The Societal Costs of Food Insecurity: Implications for Managed Care Strategies

ABSTRACT

Objectives: To estimate the direct health care costs and indirect economic implications of food insecurity among working-age adults in the US from a societal perspective.

Study Design: We conducted a retrospective longitudinal cohort study using the 2016-2017 and 2021-2022 Medical Expenditure Panel Survey.

Methods: Our sample included 17,524 working-age adults. Our outcomes included health care spending, employment status, and income. The primary independent variable was food insecurity. Lagged dependent variable models assessed the association between food insecurity in year 1 and outcomes in year 2, adjusting for baseline characteristics.

Results: Our lagged dependent variable models showed that food insecurity was not associated with total health care spending in year 2, but it was associated with a $132 (95% CI, $25-$239) increase in emergency department spending. Although food insecurity was not associated with employment, it was associated with increased likelihood of having a seasonal job or a temporary job and missing work due to illness in year 2 by 2.4 (95% CI, 1.0-3.8), 3.7 (95% CI, 1.1-6.3), and 4.8 (95% CI, 1.9-7.7) percentage points, respectively. Furthermore, food insecurity was associated with a decrease of $2521 (95% CI, –$4129 to –$914) in annual individual-level income in year 2, which was primarily driven by a decline in wages of $2030 (95% CI, –$3438 to –$621).

Conclusions: Food insecurity was associated with economic burdens beyond medical spending, particularly through lost income and employment instability. These findings suggest that food insecurity is linked to economic instability while placing targeted demands on the health system, underscoring the importance of addressing food insecurity as a policy and public health priority.

Am J Manag Care. 2026;32(6):In Press

Takeaway Points

• Food insecurity is associated with greater emergency department spending.
• Food insecurity is also associated with higher levels of job instability and income insecurity.
• These findings underscore the broader economic implications of food insecurity.

Food insecurity, defined as limited or uncertain access to adequate food necessary for an active, healthy life, remains a critical public health challenge in the US.¹ In 2022, approximately 12.8% of US households experienced food insecurity,¹ reflecting its widespread impact across communities. Although research has established strong associations between food insecurity and adverse physical and mental health outcomes, including higher rates of chronic illness, depression, anxiety, and psychological distress,^2-10 the broader economic implications of food insecurity are less well understood.^11,12

Most existing literature has focused on the direct health care costs of food insecurity, highlighting increased use of medical services and higher overall expenditures.^13-15 However, food insecurity may also impose substantial indirect economic burdens, including employment instability, reduced work attendance, and lower income. These often-overlooked indirect costs have significant implications for population health management and total cost of care when viewed from a societal perspective.

As managed care continues to shift toward value-based payment and risk-adjusted delivery models, there is increasing emphasis on addressing upstream social determinants of health that influence both health outcomes and health care spending. Food insecurity is a central example of such determinants. From a societal perspective, managed care is expected to align health care delivery with broader public goals: reducing preventable utilization, sustaining workforce participation, and promoting long-term economic resilience. Addressing food insecurity, therefore, represents not only a public health priority but also an opportunity for managed care systems to advance whole-person care.

To help fill this gap, we used nationally representative longitudinal data to quantify both the direct health care costs and the indirect economic implications of food insecurity among working-age adults in the US. Specifically, we examined how food insecurity in one year (year 1) is associated with health care spending, employment outcomes, and income in the subsequent year (year 2) using lagged models to account for potential bidirectional associations.

METHODS

Study Sample

We conducted a retrospective longitudinal cohort study using panel data from the 2016-2017 and 2021-2022 Medical Expenditure Panel Survey (MEPS). MEPS is a longitudinal cohort study conducted by the Agency for Healthcare Research and Quality and provides a representative sample of the noninstitutionalized civilian US population. The survey follows the same cohort of households over a 2-year period and collects information on demographic and socioeconomic characteristics, health status, and health care utilization. We specifically selected the 2016-2017 and 2021-2022 panels because these were the only periods in which MEPS included data on food insecurity. Because constructing a panel requires 2 consecutive years of data for each cohort, only these 2 cycles were eligible for inclusion.

