Online Program

The Causal Effect of Substance Use Latent Class Membership on a Distal Outcome

*Bethany C Bray, The Methodology Center, Penn State 
Donna L Coffman, The Methodology Center, Penn State 
Stephanie T Lanza, The Methodology Center, Penn State 

Keywords: causal inference, propensity scores, latent class analysis, alcohol use, cigarette use, marijuana use

Adolescent substance use, abuse, and dependence are known to be associated with adulthood drug use, abuse, and dependence, yet it is unclear whether earlier substance use behavior profiles have long-term causal impacts on later outcomes. Recent advances in causal inference methods have been proposed to account for possible confounders (i.e., variables that jointly influence substance use and subsequent outcomes) so that causation may be inferred. Modern causal inference techniques are rapidly being adopted by prevention scientists, but methods for drawing causal inferences about the consequences of behaviors measured as latent classes have not been available. This talk will demonstrate a new approach to estimating causal effects of latent class membership (adolescent alcohol use) on a distal outcome (adult substance use).

The sample comprised 409 adolescents (53.3% female; mean age=14.9) from the 9th-grade cohort of the National Longitudinal Study of Adolescent Health (Add Health) who had tried alcohol. Potential confounders included individual, family, peer, school, and neighborhood characteristics. The latent exposure, alcohol use behavior latent class membership, was assessed at Grade 9 using indicators of alcohol use frequency, binge drinking, friends' alcohol use, and alcohol-related problems. The outcomes, heavy drinking, regular cigarette use, and marijuana use, were assessed during emerging adulthood. We estimated the causal effect of adolescent alcohol use behavior profile membership on substance use in emerging adulthood using the R package LCCA.

We identified 3 latent classes: Non-recent drinkers (27% of sample) were unlikely to report any alcohol use; Light drinkers (43%) used alcohol occasionally but did not binge drink or have alcohol-related problems; Frequent Binge drinkers (30%) use alcohol multiple times per month, binge drank, and were likely to experience alcohol-related problems. Prior to adjusting for confounding, membership in the Light and Frequent Binge drinking classes was associated with elevated levels of adulthood heavy drinking, regular cigarette use, and marijuana use compared to Non-recent drinkers. Results based on LCCA, which adjusts for confounders, suggested that only adulthood heavy drinking and marijuana use (but not regular cigarette use) were consequences of membership in the Light and Frequent Binge drinking classes.

Modern approaches for LCA with a distal outcome can provide insight into the potential risk conferred by class membership, but as with standard regression-based approaches, coefficients describing the association between latent class membership and the distal outcome are correlational. We demonstrate a new methodological approach for estimating the causal effect of latent class membership on a distal outcome in the context of a brief software demonstration so that public health scientists may adopt this approach in their own work.