The 131st Annual Meeting (November 15-19, 2003) of APHA

Abstract #69825

William C. Kerr, PhD¹, Thomas K. Greenfield, PhD², Jason Bond, PhD¹, Yu Ye, MA¹, and Jürgen Rehm, PhD³. (1) Alcohol Research Group, 2000 Hearst Ave. Suite 300, Berkeley, CA 94709, 510 642-5208, wkerr@arg.org, (2) Public Health Institute, Alcohol Research Group, 2000 Hearst Ave, Suite 300, Berkeley, CA 94709, (3) Social, Prevention and Health Policy Research, Centre for Addiction and Mental Health, 55 Russell St, Toronto, ON M5S 2S1, Canada

Comparable measures of the prevalence of beer, wine and spirits consumption and the volume of these beverages they consume of each are analyzed across the last five National Alcohol Surveys conducted in 1979, 1984, 1990, 1995 and 2000. Logit models of any consumption of each beverage and least squares models of the yearly volume of each beverage consumed are estimated. Survey design and weighting are accounted for and models are run with and without demographic controls for ethnicity, region, employment, marital status and education. Models for all individuals, men, women, drinkers only, male drinkers and female drinkers are estimated separately. Age and birth cohort are modeled in five-year groups and period is modeled using indicator variables for each survey. Model identification requires two age, period or cohort groups be constrained as equal. Reference models constrain the 70 to 74 years old and 75 to 79 years old age groups to be equal. Sensitivity analyses constrain each set of adjacent age groups and birth cohorts to be equal. Results will identify birth cohorts with especially high or low consumption of each beverage controlling for age, period and demographic effects. The effects of age on consumption controlling for cohort effects and trends will also be presented. Association between beverage consumption and birth cohort may aid the interpretation of beverage-specific associations with alcohol-related mortality.

Learning Objectives:

Participants will
• learn the importance and difficulty of disentangling age, period and cohort effects.
• identify especially heavy and lighter drinking birth cohorts enabling them to target prevention efforts and anticipate chronic disease prevelence.
• Gain a clearer understanding of the importance of age effects in the reduction of heavy drinking.

Keywords: Alcohol Use, Aging

Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.