Economic hardship linked to cigarette-cannabis co-use in 10 states

A population-based study of adults in 10 US states found that all five measured indicators of economic instability were associated with cigarette-cannabis co-use, with the strongest adjusted relationship among people receiving Supplemental Nutrition Assistance Program benefits. Researchers analyzing the 2023 Behavioral Risk Factor Surveillance System also identified distinct associations between particular hardships and smoking cigarettes or using cannabis alone. The findings come from cross-sectional survey data and therefore describe statistical associations rather than showing that financial hardship caused substance use.

The study included 50,481 adults aged 18 or older, representing an estimated 24,896,969 people after survey weights were applied. The participants lived in Connecticut, Delaware, Illinois, Indiana, Maine, Mississippi, Montana, Nevada, New Mexico and Virginia, the states that administered both the survey’s marijuana-use module and its Social Determinants and Health Equity module. Overall, 8.1% of participants reported exclusive traditional cigarette use, 11.1% reported exclusive cannabis use and 3.6% reported using both cigarettes and cannabis.

Exclusive cannabis use was therefore more prevalent than exclusive cigarette use in this 10-state sample, while co-use was less common but had the strongest associations with economic instability. Researchers examined five indicators of economic instability reported over the preceding 12 months: employment instability, defined as losing a job or experiencing reduced working hours; receipt of food stamps through the Supplemental Nutrition Assistance Program; food insecurity; housing insecurity, including inability to pay mortgage, rent or utility bills; and threats of utility shutoffs involving electricity, gas, oil or water service. Each indicator was common in the weighted sample.

Employment instability was reported by 12.0% of participants, food-stamp receipt by 10.4%, food insecurity by 13.9%, housing insecurity by 11.6% and threatened utility shutoffs by 7.7%. All five economic-instability indicators were significantly associated with cigarette-cannabis co-use after adjustment for demographic characteristics, socioeconomic factors and selected health conditions. The largest adjusted relative risk ratio was for participants receiving food stamps, who were 2.27 times as likely to report co-use as those not receiving them, with a 95% confidence interval of 1.75–2.94.

A threatened utility shutoff carried the second-largest association, corresponding to a relative risk ratio of 2.19 and a 95% confidence interval of 1.68–2.84. The models compared each substance-use category with no substance use, rather than directly comparing co-use with either exclusive-use category. The remaining adjusted co-use estimates were also above one.

Housing insecurity was associated with a relative risk ratio of 2.03 and a 95% confidence interval of 1.59–2.60; food insecurity with 1.97 and a 95% confidence interval of 1.56–2.50; and employment instability with 1.52 and a 95% confidence interval of 1.17–1.97. Because the survey measured economic conditions and substance use at one point in time, these estimates show that the groups differed in their likelihood of reporting co-use. They do not establish a sequence in which economic hardship preceded substance use, quantify causal effects or determine how much of each association could be explained by unmeasured factors.

Economic insecurity was also associated with use of each substance by itself, but the pattern differed from that for co-use. People reporting threatened utility shutoffs had the largest adjusted association with exclusive cigarette use, at a relative risk ratio of 1.78 and a 95% confidence interval of 1.37–2.31. Food insecurity was associated with exclusive cigarette use at 1.57, with a 95% confidence interval of 1.28–1.92; receipt of food stamps at 1.55, with a 95% confidence interval of 1.24–1.93; and housing insecurity at 1.47, with a 95% confidence interval of 1.17–1.83.

Employment instability was not statistically associated with exclusive cigarette use after adjustment. For exclusive cannabis use, the strongest association was with threatened utility shutoffs, at a relative risk ratio of 1.37 and a 95% confidence interval of 1.10–1.70. Housing insecurity was associated with exclusive cannabis use at 1.27, with a 95% confidence interval of 1.04–1.55, while employment instability had an association of 1.23, with a 95% confidence interval of 1.02–1.48.

Food insecurity and food-stamp receipt were not statistically significant in the adjusted model for exclusive cannabis use. The results therefore did not map one hardship neatly onto a single substance category: some indicators were linked to cigarette use alone, others to cannabis use alone, and all five were linked to using both. The researchers used five multivariable multinomial logistic regression models, with no substance use as the reference category and each economic-instability indicator entered in a separate model.

Adjustments included age, sex, race and ethnicity, marital status, education, metropolitan residence, health insurance coverage, self-rated general health, perceived stress and histories of lung disease, cardiovascular disease and cancer. The analysis accounted for the complex survey design using sampling weights, and statistical significance was defined as a P value of 0.05 or lower. The 2023 survey had a 44.7% response rate overall.

The analytic sample was 48.6% aged 50 or younger, 52.6% women and 64.4% self-identified as non-Hispanic White. Several limitations constrain how the results can be interpreted. The participating states were selected because they administered both required optional modules, and the resulting sample is not a fully nationally representative picture of the United States.

Telephone collection and the survey’s response rate may have affected participation by people who are economically marginalized or otherwise difficult to reach. Economic conditions and substance use were self-reported, creating potential for recall errors and misclassification, while relevant influences such as family history of substance use disorders and history of trauma were not measured. Substance use was not measured in detail: the survey established current cigarette use from questions about lifetime smoking and present smoking, and current cannabis use from use on at least one of the previous 30 days, but it did not provide detailed information about duration, frequency, intensity or methods of use.

The economic-instability measures also did not fully capture the severity or persistence of hardship. These limitations leave unresolved which came first, what mechanisms might link financial conditions with substance use and how the findings differ among demographic groups. The authors concluded that the results support targeted, integrated responses involving public health, housing and social-service systems, including routine screening for economic insecurity and connections to assistance, while calling for longitudinal studies to clarify temporal relationships and mechanisms.

Subject of Research: Associations between economic instability and cigarette, cannabis and co-use patterns among adults in 10 US states

Article Title: Economic instability and substance use among adults in 10 US states: Evidence from a population-based study

Article References: Liu, P., Espinosa, T. D. J., Samami, E., Mensah, B., & Poghosyan, H. (2026). Economic instability and substance use among adults in 10 US states: Evidence from a population-based study. PLOS One, 21(10), e0360066. https://doi.org/10.1371/journal.pone.0360066

Image Credits: AI Generated

DOI: 10.1371/journal.pone.0360066

Keywords: economic instability, cannabis use, cigarette use, substance co-use, food insecurity, housing insecurity, employment instability, SNAP benefits, utility shutoffs, US adults, Economic, instability