Category: Addictive Behaviors

PS1- #A2 - Influence of Anxiety Sensitivity on Opioid Use Disorder Treatment Outcomes

Friday, Nov 17
8:30 AM – 9:30 AM
Location: Indigo Ballroom CDGH, Level 2, Indigo Level

Keywords: Substance Abuse | Anxiety Sensitivity | Relapse

For some, treatment for opioid use disorder (OUD) begins with detoxification followed by additional psychosocial treatment. Unfortunately, relapse is common and engagement in follow-up treatment is poor. One factor that may impact these outcomes is anxiety sensitivity (AS), the fear of anxiety and anxiety-related sensations. The present study examined AS within the context of OUD detoxification and as a predictor of treatment engagement and relapse. It was hypothesized that AS would be associated with a more negative experience of detoxification and increased risk of poor treatment engagement and relapse one month later.


Participants (N = 90; 45.6% female, M age [SD] = 37.5 [10.3]; 58.9% Caucasian, 35.6% African American, 5.5% other) underwent a five-day inpatient opioid detoxification program via buprenorphine. On day 4, participants completed the Anxiety Sensitivity Index–3, three visual analog scales for self-reported craving, withdrawal severity, and fear of withdrawal, and other measures of clinical interest. About 30 days later participants completed a follow-up assessment regarding treatment engagement and relapse. The follow-up rate was 84.4% (n = 76).


Sample mean on the ASI-3 was 21.3 (14.8) and about 46.7% of the sample was above the clinical cutoff. Via separate hierarchical regressions, AS was found to significantly predict withdrawal severity and fear of withdrawal after controlling for prior OUD treatments, ps < .05, and accounted for 10% and 18% of variance, respectively. Contrary to our hypothesis, AS did not significantly predict opioid craving after controlling for state anxiety, p > .05. Further, a hierarchical logistic regression (n = 53) was conducted to examine whether AS predicted treatment engagement, excluding participants who were immediately discharged into a residential program. The number of prior treatments was entered at step one and AS at step two. This overall model was not significant, ps >.05. Concerning AS predicting relapse, a hierarchical regression was conducted with opioid craving and number of days free in the community entered at step one and AS at step two. The overall model was significant, F(3, 72) = 7.27, p < .001, and explained 23.2% of the variance in the number of days using opioids. However, only craving, b = .24, t(75) = 2.28, p = .026, and days free in the community, b = .40, t(75) = 3.82, p < .001, predicted the outcome. AS was not significant in the model, p > .05. When relapse was coded dichotomously, a hierarchical logistic regression using the same predictors was significant; however, only days in the community was associated with increased odds of opioid relapse (OR = 1.10, 95% CI 1.04 – 1.17), p = .002.


Overall, AS predicted greater self-reported opioid withdrawal severity and fear of withdrawal during detoxification. However, only state anxiety predicted opioid craving during detoxification. While AS did not significantly predict relapse, opioid craving during detoxification and days free in the community predicted relapse. Notably, AS was not a significant predictor of treatment engagement. Findings suggest AS may lead to a more difficult experience during detoxification, though more work is needed to further understand the underlying processes of treatment engagement and opioid relapse.

Catherine Baxley

Graduate Student
Saint Louis University
St. Louis, Missouri

Jeremiah Weinstock

Associate Professor
Saint Louis University
Saint Louis, Missouri

Annie A. Garner

Assistant Professor
St. Louis University
Saint Louis, Missouri

Alexandra D’Agostino

Saint Louis University

Desirae Rowan

Saint Louis University

Brian Woods

Saint Louis University