Abstract
Dear Editor,
Food insecurity, hunger, and malnutrition are highly complex, multilevel challenges worldwide. 1 We appreciate the Letter to the Editor regarding our paper published in this Journal. 2 We share the author’s conviction that tackling food insecurity is essential and agree that new intervention paradigms are needed to address this global challenge. The context of food systems and the drivers of food insecurity differ widely across regions shaped by contextual factors, such as political conflict, climate change, and supply chain dynamics, and these underlying factors have variable implications for health. 1 Our response therefore situates our study within the context of a high-income country and clarifies several points about intervention, outcomes, and interpretation. We conclude with clarifications on a systems perspective and intersectional approaches to intervention.
In the U.S., food insecurity has historically been measured at the household level by the U.S. Department of Agriculture and occurs when the “ability to acquire adequate food is limited by a lack of money or other resources.” 3 Food insecurity remains a pervasive public health concern that is consistently linked with poorer physical and mental health outcomes as well as higher healthcare utilization and costs, and mortality.4-7 Many healthcare systems have invested in addressing food insecurity as a mechanism to improve health and reduce costs. 8 This includes implementing routine screening and resource navigation for food assistance and adopting various models of “Food is Medicine” programs for chronic diseases that emphasize food quality as well as quantity. 8 These efforts vary greatly, underscoring that 1 model of addressing food insecurity does not fit all contexts.
Our study explored whether 1 method of addressing food insecurity could offer improvements in health within a specific domain of health behavior related to smoking cessation. 2 Cigarette smoking is a leading cause of preventable disease and death and is concentrated in populations with socioeconomic disadvantage. 9 Food insecurity and smoking often co-occur, driven by shared structural vulnerabilities. Our prior research has examined these connections extensively: the reciprocal and intersecting nature of the 2 variables,10,11 their dose-response relationship,12,13 the lived experience of using tobacco to cope with hunger and stress,14,15 and how changes in food insecurity correspond to changes in smoking over time.16,17 We have prior and ongoing research examining food distribution settings as venues for smoking cessation outreach. 18 Within this broader program of research, we hypothesized that reducing the burden of food insecurity while individuals attempt to quit smoking might improve cessation-related outcomes.
Translating epidemiological insights into individual-level intervention research is inherently complex. Our study was designed as a pilot test to assess feasibility and acceptability, rather than to test the intervention’s efficacy. With a sample size appropriate for pilot work (n = 55) and a short-term, 3-month duration, we found that the intervention was feasible and acceptable. A key distinction of our intervention was the use of a flexible, unconditional cash transfer provided as an adjunct to resource navigation for smoking cessation and food assistance. Participants in the intervention arm received monthly cash transfers equivalent to roughly 1 week of groceries, based on the USDA’s Moderate Food Plan, for 3 months. The funds could be used for any needs, not just food purchases. The flexibility was intentional. Food insecurity occurs due to a lack of money and resources (by definition), and we believe participants know best how to allocate resources to address their needs. The qualitative feedback indicated that participants valued this autonomy, using the funds to purchase non-food necessities, offset rising food prices, or supplement existing food assistance benefits.
It is important not to overstate the intervention’s impact on food insecurity or cessation-related outcomes. Although the intervention was associated with a reduction in the severity of food insecurity, the amount was not sufficient to shift overall food security status. Among food-insecure participants, 55% reported very low food security, which is a severe form of food insecurity characterized by reduced food intake and disrupted in eating patterns. This is a stark contrast to the 38% of food insecure households in the U.S. who experience very low food security. 3 At follow-up, the prevalence of very low food security declined to from 57% to 36% in the intervention group, while the control group showed limited change (from 52% to 48%). At baseline, 44% reported making a quit attempt 3 months before the intervention and smoked an average of 13 cigarettes a day. At follow-up, those in the intervention group were more likely to report having made a quit attempt (from 50% to 79%) and smoked fewer cigarettes (from 13 to 7 cigarettes per day) compared to the control group where there was limited change in the prevalence of quit attempts (from 37% to 44%) and in cigarettes per day (from 13 to 12). While these findings represent an encouraging signal that providing flexible funds may support cessation behaviors, results must be interpreted cautiously as policy and economic shifts occurring during the study may have muted larger effects. However, these findings suggest that small but meaningful flexible supports can make a tangible difference.