MEPS is reviewed and approved annually by the Westat Institutional Review Board (IRB). Because our study used deidentified publicly available data, no additional IRB approval was required.

Using the longitudinal data file, we first identified US adults (aged 18-60 years). We then excluded observations due to incomplete information on food insecurity and observations with missing data on other variables. After these exclusions, our final sample included 17,524 US working-aged adults with complete information.

Outcomes

We evaluated both the direct and indirect costs of food insecurity. In economic evaluations, direct costs refer specifically to medical expenses, whereas indirect costs reflect broader economic consequences such as lost productivity and reduced earning potential. Following this conceptual framework, we distinguished between direct and indirect costs accordingly. For direct costs, we measured health care spending, including total spending and services-specific spending. Specifically, data on health care spending were collected through household interviews with respondents from the MEPS Household Component (HC). Then the MEPS Medical Provider Component (MPC) supplements and validates information reported in the HC through telephone interviews and survey materials mailed to medical providers and pharmacies named by HC respondents. The MPC sample includes hospitals and hospital-based physicians, home health care agencies, office-based physicians, and pharmacies. MPC data are used to replace expenditure data reported by HC respondents with data reported by their providers because the latter data are generally more complete and less prone to reporting errors. MPC data are also used as an imputation source for item nonresponse to reduce bias in survey estimates of medical spending.

For indirect costs, we analyzed employment status and income. Employment status was examined across various categories: being employed, being self-employed, holding multiple jobs, having a seasonal job, having a temporary job, and missing work due to illness. Income was assessed in total and by specific types, including wage income, business income, sales income, dividend income, interest income, Social Security income, pension income, trust/rent income, and other forms of income. HC data for each household member are collected on most types of taxable and nontaxable income. In addition, long-form filers are asked about tax refunds, alimony, business income, sales of assets, and trusts. Then MEPS income data are carefully edited to fill in for missing or incomplete data prior to public release. We analyzed health care spending and income as continuous outcomes and employment status as a categorical outcome. Health care spending and income were adjusted to 2021 US$ using the Consumer Price Index.

Primary Independent Variable

The primary independent variable was food insecurity, measured using the 10-item US Department of Agriculture (USDA) Adult Food Security Survey Module with a 30-day look-back period.¹² The USDA methodology assigns 1 point for each affirmative response, which produces a raw score ranging from 0 to 10, with higher scores indicating greater food insecurity. Details of the 10 items are available in the eAppendix (available at ajmc.com). Following standard USDA scoring,¹⁶ households with raw scores between 3 and 10 were classified as food insecure, reflecting high levels of anxiety about food sufficiency, insufficient food availability, or limited food intake due to affordability issues.

Covariates

To adjust for differences in sample characteristics, we included the following variables: age, sex, race/ethnicity, marital status, education, health insurance, US census region, presence of chronic conditions, and functional limitations. Nineteen chronic conditions commonly encountered in clinical practice were considered based on criteria established by a working group on multiple chronic conditions within the HHS Office of the Assistant Secretary for Health.¹⁷ These conditions were arthritis, asthma, cancer, cardiac arrhythmias, chronic kidney disease, chronic obstructive pulmonary disease, congestive heart failure, coronary artery disease, dementia, depression, diabetes, hepatitis, HIV, hyperlipidemia, hypertension, osteoporosis, schizophrenia, stroke, and substance use disorders. Chronic conditions were grouped into 4 levels: 0, 1 or 2, 3 to 5, and 6 or more. Functional limitations were assessed using 6 standardized questions that address difficulties with seeing, hearing, memory or concentration, walking, self-care, and performing errands due to physical, mental, or emotional conditions.¹⁸ These questions align with HHS’ standardized survey measures for disability.¹⁹ Consistent with a prior study, functional limitations were categorized as none (no reported difficulty), moderate (1-2 difficulties), and severe (≥ 3 difficulties).²⁰

