maximum (e.g., 5 if their maximum drinks were 7). Individuals who never drank more than their “typical” amount were assigned frequencies of 0 for the other categories. Weighting these 3 quantity estimates by their frequencies produced an estimate of the average number of drinks consumed per month, which we term past-year drinking density (PYDD). This variable provides a more accurate estimate than a simple quantity × frequency measure. PYDD was subsequently converted to a weekly estimate, truncated at 168 (i.e., 24 drinks/d), and then log-transformed to adjust for skewness arising from a small number of individuals who reported consuming extremely large quantities of alcohol.
