Substance use is highly prevalent among adolescents and young adults (AYAs) with polysubstance use being particularly concerning due to its association with greater substance-related consequences and heavier use patterns compared to single substance use. While research has established associations between polysubstance use and various risk factors including impulsivity, depressive symptoms, and risky sexual behaviors, most studies have examined these relationships at the person level rather than investigating nuanced daily patterns of co-occurring substance use. Given that females demonstrate higher substance use rates and increased likelihood of same-day polysubstance use, the present study used latent class analysis (LCA) to (1) identify daily substance use patterns in AYA females reporting recent binge drinking, and (2) examine associations between person-level and day-level factors with these daily use patterns.
The study utilized a longitudinal design with 149 female participants aged 14-24 years recruited from the south-central US (2017-2018), with inclusion criteria requiring at least one binge drinking episode within two weeks at screening. Participants completed baseline questionnaires assessing impulsivity (Barrett Impulsiveness Scale Version 11), depressive symptoms (Center for Epidemiologic Studies Depression Scale Revised), and demographics, followed by weekly visits over one month reporting daily substance use via Timeline Followback methodology. Day-level latent class analysis was conducted using five substance use indicators, with model selection based on information criteria and interpretability. Associations between latent classes and person-level (demographics, impulsivity, depression) and day-level factors (weekday/weekend, condomless sex) were examined using the BCH 3-step approach with Bonferroni correction.
According to the study findings, most were under 21 years old (67.1%), partnered (60.4%), and attracted to males (73%), with 62% being non-Hispanic White and 38% from minority racial/ethnic groups. Day-level latent class analysis of 4,224 person days (149 participants) identified a 4-class solution as optimal: “Little-to-no-use days” (54.1% of days) characterized by minimal substance use; “Vaping-only days” (28.2%) with high vaping probability and low other use; “Cannabis-only days” (10.5%) with high cannabis use probability; and “Heavy-alcohol-and-cannabis days (some smoking and vaping)” (7.2%) involving high probabilities of heavy alcohol and cannabis use with moderate smoking and vaping. After Bonferroni correction, significant associations emerged where vaping-only days were linked to younger age and less full-time employment, while little-to-no-use days were associated with minority race/ethnicity and lower depressive symptoms compared to all other day types. Heavy-alcohol-and-cannabis days and vaping-only days showed higher impulsivity than little-to-no-use and cannabis-only days, with heavy-alcohol-and-cannabis days being more likely on weekends and associated with condomless sex compared to little-to-no-use and vaping-only days.
Takeaway: AYA females exhibit four distinct daily substance use patterns, with the most problematic pattern (heavy alcohol and cannabis days) associated with higher impulsivity, weekend timing, and condomless sex, highlighting the need for targeted interventions addressing both substance use and co-occurring risky behaviors.