Journal Logo

Feature Articles

Original Research: The Relationship Between Food Insecurity and Cost-Related Medication Nonadherence in Older Adults: A Systematic Review

Caouette, Shaunna MSN, RN, AGPCNP-BC; Boss, Lisa PhD, RN, CNS, CEN, CNE; Lynn, Monika MSN, APRN, AGNP-C, CMSRN

Author Information
AJN, American Journal of Nursing: June 2020 - Volume 120 - Issue 6 - p 24-36
doi: 10.1097/01.NAJ.0000668732.28490.c1
  • Free
  • Videos

By 2030, the number of Americans ages 65 years and older is projected to grow from 56 million to 74 million, constituting 20% of the population—one in every five people.1 Currently, more than 60% of this age group have at least two chronic illnesses2; and more than 40% require five or more prescription drugs daily, resulting in higher out-of-pocket health care costs.3, 4 In 2012, prescription drug costs for a person with more than five chronic illnesses averaged $8,263 annually.5 And the cost of medications continues to rise as newer drugs are made available.5 Not only are many older adults burdened by living with chronic illnesses and high drug costs, but almost 10% of this population lives at or below the federal poverty level.6 In 2017, according to a report by Feeding America, nearly 8 million older adults were food insecure.7

Food insecurity, which may be defined as having limited access to nutritional and safe foods because of insufficient financial resources, can profoundly influence health.8 Limited access to healthful food can have severe health consequences for older adults, contributing to the worsening of chronic illnesses, especially those related to diet, such as diabetes, cardiovascular disease, hypertension, and obesity.9-11 Moreover, people who are food insecure may repeatedly have to choose between purchasing essential medications or meeting other basic needs such as food. Such trade-off decisions can result in cost-related medication nonadherence (CRN) behaviors, such as skipping or reducing doses, delaying or failing to fill or refill prescriptions, and stopping medications altogether. Unable to comply with their medical regimens, these older adults face higher risks of poor health outcomes, as well as increased use of health care services (such as ED and inpatient services) and higher health care costs.9, 12, 13 Indeed, recent estimates put the national costs associated with food insecurity at $160 billion in 2014,14 a substantial increase from $90 billion in 2005.15 A further complication is that, as unprecedented numbers of older adults reach retirement age, many will leave their employer-provided health insurance and become Medicare beneficiaries, straining government funds further still.

Food insecurity and CRN are significant public health issues that demand a better understanding of their causes and consequences, so that effective actions can be developed and implemented to address them. This systematic review sought to explore the relationship between food insecurity and CRN in U.S. adults ages 60 years and older.


A comprehensive electronic literature search was conducted using the CINAHL, PsycInfo, and PubMed databases. The main search terms included, but were not limited to, food assistance, food insecurity, food security, food supply, medication adherence, medication compliance, medication nonadherence, patient compliance, treatment adherence, treatment compliance, aged, elder, geriatrics, senior, and older adult. Although hunger and food insecurity are two distinctly different conditions,16 the terms are often used interchangeably, so we also included hunger as a search term. The search targeted original peer-reviewed primary research studies that involved only human subjects who were 60 years of age and older and lived in the United States. Although no time period was specified for the search, all studies that met the above criteria were published between 2010 and 2018.

The initial search yielded 191 articles: 148 from PubMed, 35 from CINAHL, and three from PsycInfo; another five articles were identified through a manual search of references. After removing 23 duplicates, the 168 remaining records were reviewed and evaluated for a relationship between food insecurity and CRN in older community-dwelling adults. Those that involved hospitalized or institutionalized patients were excluded, as well as any articles that weren't primary research or were written in a language other than English. Studies were also excluded if the subjects were outside the identified age group or lived outside the United States. We excluded other countries from this systematic review because of variations in health care coverage (universal health care versus private insurance) and prescription drug coverage. A full-text review of the remaining 18 articles was conducted. Six articles met all the inclusion criteria for this systematic review (see Figure 1).

