A latent class analysis of illicit drug abuse/dependence: results from the National Epidemiological Survey on Alcohol and Related Conditions

Abstract

We examine the co-occurrence of abuse/dependence across different illicit drugs and test associations between these classes and major psychiatric disorders. Latent class analyses were used to characterize polysubstance abuse/dependence (AB/D) in 43 093 individuals who participated in the National Epidemiological Survey on Alcohol and Related Conditions. Multinomial logistic regression was performed to examine the association between the classes of life-time illicit drug AB/D and gender, age and race, as well as life-time Diagnostic and Statistical Manual version IV (DSM-IV) alcohol abuse/dependence, nicotine dependence, major depressive disorder, generalized anxiety disorder, panic disorder, social phobia and antisocial personality disorder. Five latent classes were identified: no AB/D (class 1: 92.5%), cannabis AB/D only (class 2: 5.8%), stimulants + hallucinogen AB/D (class 3: 0.6%), prescription drug AB/D (class 4: 0.6%) and polysubstance AB/D (class 5: 0.5%). Major depressive disorder and nicotine dependence were associated most strongly with class 5. Anxiety disorders were associated strongly with the prescription drug AB/D class while alcohol AB/D and ASPD were associated with classes 2, 3, 4 and 5 when compared to the reference class (class 1). Significant heterogeneity exists in this US population for polysubstance AB/D patterns with evidence for a subgroup with high rates of sedative, tranquilizer and opiate AB/D and a history of anxiety disorders, a stimulant/hallucinogens group, a high-risk group with elevated rates of all psychiatric disorders and a milder cannabis AB/D only group. Replication of such classes across other samples has significant implications, such as characterizing risk groups that may be etiologically diverse.