Predictors of substance abuse in a methadone maintained sample

Abstract

Two hypotheses were tested regarding the effects of childhood trauma on present patterns of substance abuse in a methadone maintained sample. The major hypothesis stated that childhood trauma will predict patterns of substance abuse directly. The alternate hypothesis stated that childhood trauma will predict patterns of substance abuse through the possible mediation of four variables: social network roles, manifest anxiety, environmental instability, and negative life events.;One hundred patients in a methadone clinic and a balanced, non-addicted group of fifty subjects were administered a questionnaire that included five scales: childhood trauma, social network, environmental instability, Taylor Manifest Anxiety Scale, and the Life Event Schedule. For the methadone group, toxicology reports of illicit drug use, antedating the interview by six months, were used to construct a substance abuse index. This was based on the percent of toxicology reports showing use of illicit substances.;The data analysis utilized t-tests, analysis of variance, multiple regression, and path analysis. Multiple regression analyses indicated that the clinic group was significantly higher than the control group on measures of childhood trauma, negative life events, and environmental instability. Multiple regression analyses for the clinic group indicated that higher social network roles were inversely related to substance abuse, whereas, negative life events were positively associated with substance abuse. Childhood trauma did not make any significant direct contribution predicting the drug index in the full regression model. Path analysis suggested that childhood trauma affected substance abuse through the mediation of negative life events, i.e. early trauma was related to a higher level of negative life events, which were the proximal predictors of drug abuse.;Additional analyses tested the interaction between social network roles and negative life events in predicting substance abuse. A stress-buffering effect was found: negative life events were unrelated to substance abuse when the social network roles were high, but when social network roles were low, there was a strong positive relationship between negative events and substance abuse.;The major hypothesis that childhood trauma will directly predict patterns of drug abuse did not find support. The alternate hypothesis that childhood trauma will predict drug abuse through the mediation of proximal variables was supported.