Figure 1.:
PRISMA Flow Diagram of Studies


Study designs. Four studies were cross-sectional,17-20 one was longitudinal,13 and one was both cross-sectional and longitudinal.21 All data were collected from large, population-based studies with sample sizes ranging from 243 to 10,401 participants.17, 21 Subject data were collected from the Georgia Advanced Performance Outcomes Measures Project 6 and the Older Americans Act Nutrition Program (OAANP)13, 18, 21 or from the National Health Interview Survey (NHIS).17, 19, 20

Study samples. Three of the studies included subjects from the state of Georgia who were either participating in that state's OAANP or were on its waiting list.13, 18, 21 The other three studies included subjects participating in the NHIS, which collects data from participants nationwide.17, 19, 20 Only one study obtained data from a general population of older adults who weren't linked to a food program.17 The other five studies included low-income subjects who were participating in or wait-listed for a food assistance program, either the OAANP13, 18, 21 or the Supplemental Nutrition Assistance Program (SNAP).19, 20

Three studies included subjects 60 years of age and older13, 18, 20; the other three included subjects 65 years of age and older.17, 19, 21 Mean age ranged from 74 to 75 years.17, 18 Overall, proportionally more subjects were female (range, 58%17 to 77%21), white, and living on a low annual household income (less than $20,000 and/or such that they were eligible for food assistance). About half of all subjects had a high school education or less.

Two studies included only subjects diagnosed with borderline or type 2 diabetes.19, 21 One study included subjects with “any nutrition-related chronic condition (heart disease, diabetes, or stroke).”20 The remaining three studies did not focus on subjects with a specific disease13, 17, 18; commonly reported illnesses included obesity,17 cardiovascular disease, diabetes, stroke, arthritis, and osteoporosis.13, 18 Overall, in all but one study, a large proportion of subjects self-reported fair to poor health.13, 18-21

Definitions and measurements. Definitions for food insecurity and CRN were similar across all six studies and are reported in Table 1.13, 17-22 (Although two studies did not explicitly define one or both terms,19, 21 they referred to earlier published studies by members of their research team that did.) The definitions for food insecurity generally followed those developed by the American Institute of Nutrition22 (now the American Society for Nutrition) and adopted by the U.S. Department of Agriculture.23 The definitions for CRN were generally based on prior expert work by Piette and colleagues24 and Briesacher and colleagues.25

Table 1.:
Relationship Between Food Insecurity and CRN in Older Adults
Table 1-A.:
Relationship Between Food Insecurity and CRN in Older Adults
Table 1-B.:
Table 1-C.:
Table 1-D.:
Table 1-E.:

Three studies used the validated 10-item U.S. Adult Food Security Survey Module, which categorizes subjects into four levels of food security—high, marginal, low, and very low.17, 19, 20 One of these studies made use of all four categories17; the other two combined marginal and lower categories to represent food insecurity or “threatened by hunger.”19, 20 The remaining three studies used the validated six-item U.S. Household Food Security Survey Module to categorize subjects as either food secure or food insecure.13, 18, 21 Both surveys asked subjects about their experiences over the past 30 days.

All six studies used some variation of three to five yes/no questions to measure CRN. Afulani and colleagues asked about subjects' experiences over the past 12 months using a five-item survey adapted from the Agency for Healthcare Research and Quality (AHRQ) Medical Expenditure Panel Survey (MEPS), which is a validated instrument.17 Three studies did not use validated tools to measure CRN.13, 18, 21 These studies assessed CRN over the past 30 days using items adapted from prior expert research and based on documented CRN behaviors; however, Sattler and colleagues did not report which specific questions were used.21 The remaining two studies used the same three-question survey to determine adherence to a prescription regimen19, 20; its validated items were also based on the AHRQ's MEPS.

Relationship between food insecurity and CRN. All six studies used multivariate logistics regression analysis to examine the relationship between food insecurity and CRN and to control for multiple confounding variables. Two studies found a significant dose-response relationship between food insecurity and CRN,17, 18 and a third reported an increased risk of CRN in individuals with persistent food insecurity.13 Srinivasan and Pooler also found that SNAP subjects who were food insecure or threatened by hunger were significantly less likely to participate in CRN behaviors than SNAP-eligible nonparticipants.20

Demographics. Five studies reported that subjects who reported greater food insecurity and CRN were significantly more likely to be younger and to have lower incomes than those reporting less food insecurity and CRN.13, 17-19, 21 In four studies, greater food insecurity and CRN were associated with having less education,13, 18, 19, 21 although this finding reached statistical significance in only one study.18 Regarding age, Afulani and colleagues found that CRN decreased significantly as age increased, with the old-old (adults ages 85 years and older) reporting the fewest CRN behaviors.17 Pooler and Srinivasan found that adults ages 65 to 69 years reported significantly more CRN behaviors than their older counterparts.19 Two studies found that the young-old (adults ages 60 to 84 years) were more likely to report CRN behaviors.13, 18