Statistical Analyses

We first estimated baseline sample characteristics and outcomes between working-age adults with and without food insecurity. Then we estimated adjusted outcomes using logistic regression models for binary outcomes or linear regression models for continuous outcomes. Due to the skewed distribution of health care spending, we initially considered 2-part models; however, convergence issues with certain spending measures led us to use Poisson regression as an alternative approach. Drawing causal inferences from survey data is challenging, as several major issues, including reverse causality, can lead to biased estimates. To address this issue, we used a lagged dependent variable model.¹² Although this approach is not a formal difference-in-differences design, it follows a similar conceptual framework by assessing how prior exposure to food insecurity relates to subsequent outcomes. Specifically, we examined the relationship between food insecurity in year 1 and outcomes in year 2, adjusting for individual-level covariates and the baseline value of each outcome measured in year 1. We then used predictive margins to calculate the mean adjusted values of the outcomes for each group and conducted postestimation tests to estimate the differences in outcomes between those with and without food insecurity. Because our data set includes the COVID-19 pandemic period, we conducted a sensitivity analysis using only the 2016-2017 panel to assess the robustness of our findings in a prepandemic context. However, the lagged dependent variable model may be vulnerable to unobserved heterogeneity. Thus, we also conducted the analysis using a fixed-effects model, which allowed us to account for unobserved, time-invariant individual characteristics. To account for differences in sample characteristics between individuals with and without food insecurity, we estimated inverse probability of treatment weights (IPTW) based on individual-level characteristics and then conducted the lagged model incorporating these weights. For all analyses, we used survey weights to adjust the sample characteristics to be representative of the US adult population. We also accounted for the complex survey design in our SE estimation. The data were analyzed using Stata 16.1 (StataCorp LLC).

RESULTS

Our final sample included 17,524 working-age adults, which represented a population of 201,729,816 (Table 1). Of those, 15,058 were classified as food secure and 2466 as food insecure in year 1. Compared with food-secure adults, food-insecure adults were more likely to be female, be Hispanic, be non-Hispanic Black, have lower educational attainment, lack health insurance or have public insurance, and experience chronic conditions and functional disabilities, and they were less likely to be married and have private health insurance. We compared sample characteristics before and after excluding observations with missing values. The results showed minimal differences, suggesting that the missing data were likely missing at random and unlikely to introduce bias.

Our lagged dependent variable model identified certain direct costs associated with food insecurity (Table 2). Specifically, food insecurity in year 1 was not significantly associated with total health care spending in year 2 (adjusted differences in total health care spending between adults with and without food insecurity). Similarly, we observed no significant differences in inpatient admissions, outpatient visits, or prescription drug spending. However, adults with food insecurity had significantly higher emergency department (ED) spending ($132 higher; 95% CI, $25-$239) vs those without food insecurity.

Our lagged dependent variable model also identified indirect costs associated with food insecurity (Table 3). Although food insecurity in year 1 was not significantly associated with overall employment rates in year 2, it was associated with more precarious employment outcomes. Specifically, compared with adults without food insecurity, adults with food insecurity were 2.4 percentage points (95% CI, 1.0-3.8) more likely to hold seasonal jobs, 3.7 percentage points (95% CI, 1.1-6.3) more likely to hold temporary jobs, and 4.8 percentage points (95% CI, 1.9-7.7) more likely to miss work due to illness in year 2. No significant differences were observed in being employed, being self-employed, and having multiple jobs. Additionally, adults with food insecurity had significantly lower annual total individual-level income in year 2—$2521 less on average (95% CI, –$4129 to –$914)—compared with adults without food insecurity. This reduction was primarily driven by a decline in wages of $2030 (95% CI, –$3438 to –$621). Other income sources, including income from sales, dividends, interest, Social Security, pensions, rental or trust funds, and miscellaneous sources, did not show statistically significant differences.

Our sensitivity analysis, which excluded data from the COVID-19 period, demonstrated that the main findings remained largely consistent. Although the association between food insecurity and income was no longer statistically significant, the magnitude and direction of the coefficient were similar. To account for skewness in the income distribution, we reestimated our models using log-transformed income. These models yielded consistent results regardless of whether 2021 data were included, with comparable coefficient estimates in terms of magnitude, direction, and statistical significance. Furthermore, results from both fixed-effects and IPTW-adjusted models were generally consistent with our main findings (eAppendix), further supporting the robustness of our conclusions.

