Using the Past to Understand the Future of U.S. and Global Smoking Disparities: A Birth Cohort Perspective

Abstract

U.S. smoking-related disparities persist, but data evaluating how smoking patterns across diverse populations have changed by birth cohort are lacking. Worldwide, smoking continues to exact harm, especially to low- and middle-income nations with less historical data for smoking analyses. The Cancer Intervention and Surveillance Modeling Network (CISNET) Lung Working Group previously generated smoking histories for the whole U.S. population using an age, period, and birth cohort (APC) methodological framework. These inputs have been used in numerous models to simulate future patterns of smoking and to evaluate the potential impact of policies. However, the absence of detailed model-ready inputs on smoking behaviors for diverse U.S. populations has been a barrier to research evaluating future trends in smoking-related disparities or the projected impacts of policies across sociodemographic groups. This supplement issue provides new estimates of smoking behaviors with detailed historical data by race/ethnicity, educational attainment, family income, and for each of the 50 U.S. states and Washington, DC. All-cause mortality relative risks associated with smoking by race/ethnicity and educational attainment are also available for the first time. Finally, the supplement issue presents comprehensive smoking histories for Brazil, demonstrating the application of this methodology to resource-limited settings. Collectively, these data aim to offer insight into future U.S. and global smoking disparities and accelerate research on tobacco control policies that advance health equity. This effort will allow tobacco simulation models to account comprehensively for population diversity, thereby enabling researchers to develop more sophisticated analyses of tobacco use and control interventions.

Figure 1.

Smoking prevalence by birth cohort for the U.S. population. Notes: Lines represent individual birth cohorts from 1910 (black line, farthest to left) to 2000 (teal line, farthest to right). Data reflect increasing smoking prevalence among females from 1910‒1940 birth cohorts, and declines from 1950‒2000 birth cohorts; among males, smoking prevalence increased from the 1910 to 1920 birth cohorts, with declines in most subsequent cohorts. Estimates of current smoking prevalence by cohort and gender were generated by the Cancer Intervention and Surveillance Modeling Network (CISNET) Lung Working Group using the age, period, and birth cohort (APC) methodological framework.^,

Figure 2.

The CISNET Lung Working Group approach to modeling smoking, smoking-related mortality, and lung cancer. Notes: Black lines with text in bold typeface represent research and data published in this supplement. Age, period, and cohort analyses were conducted for the U.S. population by race/ethnicity, education, income, and state, and for the Brazil population. Mortality analyses were conducted for the U.S. population by race/ethnicity and education only.

Figure 3.

Extensions of the age-period-cohort framework for population smoking histories. Notes: Gray outlines represent previously published applications of the Age-Period-Cohort (APC) framework for smoking histories. Black outlines represent APC extensions to new populations published in this supplement.