Longitudinal Patterns and Predictors of Alcohol Consumption in the United States
Alison A. Moore, MD, MPH,
Robert Gould, PhD,
David B. Reuben, MD,
Gail A. Greendale, MD,
M. Kallin Carter, MS,
Kefei Zhou, MS and
Arun Karlamangla, MD, PhD
At the time of the study, Alison A. Moore, David B. Reuben, Gail Greendale, Kefei Zhou, M. Kallin Carter, and Arun Karlamangla were with the David Geffen School of Medicine, Division of Geriatrics, University of California at Los Angeles. Robert Gould is with the Department of Statistics, University of California at Los Angeles.
FIGURE 1—Predicted longitudinal alcohol consumption in 3 birth cohorts (1905, 1925, 1945) of men (M) and women (F) from the NHANES I sample. Note. Values were predicted with the following demographic covariates at baseline: White race, not married, working, nonsmoking, higher income ( $7000), and high school education. Per capita alcohol consumption was set at 2.5 gal/y. To clearly demonstrate the cohort modification of the age effect, we extrapolated alcohol consumption beyond the ages observed in the data (6687 years for the 1905 cohort, 4667 years for the cohort, and 2647 years for the 1945 cohort).