TABLE 3.
AORs of Individual-Level Youth Drinking and Binge Drinking Associated With a 10 Percentage Point Increase in Modified APS Score Based on Age-Related Policy Subgroups, YRBS, Biennial Years 1999–2011
| Policy Group Exposure Variablea | Youth Drinking,b AOR (95% CI) | Youth Binge Drinking, AOR (95% CI) |
|---|---|---|
| Population-oriented policy subgroup (n = 19) | ||
| Adjusted modelc | 0.93 (0.91–0.96) | 0.95 (0.93–0.97) |
| Also controlling for youth policiesc | 0.94 (0.92–0.97) | 0.96 (0.94–0.99) |
| Youth-oriented policy subgroup (n = 10) | ||
| Adjusted modelc | 0.96 (0.94–0.98) | 0.95 (0.94–0.97) |
| Also controlling for population policiesc | 0.98 (0.96–0.99) | 0.97 (0.95–0.99) |
Population-oriented policies consisted of policies that target the general population (eg, alcohol taxes, hours of alcohol sales). Youth-oriented policies target primarily underage youth (eg, policies related to the minimum legal drinking age). Methods used to calculate the 2 modified APS scores were identical to those used for calculating the APS score based on all 29 policies, but they used restricted sets of mutually exclusive policies.
Youth drinking was defined as ≥1 d of consuming alcohol during the past 30 d. Binge drinking was defined as ≥1 d of having ≥5 drinks of alcohol “in a row, that is, within a couple of hours” during the past 30 d.
For all models, the modified state APS scores were used to predict odds of individual-level youth drinking and youth binge drinking in the same state-year. AOR is based on absolute 10 percentage point increase of the APS scores, with generalized estimating equations adjusted for YRBS weight with multiple-year adjustment and the clustering of individuals within sampling units. Individual-level covariates included age, gender, race or ethnicity, and past month tobacco use. State-level covariates include proportion of adults ≥21 y of age, gender distribution, race or ethnicity distribution, degree of urbanization, median household income, religiosity, police officers per capita, and geographic region. Year was treated as a categorical variable.