Comparing the Psychometric Properties of Reward and Relief Drinking Measures

Abstract

Previous work examining the extent to which individuals seek alcohol to enhance positive experiences (reward drinking) or relieve aversive states (relief drinking) has shown that reward/relief drinking predicts response to naltrexone and acamprosate treatment for alcohol use disorder (AUD). Yet, various measures of reward/relief drinking have been used in prior research, and the comparative psychometric properties of these measures are unknown. Evaluating and comparing the psychometric properties of these reward/relief drinking measures could identify measures with the most promise for translating precision medicine findings to clinical practice. In a community sample of 65 individuals with heavy/hazardous alcohol use on the Alcohol Use Disorder Identification Test, we showed good internal consistency reliability, test-retest reliability, and concurrent validity for theoretically aligned measures (e.g., reward drinking and reward responsiveness, relief drinking and depression/anxiety symptoms) of the reward and relief subscales across the six measures. We then used ecological momentary assessment (EMA) to determine whether reward and relief drinking subscales predicted within-person associations between contextual factors of interest (e.g., negative affect, positive affect, distress intolerance, physical pain, hangover symptoms, social drinking situations, alcohol cues) and same-moment alcohol craving. All six measures demonstrated limited predictive validity for alcohol craving contexts in daily life assessed via EMA. Despite these findings, reward and relief drinking measures show good reliability and concurrent validity and previously demonstrated clinical utility for predicting outcomes of naltrexone response and other AUD treatments. Future research should aim to elucidate the mechanisms underlying the association between responses to reward/relief drinking measures and pharmacotherapy response.