moeller S.; Macatee R. · International journal of psychophysiology : official journal of the International Organization of Psychophysiology · 2026

**Background & Methods** This study examined neurophysiological markers of affective reactivity in 44 individuals with Cannabis Use Disorder (CUD) using the late positive potential (LPP), an event-related potential measured via EEG. Participants completed two identical lab visits one week apart, during which they performed a probabilistic choice task and viewed cannabis-related and affective pictures while EEG was recorded. K-means cluster analysis was applied to LPP amplitudes to classify participants into neuroaffective reactivity profiles based on their neural responses to different stimulus types. **Key Findings** • Participants were reliably classified into two distinct neuroaffective reactivity clusters at both timepoints: one group showing greater neural reactivity (larger LPP amplitude) to cannabis cues compared to pleasant images, and another group showing the opposite pattern (greater reactivity to pleasant images). This two-cluster classification was consistent with previous findings in other substance use populations. • Despite reliable group-level classification patterns, individual-level retest reliability over one week was low, meaning that specific individuals could not be reliably assigned to the same cluster across visits. Residualized and raw LPP scores performed better than cluster membership for tracking individual changes. • LPP clusters did not differentiate participants on choice behavior outcomes, problematic use indicators, or other clinical markers, suggesting limited convergent validity between the neuroaffective profiles and behavioral/clinical measures despite their theoretical relevance. **Dosage & Administration** Not reported. **Safety & Adverse Effects** Not reported. **Evidence Quality** This study is limited by its small sample size (n=44) and short retest interval (one week), which may not capture longer-term stability. The lack of association between LPP profiles and behavioral outcomes raises questions about clinical utility. The findings suggest that while group-level neuroaffective reactivity patterns appear consistent across addiction populations, individual-level reliability remains problematic for clinical application. The authors acknowledge the gap between theoretical predictions (based on incentive sensitization model) and empirical associations with actual choice behavior, indicating that LPP clustering alone may be insufficient for predicting CUD outcomes without additional biomarkers or behavioral measures.

More recent cannabis research