Abstract
Objective: The high comorbidity between posttraumatic stress disorder (PTSD) and alcohol use disorder (AUD) is well-established and complex. However, there is a need to explore transdiagnostic constructs that may underlie this association to better understand what accounts for this comorbidity and to inform treatment development. Method: Thus, the present study utilized a large, cross-sectional dataset (N = 513; Mage = 38.25 years, SD = 10.07; 49.9% female), based on national recruitment, to (1) examine whether the associations between PTSD symptom severity (PCL-5) and alcohol use severity (AUDIT) were statistically mediated by (a) anxiety sensitivity (SSASI); and (b) difficulties with emotion regulation (DERS-16); and (2) examine whether coping motives for drinking moderate this indirect effect. Sex assigned at birth was included as a covariate. Results: When examining the hypothesized mediators (SSASI and DERS-16) in separate models, there was a statistically significant indirect effect of PCL-5 on AUDIT through both SSASI and DERS. However, when both SSASI and DERS were entered into a model simultaneously, only SSASI served as a statistically significant mediator. Coping motives for drinking did not moderate the observed indirect effect. Conclusions: The current findings highlight anxiety sensitivity and emotion regulation as transdiagnostic processes that may explain, at least partially, the relationship between PTSD symptom severity and alcohol use; however, stronger evidence was evident for anxiety sensitivity. These findings may help inform the development of refined, streamlined interventions for PTSD and alcohol use that directly target these processes.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 In all analyses, with regard to the effect of the covariate, there was no significant effect of sex on any of the mediators (ps = .13), but female sex was associated with lower AUDIT scores in both the direct effects and total effects models (ps < .001). This is reported once here to reduce redundancy. Note that parallel analyses were conducted with and without the covariate and the pattern of findings was identical.