Research That Matters (January 17 - 20, 2008)

Background: Studies indicate that an estimated 9 % to 20 % of adolescents have engaged in “huffing” or “sniffing” of inhalants such as gasoline, glue, shoe polish, paint, and correction fluid (Johnston et al., 2006; Wu et al., 2004). Little if any systematic study of the reasons for using inhalants has been undertaken. This is unfortunate because knowledge of the reasons for using inhalants may eventually lend itself to the development of targeted prevention and treatment strategies. The purpose of this study was to identify specific subgroups of adolescents based on their reasons for using inhalants and translate findings into concrete treatment implications.

Methods: The present study sample (N= 277) of inhalant users was drawn from a larger survey of the population of current residents (N= 740) in the Missouri Division of Youth Services (DYS). All youth providing written informed consent completed the structured face-to-face interview assessing demographic characteristics, lifetime and annual use of inhalants, other drug use, substance-related problems, current and lifetime psychiatric symptoms, and externalizing behaviors. A set of fifteen items reflecting a broad range of reasons for using inhalants were selected for this study. Latent Profile Analysis (LPA) was used to identify subgroups of subjects who provided qualitatively different sets of reasons for using inhalants. The empirical fit of the model was based on the Bayesian Information Criterion (BIC). A validation analysis of the final model was conducted by testing associations of class characteristics with demographic and clinical measures.

Results: Overall, the three class model provided the best statistical and conceptual fit to the data. Classes consisted of 1) experimental user's, 2) experimental users/moderate distress, and 3) abusers/high distress. Chi-square tests revealed that class composition did not differ by race, gender, receipt of public assistance, or urbanicity. One-way analysis of variance (ANOVA) with Dunnett's T3 post-hoc comparisons revealed significant differences on all the clinical variables. The strongest effect sizes observed were on the Substance Use Problems Index (η² = .13), Anxiety-BSI (η² = .14), and the Global Severity Index (η² = .13). Although these effect sizes appear low relative to their zero-order counter-parts (e.g., Cohen's d), they are squared multiple correlations. Thus, they are of moderate strength and considered clinically significant.

Implications for practice: Results indicate that class profiles reflect stages of a continuous process from experimentation to abuse to dependence. Thus, study findings potentially provide a framework to assess and interrupt this progression. Future taxonomic research on inhalant users should study the stability of identified classes across time, employ diagnostic mental health inventories and assess the moderating effects of situational influences, family problems and other drug use progression.