Five studies described subjects with food insecurity, CRN, or both as self-reporting poorer health and more chronic illnesses.13, 18-21 Sattler and Lee found that “increasing severity of CRN” was significantly associated with poorer health, multiple comorbidities, and polypharmacy.13

Income and health insurance. Afulani and colleagues reported no association between deeper poverty levels and increased CRN.17 But subjects with incomes between 100% and 299% of the federal poverty level were almost twice as likely to have a higher risk of CRN than subjects with incomes below 100% or above 299% of that level. Afulani and colleagues also found that subjects who received Medicare plus Medicaid or another form of public health insurance were significantly less likely to report CRN than were those receiving Medicare alone. Srinivasan and Pooler found that SNAP participants with prescription drug coverage and lower out-of-pocket health care costs were significantly less likely to report CRN than were SNAP-eligible nonparticipants.20

Prescription costs. Higher out-of-pocket prescription costs were associated with significantly increased CRN behaviors in four studies.13, 18-20 Pooler and Srinivasan found that SNAP participants with $500 or higher out-of-pocket medical costs (including prescriptions) were more likely to engage in CRN compared with subjects with lower costs.19 Similarly, Bengle and colleagues reported that subjects with $101 or higher monthly prescription costs were more likely to engage in CRN than were those with lower or no costs.18


Relationship between food insecurity and CRN. The findings of this systematic review suggest a correlation between food insecurity and CRN in older adults. Although it's unclear whether there is a causal relationship between these two variables, two of the six studies did find a significant dose-dependent relationship, as noted above.17, 18 In the literature, at least one earlier study reported similar findings with regard to food insecurity and CRN in younger adults ages 18 to 64 years.26 But the consequences of these two variables tend to be more severe in older adults, who are more likely to face the complex interplay of factors such as chronic illnesses, mobility limitations, and the need for multiple medications.16, 27

Demographics. Although subjects' age ranges varied in the six studies, the findings suggest that young-old adults (those ages 60 to 84) with food insecurity are at higher risk for CRN than old-old adults (those ages 85 and older).13, 17-19 This general pattern has been noted in several earlier reports.16, 25, 28 But again, age groupings vary in the literature, and it's unclear which precise age range is most at risk. According to a report by Feeding America, among adults ages 60 and older, the threat of food insecurity is highest in those ages 60 to 64 (14.2%) and declines with increasing age, though it still affects 6.2% of those ages 80 and older.7 Lower rates of CRN in old-old adults might also be a result of this subgroup's lower food requirements, higher mortality rate, and increased access to nutritional and financial assistance.7 Further research is required to better understand the risks of food insecurity and CRN in specific age subgroups of the older population.

Sex was not found to be a statistically significant predictor of CRN. That said, three of the six studies found that women with food insecurity were more likely to report CRN than men, although these findings lacked significance.13, 17, 18 In other literature, the evidence regarding associations between sex, food insecurity, and CRN is inconsistent. Briesacher and colleagues reported no association between female sex and CRN in their literature review.25 Sharkey, who focused on homebound women ages 65 and older, reported no significant differences for food sufficiency status in relation to medication use.29 But Garcia and colleagues found that, among adults ages 50 and older who were food insecure, a significantly higher proportion of women reported CRN than did their male counterparts.12 Zhang and colleagues found significantly higher rates of CRN in women versus men across all age (18 years and older) and socioeconomic categories, with lower income generally associated with higher CRN.30

Income and health insurance. In all six studies, income and health insurance were identified as critical factors that influenced food insecurity and CRN in older adults. In other literature, Zivin and colleagues reported that the main driver for CRN was associated with the ability to pay for medications; other factors associated with CRN included lower net worth, lower income, and no or partial health insurance.31 It's interesting that Afulani and colleagues found that people living at between 100% and 299% of the federal poverty level were at higher risk for CRN than those living below 100% of that level.17 Comparable findings were reported in a national study of chronically ill adults ages 20 and older, which noted that people with incomes between 101% and 200% of the federal poverty level are ineligible for many food assistance programs and financial aid.27