DISCUSSION

Our study provides empirical evidence on the direct and indirect economic implications of food insecurity among working-age adults in the US. First, we observed direct costs of food insecurity associated with health care spending primarily through increased or more intensive ED visits. Food-insecure adults are particularly vulnerable to additional social risk factors, which are linked with poorer medical conditions and more frequent ED visits.^11,14 These factors are associated with greater ED spending and further exacerbate the financial burden of care. However, we did not observe a significant increase in total health care spending. This finding offers a more nuanced perspective on the existing literature: Although previous studies have suggested that food insecurity contributes to higher total health care spending,^13,14 recent evidence suggests that much of this association diminishes once baseline comorbidities are taken into account.²¹ This suggests that higher health care expenditures among food-insecure individuals are primarily driven by poor baseline health rather than the independent effect of food insecurity itself.¹⁵

We also observed that the significant economic implications of food insecurity extend beyond health care.¹² Although prior research has found that job loss can lead to food insecurity,²² our study shows how food insecurity is associated with employment and income. Despite overall employment rates remaining unchanged, food-insecure adults were more likely to hold less stable jobs, such as seasonal or temporary positions, which could result in lower income. Additionally, food-insecure adults were more likely to miss work due to illness, potentially reducing productivity and further impacting income. These factors not only perpetuate inequities in health and health care but also reinforce inequalities in economic and social opportunities.

Our study findings suggest that food insecurity is linked with broader societal challenges that impact public health and economic stability. Addressing food insecurity as both a public health and economic issue is essential to breaking this cycle and promoting social and economic equity. Policy makers should adopt a multifaceted approach to mitigate its contributions. Expanding programs such as the Supplemental Nutrition Assistance Program and the Special Supplemental Nutrition Program for Women, Infants, and Children can help alleviate food insecurity and its associated burdens. Improving benefit levels, expanding eligibility, and streamlining administrative processes could significantly support working-age adults experiencing economic instability. At the same time, more coordinated efforts across public health, social services, and workforce systems may better position these sectors to address the intertwined challenges of food insecurity, chronic illness, and unstable employment.

These findings also have important implications for managed care, which often assumes financial risk for population health under value-based care models. Although our study did not find clear causal evidence linking food insecurity to outcomes within 1 year, the results indicate that food insecurity functions as an upstream social determinant that is associated with preventable ED visits and disrupts care coordination. From a societal perspective, the costs of food insecurity—through both increased health care utilization and reduced economic productivity—ultimately shape policy priorities, payment structures, and expectations for managed care. Efforts to improve efficiency, quality, and cost control within managed care may benefit from systematically integrating social risk screening into clinical workflows, building partnerships with community organizations, and investing in targeted interventions that address food insecurity. Health insurers and integrated delivery systems have already begun implementing such strategies, including screening programs, medically tailored meal delivery, and collaborations with food banks. Given the substantial societal and economic burden of food insecurity, managed care has both a responsibility and a strategic interest in addressing this challenge. However, sustainable progress will require broader policy support and public investment to reduce food insecurity at scale.

