Perceptions of negative reinforcement in the natural environment: Characterizing subjective alcohol- and...

Negative reinforcement is a major component of etiologic models of addiction. Recent work using ecological momentary assessment (EMA) implicates subjective substance-contingent relief as a negative reinforcer of substance use rather than improvements in negative affect. This pre-registered secondary data analysis tested predictors of subjective alcohol- and cannabis-contingent relief to understand what factors contribute to perceived relief, if not actual improvements in negative affect. Adults (N = 87, mean age 25.2 years, 60.2% female, 85.2% white) who reported using alcohol and cannabis simultaneously at least twice per week completed 14 days of EMA. Morning surveys assessed hypothetical relief expectancies for that day. All surveys assessed affect, craving, alcohol and cannabis use, and subjective relief and pleasure from use. Multilevel models tested predictors of subjective relief. During alcohol use moments (n = 868), drinking-contingent pleasure (b = 0.25, SE = 0.06, p < .001) and relief expectancy reported that morning (b = 0.29, SE = 0.10, p = .005) predicted higher drinking-contingent relief. During cannabis use moments (n = 1519), cannabis-contingent pleasure (b = 0.29, SE = 0.06, p < .001) and relief expectancy reported that morning (b = 0.33, SE = 0.11, p = .004) predicted higher cannabis-contingent relief. Changes in affect and craving were not associated with relief. Results highlight subjective pleasure, which is more consistent with positive reinforcement, as a potential driver of perceived relief. Future work should disentangle the predictive direction of relief and pleasure to inform reinforcement mechanisms. Results also highlight the role of expectancies in subjective perceptions of use, above and beyond measurable changes in mood and craving. Collectively, findings add to a growing body of work seeking to articulate reinforcement processes in the natural environment.