Structure and predictive validity of alcohol use disorder recovery domains in young adulthood

The alcohol use disorder (AUD) recovery construct has been expanded beyond remission to changes across an array of biopsychosocial domains. Drawing on expanded AUD recovery definitions, researchers have developed numerous psychosocial instruments to measure and evaluate the validity of the recovery construct. However, research examining whether the meaning of AUD recovery is invariant by disorder status and treatment use is currently scant, as is longitudinal research examining whether recovery domains demonstrate predictive validity via change in AUD symptoms. To address these gaps, the current uses young adult data from the National Longitudinal Study of Adolescent to Young Adult Health (n = 4,693) to test the structure and predictive validity of AUD recovery domains. Using exploratory structural equation modeling, we found a seven-factor structure­—psychological health, social support and connection, risky behavior and violence, physical health and daily activity, religiosity, socioeconomic status, and alcohol use behaviors—that fit the data well (χ2 = 4,452.53 [1,692], p < .001; CFI = .98; TLI = .97; RMSEA = .02[.020-.021] and appeared largely invariant between those who did and did not meet the AUD criteria, as well as between those who did and did not receive treatment or other supports. Controlling for baseline AUD severity, demographics, and use of treatment, we found that several recovery domains predicted rank-order change in alcohol use and abuse indicators that occurred across seven years, with religiosity emerging as one of the most consistent predictors of change. Effects on later symptoms of alcohol dependence, however, were trivial or non-significant. These findings provide new insight into the AUD recovery process and hold important implications for AUD recovery measurement and assessment.