Limitations

First, our sample was limited to the noninstitutionalized US population, excluding incarcerated people, nursing home residents, and those in residential treatment. This may lead to an underestimation of our findings, as these populations are more likely to be affected by food insecurity. Second, our measures relied on self-reporting, which may be subject to measurement error and could potentially bias our estimates. However, MEPS applies rigorous data review and imputation procedures to minimize biases related to self-reporting, which may reduce concerns about data accuracy. Third, the income data from MEPS were top-coded, potentially leading to an understatement of our findings. Fourth, although our analyses adjusted for potential confounders, residual confounding remains a concern. The lagged dependent variable model accounts for baseline outcomes, offering some protection against reverse causality. However, unmeasured time-varying confounders may still bias the results. Fifth, our study period included the COVID-19 pandemic, which may have introduced confounding factors such as expanded government relief programs, changes in employment patterns, and disruptions in health care utilization. Although we adjusted for survey year and conducted sensitivity analyses excluding 2021 data, the widespread and multifaceted nature of the pandemic likely limited our ability to fully account for its impact. Therefore, residual confounding related to pandemic-era policy shifts and care delivery disruptions may persist. Finally, our analyses are associational, so the results should not be interpreted as causal. Further research is needed to provide robust evidence of a direct causal relationship between food insecurity and the outcomes examined. Establishing such evidence may enhance the authority, incentives, and institutional capacity of the managed care community to address this complex public health challenge independently and at scale.

CONCLUSIONS

Our findings highlight the direct and indirect economic implications of food insecurity among working-age US adults. Food insecurity was associated with greater ED spending, increased job instability, increased likelihood of missing work due to illness, and significant reductions in income. These indirect costs, often hidden from traditional health care spending, have critical implications for managed care organizations operating under value-based models.

By showing how food insecurity is associated with health care costs and economic productivity, our study underscores the importance of integrating social risk factors into population health strategies. Addressing food insecurity is not ancillary to managed care; rather, it is central to achieving cost containment, improving health outcomes, and advancing equity. Managed care organizations, policy makers, and organizational partners should prioritize scalable interventions that reduce food insecurity as part of comprehensive efforts to improve workforce stability and reduce avoidable health care expenditures.

Author Affiliations: Department of Health Policy and Management, College of Health Science, Korea University (SP), Seoul, Republic of Korea; Thompson School of Social Work and Public Health, University of Hawaiʻi at Mānoa (ANO), Honolulu, HI; Department of Health Policy and Management, School of Public Health, University of Maryland (JC), College Park, MD; Department of Health Management and Policy, Herbert Business School, University of Miami (KM), Coral Gables, FL; Department of Health Policy and Management, Fielding School of Public Health, and Latino Policy and Politics Institute, University of California, Los Angeles (AVB), Los Angeles, CA.

Source of Funding: None.

Author Disclosures: Dr Chen is supported by grants P30AG097158 and RF1AG083175. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (SP, ANO, JC, KM, AVB); acquisition of data (SP); analysis and interpretation of data (SP, KM, AVB); drafting of the manuscript (SP, ANO, JC, KM); critical revision of the manuscript for important intellectual content (SP, ANO, JC, KM, AVB); statistical analysis (SP); and supervision (SP, ANO).

Address Correspondence to: Sungchul Park, PhD, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul, Republic of Korea 02841. Email: sungchul_park@korea.ac.kr.

REFERENCES

1. Rabbitt MP, Hales LJ, Burke MP, Coleman-Jensen A. Household Food Security in the United States in 2022. US Department of Agriculture Economic Research Service; October 2023. Accessed May 1, 2026. https://ers.usda.gov/sites/default/files/_laserfiche/publications/107703/ERR-325.pdf?v=68979

2. Gundersen C, Ziliak JP. Food insecurity and health outcomes. Health Aff (Millwood). 2015;34(11):1830-1839. doi:10.1377/hlthaff.2015.0645

3. Berkowitz SA, Palakshappa D, Seligman HK, Hanmer J. Changes in food insecurity and changes in patient-reported outcomes: a nationally representative cohort study. J Gen Intern Med. 2022;37(14):3638-3644. doi:10.1007/s11606-021-07293-4

4. Berkowitz SA, Baggett TP, Wexler DJ, Huskey KW, Wee CC. Food insecurity and metabolic control among U.S. adults with diabetes. Diabetes Care. 2013;36(10):3093-3099. doi:10.2337/dc13-0570

5. Seligman HK, Jacobs EA, López A, Tschann J, Fernandez A. Food insecurity and glycemic control among low-income patients with type 2 diabetes. Diabetes Care. 2012;35(2):233-238. doi:10.2337/dc11-1627

