Estimating Drinking Trajectories in the United States From Adolescence to Old Age: A Three-Step Approach

OBJECTIVES: Understanding how people transition over time from adolescence to old age across levels of drinking is crucial when trying to understand the lifetime effects of alcohol interventions delivered to adolescents. This study estimates the transition probabilities between different levels of alcohol consumption to understand the long-term drinking trajectories of adolescents in the United States.

METHODS: Using Bayesian Simulation techniques, we computed transition probabilities from adolescence to old age (12-65) in the United States, specifically focusing on the four drinking risk levels defined by the World Health Organization (low, medium, high, and very high risk), as well as abstinence and dependence. We used the National Longitudinal Survey of Youth 1997 cohort (NLSY97) and the National Survey on Drug Use and Health (NSDUH) datasets to calculate these transition probabilities. A simulated annealing approach was used for the calibration of transition probabilities to minimize the difference between real and simulated distributions. We validated the transition probabilities using the NSDUH 2018 cohort which was not used to calculate transition probabilities.

RESULTS: Across 16 age groups, the 1-year probability of remaining in the same state averaged 0.84, 0.47, 0.31, 0.33, 0.33, and 0.49 for abstinent, low, medium, high, very high-risk, and dependence, respectively. The probability of remaining in the abstinent state decreases significantly at age 21 and then increases gradually as people get older. People from higher-risk drinking states have a higher probability of going to the dependence state.

CONCLUSIONS: These transition probabilities are essential for understanding the progression of alcohol use over a lifetime and can be used to estimate the long-term impacts of alcohol interventions for adolescents.