Background:
Since 1964 when the first Surgeon General’s report on smoking was published, the decline in smoking has been steepest among those with the most education. As smoking has become denormalized, it is increasingly concentrated in vulnerable populations that are marginalized by socioeconomic status, mental health, race/ethnicity, or sexual orientation. This analysis uses the National Health Interview Survey (NHIS) to assess current trends in cigarette smoking among vulnerable vs. non-vulnerable populations.
Methods:
Current cigarette smoking among adult NHIS respondents was tracked from 1999-2018. Vulnerable populations were defined by low socioeconomic status (<100% federal poverty level, <high school education, Medicaid recipient), low mental health (Kessler 6 score≥5), or race/ethnicity with elevated smoking (American Indian, multiracial). Sensitivity analyses added food insecurity and minority sexual orientation (non-straight), available starting 2013, to the definition of vulnerability. Regression analysis projected future smoking trends.
Results:
Smoking rates in all sub-groups declined from 1999 to 2018. Among the non-vulnerable, smoking declined from 19.2% to 9.7%; among the vulnerable, smoking rates were 10 percentage points higher, declining from 30.5% to 19.5%. Smoking rates among vulnerable populations are now double the rates in non-vulnerable populations. The proportion of smokers classified as vulnerable increased from 47.7% to 58.6%; the proportion of cigarettes consumed by vulnerable smokers increased from 50.1% to 60.4%. Expanding the definition of vulnerability to include those food insecure and in sexual minorities, the proportion of smokers classified as vulnerable (2018) was 61.63%. The most common vulnerabilities in the population (2018) were low socioeconomic status (25.5%) and poor mental health (22.0%); (7.8% were both). In 2018, as the number of vulnerabilities increased (0, 1, 2, ≥3), so did the smoking rate (9.4%, 14.6%, 21.7%, 31.7%, respectively, p<0.001). Using linear regression to extrapolate current trends, non-vulnerable populations will eliminate cigarette smoking in 2037, vulnerable populations in 2060.
Conclusion:
Existing public health policy and social trends are reducing smoking rates among all smokers, but absent concentrated effort to reduce smoking in vulnerable populations, health disparities from disproportionate tobacco use will increase and reverberate for decades.
Do shift-and-persist strategies predict the mental health of low-socioeconomic status individuals?