A few notable points within our data require additional interpretation. In the longitudinal child and adolescent sample of baseline nonusers (i.e., the TAOS sample), we found that DLPFC GMV prospectively predicts an early age of drinking initiation. In the DNS longitudinal prospective data of young adults, reduced GMV in these regions also predicted future alcohol use, even after accounting for the extent of baseline alcohol use. However, this characteristic was only predictive up until 20.85 years of age. It is possible that risk conferred by reduced GMV in the DLPFC is developmentally constrained or may be minimized by environmental differences in permissivity or legality, as the legal drinking age in the United States is 21 years (66).
