Educational differences in alcohol consumption and heavy drinking: An age-period-cohort perspective

Background: Low socioeconomic status (SES) has been associated with lower alcohol consumption, but also with heavier drinking. To explain this contradictory relationship, we examined SES differences in drinking patterns from an age-period-cohort (APC) perspective. Methods: Data are from seven waves of the U.S. National Alcohol Surveys from 1979 to 2010. As a proxy for SES, educational attainment was used. Past-year alcohol volume was calculated from frequency (never-to-every day) and usual quantity (1–2, 3–4, or 5–6 drinks). Past-year frequency of heavy episodic drinking was labelled as total days of 5+ drinks. Gender-stratified APC fixed-effects models were conducted controlling for demographics and adjusting for survey design and weights. Results: Significant APC effects by education were found, but the direction varied by alcohol measure. Education and total volume were positively associated across APC. Cross-over effects for age occurred with a positive education-heavy drinking relationship in young adulthood and negative relationship in mid-adulthood. Cohort-by-education effects showed greater heavy drinking among less educated women in 1956–60 cohort and more educated men and women in younger cohorts (post-1976). Conclusions: Higher SES is consistently associated with total volume across age, period, and cohort, but less consistently with heavy drinking. While there are currently significant intervention efforts to reduce heavy drinking in young adulthood, our study suggests the need for age-specific strategies targeting lower-SES groups in mid-adulthood and cohort-specific strategies for lower-SES women in the baby boomer cohort and higher-SES men and women in younger birth cohorts.