Understanding Alcohol Use Over Time in Early Mid-Adulthood for At-Risk Men

Oregon Social Learning Center, Inc., Eugene OR

Investigators

Linked publications, trials & patents

Abstract

DESCRIPTION (provided by applicant): The proposed study would examine issues related to the course, including persistence and desistance, of alcohol use and AUDs in early midadulthood for men. Such behaviors are costly to society both financially and in personal and family distress. These issues would be examined by utilizing an existing longitudinal data set currently spanning later childhood to ages 31-32 years, including 23 yearly assessments of alcohol use, and adding two new waves of data collection at ages 35-36 and 37-38 years. Combining these two new waves with existing data from ages 29-30, 30-31, and 31-32 will provide a total of 5 waves in the early midadult period. The multimethod/multiagent study of a community sample at risk for conduct problem and other risk behaviors began with two cohorts of fourth-grade boys (total N = 206); the retention rate at the last wave was 95% of living subjects (N = 191, ages 31-32 years). The study focused on contextual, family of origin, and peer risk factors based on a social learning model, and data are available on romantic partners. We plan to use a Dynamic Developmental Systems approach, based on aspects of past and current social- contextual influences, to study heterogeneity in the course of alcohol use and AUDs and predictors of the course, including general and specific risk factors within the realms of family, peer, psychopathology, and early to midadult risk and social influences, including romantic partner behaviors. Early onset of alcohol use was assessed prospectively by age at first drink and age at first drunken experience. Association of growth and heterogeneity in growth of alcohol use will be examined also in relation to heterogeneity in the course of co- morbid substance use (tobacco, marijuana, other illicit drugs, non-prescription and over-the-counter drugs), and gambling behavior. In addition, prediction models of treatment seeking will be examined. The two new assessments will include self-report, peer report, and records data (police and DMV records). Reports from intimate partners are available in a companion study. A variety of complementary developmental modeling and prediction techniques will be used to capture the degree to which prior family and contextual factors, and prior behaviors in childhood and adolescence (e.g., antisocial behavior in childhood) and similar contextual, social relationships, and behavioral factors in adulthood, can explain changes in the course of behaviors across the early adult period.