Constructs derived from the addiction cycle predict alcohol use disorder treatment outcomes and recovery 3 years following treatment.

Abstract

Objective: The addiction cycle has been proposed as a framework for understanding the progression of alcohol use disorder (AUD) in terms of psychological and biological domains, including reward drinking/incentive salience, relief drinking/negative emotionality, and loss of control/executive functioning impairment. To have utility in clinical practice, self-report measures of these domains that are applicable across sociodemographic groups and associated with clinical outcomes are needed. This study sought to validate domains from self-report measures and to test whether domains are measurement invariant across sociodemographic groups and associated with treatment outcomes. Method: Secondary analysis of individuals with AUD (n = 3,092) who participated in two alcohol clinical trials, Project Matching Alcohol Treatment to Client Heterogeneity (MATCH) and COMBINE. Factor analytic methods were used to derive addiction cycle domains at baseline. These domains were then examined as predictors of outcomes. Results: Fifteen self-report items were used as indicators of the addiction cycle domains, with sociodemographic differences in measurement by sex, age, race, education, and AUD symptoms. Relief/negative emotionality and reward/incentive salience were significantly associated with outcomes at 1 and 3 years following treatment, and executive functioning also predicted nonabstinent recovery at 3 years. Conclusions: The results support the utility of domains relevant to the addiction cycle in predicting AUD treatment outcomes and recovery among individuals who sought treatment for AUD. The addiction cycle domains were more strongly associated with outcomes than other measures clinicians might use to predict outcomes (e.g., AUD symptoms). Future research should continue to develop and refine the items and test whether the addiction cycle domains can inform treatment planning.

Examining the Validity of the Addictions Neuroclinical Assessment Domains in a Crowdsourced Sample of Adults with Current Alcohol Use

“…In fact, by blurring the distinction between dispositional and acquired features, we might attenuate observed associations between negative emotionality and drinking-related outcomes, given that the link between dispositional negative emotionality and alcohol use disorder is relatively weak (Hussong et al, 2011). Only one study has examined the ANA domains using alcohol-specific indicators, and findings showed the utility of these domains for predicting treatment outcomes one and three years following treatment (Witkiewitz, Stein, et al, 2022). An essential step in refining ANA measurement will be examining whether alcohol-specific, specifically acquired, or general indicators of the three ANA domains are more valid or meaningful.…”

“…Further, although self-report indicators increase the feasibility of the ANA battery in clinical practice by reducing the burden on clinicians and clients, the indicators evaluated thus far have several limitations that hinder interpretability (Boness et al, 2021;Witkiewitz, Stein, et al, 2022). Previous studies have used a wide range of indicators, including many general or trait measures of reward/incentive salience, negative emotionality, and cognitive control/executive function that were not designed to assess core AUD mechanisms and do not reflect the intention of the ANA to measure constructs reflecting the addiction cycle (Boness et al, 2021;Witkiewitz, Stein, et al, 2022).…”

“…Further, although self-report indicators increase the feasibility of the ANA battery in clinical practice by reducing the burden on clinicians and clients, the indicators evaluated thus far have several limitations that hinder interpretability (Boness et al, 2021;Witkiewitz, Stein, et al, 2022). Previous studies have used a wide range of indicators, including many general or trait measures of reward/incentive salience, negative emotionality, and cognitive control/executive function that were not designed to assess core AUD mechanisms and do not reflect the intention of the ANA to measure constructs reflecting the addiction cycle (Boness et al, 2021;Witkiewitz, Stein, et al, 2022). For example, many studies (including Kwako et al, 2016) rely on measures of dispositional negative emotionality even though Koob and LeMoal's (1997) modelwhich, again, forms the basis of ANAcenters around negative emotionality that is acquired throughout the addiction cycle, arising along with withdrawal from a shift in hedonic tone attributable to allostatic burden.…”

“…Thus far, research has examined the factor structure and construct validity of the ANA domains in clinical samples diagnosed with AUD (Stein et al, 2021;Votaw et al, 2020;Witkiewitz, Stein, et al, 2022), combined treatment-seeking and non-treatment-seeking samples of people with and without diagnoses of AUD (Kwako et al, 2019;Nieto et al, 2021), nontreatment-seekers who consume alcohol (DeMartini et al, 2021), and individuals who use methamphetamine (Nieto & Ray, 2022). Although these studies have been interpreted as supporting the three-factor structure of the ANA domains, select results are perplexing concerning the interpretation of factors.…”

“…For example, Kwako and colleagues (2019) identified extraversion as a statistically-significant indicator of the negative emotionality domain (Boness et al, 2021). Nevertheless, there is initial evidence that ANA domains, particularly negative emotionality, may predict AUD treatment outcomes over time (Votaw et al, 2021;Witkiewitz, Stein, et al, 2022). Also, more limited research has examined measurement invariance of the ANA negative emotionality and reward/incentive salience domains across gender and time (Stein et al, 2021;Votaw et al, 2020), which is critical to establish whether the ANA captures the same constructs across genders and over time.…”