Exploring neural markers of incentive salience and real-world drinking among individuals with alcohol use disorder

• Two-week EMA plus alcohol biosensor monitoring • Laboratory EEG visual oddball with standard, alcohol and non-alcohol images. • Alcohol cues elicit larger P3b than non-alcohol ones • Effect concentrated in moderate–severe Alcohol Use Disorder (AUD) participants. • Binge-drinking and peak eBAC amplify alcohol-specific P3b. Alcohol cue salience is theorized to play a mechanistic role in alcohol use disorder (AUD), yet links between neural cue reactivity and naturalistic drinking remain undercharacterized. This study combined laboratory event-related potential (ERP) measures with two weeks of ambulatory assessment to evaluate whether alcohol-cue P3b relates to real-world drinking and individual differences in AUD severity. Heavy drinking participants (52% Female; Ages 21-32) were recruited from the local community. Participants completed two weeks of ecological momentary assessment with continuous transdermal alcohol monitoring and attended three laboratory visits scheduled at one-week intervals. At the final study visit, participants completed an EEG visual oddball task involving the presentation of both infrequent alcohol and non-alcohol beverage target images and frequent household-object standards. Alcohol images elicited significantly larger P3b amplitudes than non-alcohol images across the sample ( N = 47), b = 2.13, p = .002. Critically, this alcohol-specific P3b enhancement was concentrated among individuals with pronounced AUD (moderate–severe, N =20). Objective transdermally-measured ambulatory drinking further moderated neural cue reactivity in the pronounced AUD group: more binge-level days, b =0.51, p =.020, and higher peak estimated consumption, b =34.09, p =.044, were associated with stronger alcohol-specific P3b responses. In contrast, neither retrospective baseline nor in-vivo ambulatory self-reports of drinking demonstrated consistent associations. Together, findings indicate that alcohol cue–elicited P3b is (a) sensitive to clinically meaningful severity distinctions and (b) larger among individuals with heavier real-world drinking as captured with objective sensors, supporting its utility as a neurocognitive marker with ecological validity for understanding individual differences in AUD.