| Research has shown that age of initial substance use is related to continued substances use, intravenous substance use, trading sex for illicit substances, and arrest rates. However, this research has never focused on arrestees in general or substance using arrestees in particular. Thus, the purpose of this study was to determine whether age of initial illicit substance use can predict these four risky behaviors among a sample of substance using adult arrestees (n = 16,445). To determine this, four hypotheses were tested where age of initial illicit substance use predicted current illicit substance use, intravenous substance use, trading sex for illicit substance acquisition, and arrest rates. Because marijuana use is more prevalent than powder cocaine, crack cocaine, methamphetamine, or heroin use, the age of initiation of the latter four substances were combined to create one variable known as "age of initial hard drug use." All four hypotheses were tested with two independent variables: age of initial marijuana use and age of initial hard drug use. Logistic regressions were used to analyze the first three hypotheses and a generalized linear model that assumed a negative binomial distribution for count data was created to analyze the last hypothesis. As expected, age of initial marijuana use and age of initial hard drug use predicted current illicit substance use, past-year intravenous substance use, and arrest rates. However, none of the statistical models were valid in predicting the trade of sex for illicit substance acquisition. These results are likely due to the very low response rate in general and even lower positive response rate in particular. The primary limitation of this study is the robust sample size for a few of the statistical models used to make predictions. While some of the models (current use of powder cocaine, methamphetamine, heroin, and hard drug use; arrest rates) had acceptable fit, others did not (current use of marijuana and crack cocaine, past-year intravenous substance use); this is believed to be due to the influence of sample size on chi-square analyses. The results of this study adds to the body of knowledge where it was lacking, thereby providing information that could benefit prevention and treatment programming, and public health and criminal justice policy. |
