Method: A random sample of 230 incarcerated women was obtained from two state prisons. Participants completed the Childhood Trauma Questionnaire, the Mini International Neuropsychiatric Interview, and the Substance Abuse Module-Revised. Psychosis was defined as audio/visual hallucinations and/or delusions not directly related to drug use. Binary logistic regression models assessed the relationship between frequency of marijuana and crack-cocaine use in the year prior to incarceration on current (past 30 days) psychosis. For each regression model we examined the following: case-to-predictor variable ratio, multicollinearity, model fit, and relationships between predictor variables and the dependent variable. Psychosis was a dichotomous variable in all models.
Results: Eighty-five women experienced at least one symptom of psychosis in their lifetime (37.4%), with 44 reporting symptoms in the previous 30 days (19.4%). Sixty percent (n=139) used marijuana in the year prior to incarceration, with almost 48.2% of users (n=67) reporting daily use. Seventy-three (37.1%) used crack-cocaine in the year prior to incarceration, with 46.6% of users (n=34) reporting daily use. Frequency of marijuana use in the year prior to incarceration did not significantly predict current psychosis (p=.853). Frequency of crack-cocaine use in the year prior to incarceration significantly predicted current psychosis (OR = 1.285, 95% CI = 1.053-1.569, p=.014). A one-unit increase in frequency of crack-cocaine use was associated with being 28.5% more likely to report current psychosis. However, when frequency of CV was added to the model, only the CV variable remained significant (OR=1.725; 95% CI: 1.216-2.445; p=.002). A one-unit increase in frequency of CV was associated with being 72.5% more likely to report current psychosis.
Conclusions and Implications: Although the relationship between frequency of marijuana us in the year prior to incarceration and current psychosis was non-significant, frequency of crack-cocaine use in the year prior to incarceration was a significant predictor of current psychosis for women in the sample. However, adding CV to the model moderated the importance of crack-cocaine use in the incidence of current psychosis for affected women, making the drug use variable non-significant. This analysis supports findings which suggest that marijuana and crack-cocaine use may function as a method of coping with previous victimization for incarcerated women. Research is needed to better understand the interaction between CV and drug use for this population as a vehicle to improve the design and implementation of trauma-informed substance misuse interventions delivered within the prison milieu.