The reciprocal relationship between posttraumatic stress disorder symptoms and alcohol use in a large multisite...

Posttraumatic stress disorder (PTSD) and alcohol use disorder (AUD) often co-occur. There is a lack of longitudinal studies measuring the naturalistic development of PTSD and alcohol use problems in individuals with recent trauma exposure. This study aimed to compare the temporal relationships between posttraumatic stress symptoms and alcohol use over 6 months following trauma exposure in males and females.

Large-scale longitudinal observational emergency department (ED)-based study of individuals with recent trauma exposure. Individuals with recent trauma exposure (n = 2942, 62% female) were recruited from 29 EDs across the United States within 72 h of trauma exposure from 2017 to 2021. PTSD symptoms, measured via the PTSD Checklist for the DSM-5, and alcohol use measured via the PhenX toolkit, were assessed at five time points: ED visit, 2 weeks, 8 weeks, 3 months, and 6 months following trauma.

PTSD symptoms predicted lower future alcohol use between the pretrauma to two-week time points (b = -0.08, p = 0.01) and higher use between the 3- to 6-month time points (b = 0.06, p = 0.01). There were no time points in which alcohol use predicted future PTSD symptoms. When stratifying by sex, male participants showed reciprocal associations, with alcohol use early after trauma predicting PTSD symptoms between 2 and 8 weeks (b = 0.08, p = 0.01), while PTSD symptoms predicted alcohol use between the 3- to 6-month time points (b = 0.10, p = 0.01). Female participants showed a different reciprocal pattern, with pretrauma PTSD symptoms predicting lower alcohol use 2 weeks posttrauma (b = -0.08, p = 0.04), while alcohol use subsequently predicted greater PTSD symptoms from 8 weeks to 3 months (b = 0.04, p = 0.04); these findings did not survive Bonferroni correction.

Males and females exhibit complex temporal development patterns of PTSD symptoms and alcohol use that align with the mutual maintenance hypothesis in males but the susceptibility hypothesis in females. These patterns are masked in analyses that do not stratify by sex.