Three of the six studies reported more CRN behaviors in older adults who either lacked health insurance or received only Medicare, compared with those with Medicare plus Medicaid or another private or public health plan.13, 17, 18 In keeping with Bakk,32 Sattler and colleagues noted that the Medicare Part D coverage gap may prompt people to engage in more CRN behaviors in an effort to restrict out-of-pocket spending.21 Elsewhere, Naci and colleagues reported that the odds of CRN increased by 20% between 2009 and 2011 for Medicare recipients with four or more chronic conditions.3 It's also relevant that over the next decade, the annual growth rate for Medicare Part D beneficiary costs is projected to rise from 2.3% to 4.4%.33

Prescription costs. In five of the six studies, older adults who took more prescription medications and had higher out-of-pocket drug costs were at higher risk for food insecurity and CRN.13, 17, 18, 20, 21 Dual beneficiaries who have Medicare plus Medicaid or other health care insurance have lower copays and increased access to health assistance programs, unlike those receiving only Medicare.17, 19 For dual beneficiaries, this “protective effect” may help to fill in financial gaps, giving them more disposable income for food and other basic needs.12

Five of the six studies included only subjects who participated in or were wait-listed for food assistance programs. An important finding was that, even with access to food aid, a lack of prescription drug coverage and higher out-of-pocket health care costs were predictive of increased CRN behaviors in subjects with food insecurity.19 In light of this, it's vital that health care providers be knowledgeable about costs of prescription drugs and possibly cheaper alternatives. Ultimately, lower drug prices and improved Medicare drug coverage are needed for older adults.

Practice implications. Our findings highlight the financial and physical burdens that food insecurity and CRN place on older community-dwelling adults. It's vital that health care professionals become more aware of and better educated about food insecurity and the populations at risk. Nurses and other clinicians should consider screening older adults for food insecurity, especially women, those with multiple chronic conditions, and those who receive only Medicare.

Besides identifying older adults with food insecurity, nurses and other clinicians must continue to educate patients regarding healthy food choices, and help them connect with appropriate community resources where they can receive nutritious low-cost food and other needed assistance.34 According to Gundersen and Ziliak, over 60% of older adults eligible for SNAP do not receive assistance.35 Reasons for this might include a lack of knowledge about food assistance programs or an inability to access such resources.16

Lower- or no-cost prescription plans have been shown to improve health outcomes.31 The findings of this review reinforce the importance of reforming Medicare prescription drug coverage, lowering drug costs, and improving financial aid for older adults. As patient advocates, nurses must advocate policy changes that will remove financial barriers and make medications more available and affordable for this population.

Food insecurity and CRN in older adults remain understudied, and there are gaps in the current literature that require further investigation. Future research should include longitudinal trials conducted with large samples of older Americans. Additional research will better inform us as to whether there is a causal relationship between food insecurity and CRN, and how changes to the predictors of these two variables influence long-term outcomes. Studies to clarify the interplay of factors such as race, ethnicity, and sex are also warranted.

Limitations. All studies used self-reported mailed surveys and questionnaires to gather subject data on demographics, health status, food insecurity, and CRN practices. Most studies excluded subjects for reasons such as illiteracy, inability to read and write English, or physical disabilities (such as blindness or cognitive issues), with the exception of Sattler and colleagues,21 who included subjects with dementia, including Alzheimer disease. Unfortunately, it's estimated that one in 10 U.S. adults ages 65 years and older has been diagnosed with Alzheimer disease.36 Although this is an important subgroup that should not be dismissed from this research, the methodology of self-administered questionnaires may have biased reporting, especially in those with cognitive disabilities. Recall error may also have limited the accuracy of some self-reported responses.

The Georgia studies were completed during the 2008-2009 economic recession, which may have contributed to increased reporting of food insecurity and CRN by some subjects.13, 18, 21 Three studies were completed with subjects drawn from the same larger sample of low-income older adults,13, 18, 21 and two studies were secondary analyses of original data from previous studies,13, 19 limiting the generalizability of the results. Two studies had the same lead author and this too may have risked bias in the data.13, 21


Our findings demonstrate an increased risk of poorer health outcomes in older adults faced with food insecurity and CRN. It's critical that health care providers and organizations find ways to help manage these risks, in order to improve health outcomes, reduce use of health care services, and lower long-term costs for older adults. As frontline health care professionals, nurses can be pivotal in taking the lead to develop appropriate interventions and programs for patients, advocate better government-funded health care, and work toward health policy reform in order to improve health outcomes in this vulnerable population.


