Age-related differences in cigarette smoking among whites and African-Americans: evidence for the crossover hypothesis.

Background: Age crossover describes the age-related reversal in prevalence of current cigarette smoking among non-Hispanic whites and African-Americans, with prevalence higher among whites than African-Americans in adolescence but lower in adulthood. Prior studies have examined smoking patterns in separate adolescent and adult samples and have not sought to identify factors that could account for crossover. We conducted analyses using national samples to identify factors that account for crossover and estimate their impact on crossover age.

Methods: Analyses are based on national samples of lifetime smokers 12-49 years old in the 2006-2008 aggregated National Surveys on Drug Use and Health (N=61, 757) (SAMHSA, 2007-2009) and on multiple birth cohorts followed over 21 cross-sectional surveys.

Results: We identified crossover for cigarette smoking in the US population at about age 29. Crossover is partially explained by differences between whites and African-Americans in education and marital status, and more weakly by the opposite impact of age of smoking onset on persistence of smoking in the two groups. Controlling for smoking history, education and social role participation would raise crossover in current smoking by more than 14 years. Rates of current smoking among lifetime smokers at four different age categories in multiple birth cohorts followed from ages 12-17 to 35 and over in 21 surveys spanning 24 years confirm the age-related patterns observed cross-sectionally.

Conclusion: Age crossover for current smoking appears among whites and African-Americans. Efforts targeted toward improving educational levels of young people would have the strongest impact in decreasing persistent smoking, especially among African-Americans.