Age at first use of alcohol predicts the risk of heavy alcohol use in early adulthood: A longitudinal study in the United States

• Findings are based on analysis of data from a longitudinal study.
• Younger age at first use of alcohol increased risk of heavy alcohol use.
• A newly analytical approach was used to control for unmeasured confounding effects.
• The observed association is likely to be causal and affects the general population.
• A straightforward causal mechanism is discussed.

BackgroundTwo ecological cross-sectional studies which relied on national survey data (U.S. and Australia) have shown that starting drinking at a younger age increases the frequency of heavy drinking in the general population, including those with good mental and physical health status. This study further investigates the hypothesis that age at first use of alcohol increases the risk of heavy alcohol use by applying data from a longitudinal study.MethodThis study used public-use data collected from Wave I, Wave III and Wave IV of the National Longitudinal Study of Adolescent Health in the United States. The association between age at first use of alcohol and heavy alcohol use (5+ drinks per occasion) was examined with two different multivariate analysis approaches with data from 2316 participants: ordered logistic regression models and Poisson regression models with longitudinal data settings. In addition, the newly developed proxy outcome approach was further used to estimate and adjust for unmeasured/unobserved confounding factors.ResultsAge at first use of alcohol before 18 years was associated significantly higher risk of heavy alcohol use at follow-up.ConclusionAfter adjusting for known and residual confounders, younger age at first use of alcohol was associated with significantly higher risk of heavy alcohol use, moreover, we posit that the association observed from this longitudinal study is probably causal. Abstinence from alcohol until the age of 18 years will likely reduce individual risk of alcohol-related problems in adulthood. In the longer term, delayed onset of exposure with widespread abstinence among this age group is also likely to reduce the overall prevalence of alcohol-related problems in the general population.