The Letter raises thoughtful questions about the viability of financial incentives in behavioral change. Although our study’s cash transfer intervention was not incentive based, we agree that financial mechanisms warrant discussion in light of the available evidence. A 2025 Cochrane review concluded with high-certainty that incentives—also called conditional cash transfers—are effective at improving smoking cessation rates. 19 From a health economics and resource allocation perspective, concerns about viability of financial interventions (whether conditional or unconditional) should be weighed against cost-effectiveness. Tobacco use drives an estimated $225 billion USD 20 and food insecurity an additional $78 billion USD 21 in healthcare expenditures each year. It is estimated that reducing tobacco use by just 1% would result in Medicaid savings of $2.5 billion USD in the following year alone. 22 Access to financial intervention programs to quit smoking may thus be more limited by political will, policy priorities, and healthcare reimbursement structures, rather than individual preference, personal apprehensions, and administrative burden. For instance, the Supplemental Nutrition Assistance Program (SNAP) is the largest and most critical program to guard against food insecurity in the U.S. SNAP is effectively a means-tested cash transfer program (restricted to food purchases) that carries administrative burdens alongside substantial benefits. Participation in SNAP has been demonstrated to increase household food security, 23 reduce healthcare costs,24,25 and reduce symptoms of psychological distress. 26 Although the current level of benefits (calculated under the USDA’s Thrifty Food Plan) are insufficient in nearly all U.S. counties, 27 food insecurity would be considerably higher without the program. Overall, SNAP illustrates that cash-based approaches are not only feasible, but also scalable and integral components of existing safety net systems.
We fully agree on the value of systems thinking to address complex and intersecting public health challenges. Yet, it is important to distinguish a systems perspective from multicomponent, intersectional or syndemic approaches to intervention. There may be overlap for some situations, such as a multicomponent intervention for an intersectional intervention to address a syndemic using system dynamics (systems thinking). However, there are also cases where these are distinct, such as system dynamics being applied to a single disease without consideration or need for explanation of co-occurring diseases in a population and hence not addressing a syndemic. Furthermore, multicomponent and intersectional interventions can address complex public health challenges that do not involve syndemics, systems thinking, or more generally systems science methods. Hence, nutritional therapies focused on specific micronutrients or culinary education may be valuable components of comprehensive health promotion interventions but are not inherently systems-based unless they explicitly account for system dynamics, such as feedback loops, interactions among sectors, or structural leverage points. A systems science approach might entail mapping causal pathways and identifying feedback mechanisms and accumulations that sustain the cycle between food insecurity and tobacco use. Such methodologies can reveal where smaller but strategic changes might yield the greatest population-level impact. Systems and syndemic perspectives complement one another, as both acknowledge complexity, and systems science provides a set methods for systematically modeling and testing where to intervene most effectively.
Addressing food insecurity and tobacco use requires attention to both structural context and individual experience. While disease-agnostic structural interventions will have greatest impact to population health, person-centered interventions remain essential for understanding mechanisms and improving implementation. Our study demonstrated the feasibility and acceptability of an integrated approach that offered modest but flexible supports, with some encouraging findings regarding reduction in food insecurity severity and cessation-related changes. Moving forward, systems science methods can help refine and target specific intervention leverage points alongside risk-benefit considerations of such interventions (eg, identifying unintended consequences). We appreciate the Letter to the Editor for engaging thoughtfully with our work and for underscoring the shared commitment to advancing systems-informed solutions for complex public health challenges.