6. Crews DC, Kuczmarski MF, Grubbs V, et al; Centers for Disease Control and Prevention Chronic Kidney Disease Surveillance Team. Effect of food insecurity on chronic kidney disease in lower-income Americans. Am J Nephrol. 2014;39(1):27-35. doi:10.1159/000357595

7. Leung CW, Kullgren JT, Malani PN, et al. Food insecurity is associated with multiple chronic conditions and physical health status among older US adults. Prev Med Rep. 2020;20:101211. doi:10.1016/j.pmedr.2020.101211

8. Silverman J, Krieger J, Kiefer M, Hebert P, Robinson J, Nelson K. The relationship between food insecurity and depression, diabetes distress and medication adherence among low-income patients with poorly-controlled diabetes. J Gen Intern Med. 2015;30(10):1476-1480. doi:10.1007/s11606-015-3351-1

9. Mangurian C, Sreshta N, Seligman H. Food insecurity among adults with severe mental illness. Psychiatr Serv. 2013;64(9):931-932. doi:10.1176/appi.ps.201300022

10. Pourmotabbed A, Moradi S, Babaei A, et al. Food insecurity and mental health: a systematic review and meta-analysis. Public Health Nutr. 2020;23(10):1778-1790. doi:10.1017/S136898001900435X

11. Park S, Chen J, Bustamante AV. Adverse consequences of food insecurity among U.S. adults beyond health outcomes. Am J Prev Med. 2024;66(1):146-153. doi:10.1016/j.amepre.2023.09.003

12. Park S, Ortega AN, Chen J, Mortensen K, Bustamante AV. Association of food insecurity with health, access to care, affordability of care, financial burden of care, and financial hardships among US adults during the COVID-19 pandemic. Public Health. 2024;230:183-189. doi:10.1016/j.puhe.2024.02.028

13. Berkowitz SA, Basu S, Meigs JB, Seligman HK. Food insecurity and health care expenditures in the United States, 2011-2013. Health Serv Res. 2018;53(3):1600-1620. doi:10.1111/1475-6773.12730

14. Berkowitz SA, Seligman HK, Meigs JB, Basu S. Food insecurity, healthcare utilization, and high cost: a longitudinal cohort study. Am J Manag Care. 2018;24(9):399-404.

15. Johnson KT, Palakshappa D, Basu S, Seligman H, Berkowitz SA. Examining the bidirectional relationship between food insecurity and healthcare spending. Health Serv Res. 2021;56(5):864-873. doi:10.1111/1475-6773.13641

16. Food security in the U.S. — definitions of food security. US Department of Agriculture Economic Research Service. Updated March 30, 2026. Accessed May 1, 2026. https://www.ers.usda.gov/topics/food-nutrition-assistance/food-security-in-the-us/definitions-of-food-security

17. Park S, Marcondes FO. The effect of Medicare eligibility on diagnosis of chronic conditions. Am J Manag Care. 2024;30(2):96-100. doi:10.37765/ajmc.2024.89497

18. Xie Z, Hong YR, Tanner R, Marlow NM. People with functional disability and access to health care during the COVID-19 pandemic: a US population-based study. Med Care. 2023;61(2):58-66. doi:10.1097/MLR.0000000000001765

19. U.S. Department of Health and Human Services implementation guidance on data collection standards for race, ethnicity, sex, primary language, and disability status. HHS. May 1, 2026. https://aspe.hhs.gov/sites/default/files/private/pdf/76331/index.pdf

20. Park S, Stimpson JP. Health care expenses and financial hardship among Medicare beneficiaries with functional disability. JAMA Netw Open. 2024;7(6):e2417300. doi:10.1001/jamanetworkopen.2024.17300

21. Dean EB, French MT, Mortensen K. Food insecurity, health care utilization, and health care expenditures. Health Serv Res. 2020;55(suppl 2):883-893. doi:10.1111/1475-6773.13283

22. Mabli J, Monzella K, Franckle RL, Delgado PL. Food insecurity transitions and changes in employment and earnings. Am J Prev Med. 2023;64(3):368-376. doi:10.1016/j.amepre.2022.09.028