Latent Classes of Individual Assets and Substance Use in at-Risk Adolescents and Adaptive Functioning in Emerging Adulthood

Background & Purpose: Few studies have examined the combined patterns of dimensions of individual strength and liability as a whole, especially in youth with multiple biological and environmental risks. Assessing these attributes, in the context of each other rather than in isolation, is crucial for understanding youth’s developmental trajectories and creating personalized interventions. The present study aimed to: 1) characterize subgroups using indicators of individual assets (e.g., social competencies, positive identity) and substance use in adolescents, and 2) investigate how these subgroups are associated with later adaptive functioning (e.g., high school graduation, problematic use of alcohol, tobacco, marijuana, legal system contacts) in emerging adulthood.

Methods: Participants were 358 adolescents (54% female), primarily African-American, urban, and of low socioeconomic status, who were recruited at birth for a prospective study on the effects of prenatal substance exposure. At age 15, adolescents’ individual assets were assessed using the Developmental Assets Profile, substance use via biologic assays (urine, hair, blood spots) and self-report, and substance use related problems (e.g., experiencing mood swings, forgetfulness, accidents) via the 17-item substance use/abuse scale from the Problem Oriented Screening Instrument for Teenagers. At age 21, substance use symptoms were assessed using the Substance Abuse Module, with symptoms ≥ 2 defined as a problematic use. High school graduation, legal problems (ever-experienced incarceration and/or probation), and behavioral adjustment using the Adult Self-Report, were also assessed. Latent class analysis (LCA) was conducted to classify adolescents into discrete, meaningful groups. Multivariable logistic regression analyses and ANCOVA controlling for gender were conducted to examine the validity of the LCA-derived groups. Adjusted means and standard errors and probabilities of the LCA-derived groups were calculated from the estimated models.

Results: LCA indicated a five-class solution as the optimal model (entropy = .81): 1) high-assets-with-low-substance use (SU) (10.2%); 2) moderate-assets-with-low-SU (28.7%); 3) low-assets-with-low-SU (32.0%); 4) moderate-assets-with-high-SU (9.4%); and 5) low-assets-with-high-SU (19.7%). The “low-assets-with-high-SU” group reported higher likelihood of problematic use of tobacco (65%), alcohol (28%), and marijuana (70%) than the three low SU groups at age 21. Similarly, the “moderate-assets-with-high-SU” group also reported higher likelihood of problematic use of tobacco (58%) and marijuana (50%), but not problematic alcohol (15%), than the three low SU groups. The “low-assets-with-high-SU” group reported lower likelihood of graduating high school (37%) and higher likelihood to be involved with legal system (32%) than the three low SU groups. The “moderate-assets-with-high-SU” group also reported higher likelihood to be involved with legal system (24%) than the two low SU groups, the “high-assets-with-low-SU” (9%) and the “moderate-assets-with-low-SU” (9%). However, the “moderate-assets-with-high-SU” group reported the lowest internalizing T score (M= 43, SE=2.7), while the “low-assets-with-high-SU” group reported the highest internalizing (M=53, SE=1.9) and externalizing T-scores (M= 52, SE=1.6).

Conclusions: The current findings suggest utility for a holistic assessment of adolescents’ internal assets and substance use in determining risk classes, which can inform tailored prevention and developmental asset-based drug use intervention programs. Furthermore, such an integrated approach will aid in identifying teens with the greatest risk for maladjustment in emerging adulthood.