1. Colby SL, Ortman JM. Projections of the size and composition of the U.S. population: 2014 to 2060. Washington, DC: U.S. Census Bureau; 2015 Mar. P25–1143. Current population reports;
2. Ward BW, et al. Multiple chronic conditions among US adults: a 2012 update. Prev Chronic Dis 2014;11:E62.
3. Naci H, et al. Medication affordability gains following Medicare Part D are eroding among elderly with multiple chronic conditions. Health Aff (Millwood) 2014;33(8):1435–43.
4. Wilson IB, et al. Physician-patient communication about prescription medication nonadherence: a 50-state study of America's seniors. J Gen Intern Med 2007;22(1):6–12.
5. Federal Interagency Forum on Aging-Related Statistics. Older Americans: key indicators of well-being. Washington, DC; 2016.
6. Fontenot K, et al. Income and poverty in the United States: 2017. Washington, DC: U.S. Census Bureau; 2018 Sep. P60–263. Current population reports;
7. Zillak JP, Gundersen C. The state of senior hunger in America in 2017: an annual report. Chicago: Feeding America 2019 May 14.
8. Office of Disease Prevention and Health Promotion. HealthyPeople 2020: food insecurity. n.d.
9. Bhargava V, Lee JS. Food insecurity and health care utilization among older adults in the United States. J Nutr Gerontol Geriatr 2016;35(3):177–92.
10. Lee JS. Food insecurity and healthcare costs: research strategies using local, state, and national data sources for older adults. Adv Nutr 2013;4(1):42–50.
    11. Seligman HK, et al. Food insecurity is associated with chronic disease among low-income NHANES participants. J Nutr 2010;140(2):304–10.
    12. Garcia SP, et al. Incremental health care costs associated with food insecurity and chronic conditions among older adults. Prev Chronic Dis 2018;15:E108.
    13. Sattler EL, Lee JS. Persistent food insecurity is associated with higher levels of cost-related medication nonadherence in low-income older adults. J Nutr Gerontol Geriatr 2013;32(1):41–58.
    14. Cook JT, Poblacion AP. The nourishing effect: ending hunger, improving health, reducing inequality. Appendix 2: estimating the health-related costs of food insecurity and hunger. Washington, DC: Bread for the World Institute; 2015. 2016 Hunger report—26th annual report on the state of world hunger;
    15. Shepard DS, et al. Hunger in America: suffering we all pay for. Washington, DC: Center for American Progress; 2011 Oct.
    16. Strickhouser S, et al. Food insecurity among older adults [2015 update]. Washington, DC: AARP Foundation 2014 Sep.
    17. Afulani P, et al. Food insecurity and health outcomes among older adults: the role of cost-related medication underuse. J Nutr Gerontol Geriatr 2015;34(3):319–42.
    18. Bengle R, et al. Food insecurity is associated with cost-related medication non-adherence in community-dwelling, low-income older adults in Georgia. J Nutr Elder 2010;29(2):170–91.
    19. Pooler JA, Srinivasan M. Association between Supplemental Nutrition Assistance Program participation and cost-related medication nonadherence among older adults with diabetes. JAMA Intern Med 2019;179(1):63–70.
    20. Srinivasan M, Pooler JA. Cost-related medication nonadherence for older adults participating in SNAP, 2013-2015. Am J Public Health 2018;108(2):224–30.
    21. Sattler EL, et al. Food insecurity and medication adherence in low-income older Medicare beneficiaries with type 2 diabetes. J Nutr Gerontol Geriatr 2014;33(4):401–17.
    22. Anderson SA. Core indicators of nutritional state for difficult-to-sample populations. J Nutr 1990;120 Suppl 11:1559–600.
    23. Bickel G, et al. Guide to measuring household food security (revised 2000). Alexandria, VA: U.S. Department of Agriculture, Food and Nutrition Service; 2000 Mar. Measuring food security in the United States: reports of the Federal Interagency Food Security Measurement Project;
    24. Piette JD, et al. A conceptually based approach to understanding chronically ill patients' responses to medication cost pressures. Soc Sci Med 2006;62(4):846–57.
    25. Briesacher BA, et al. Patients at-risk for cost-related medication nonadherence: a review of the literature. J Gen Intern Med 2007;22(6):864–71.
    26. Herman D, et al. Food insecurity and cost-related medication underuse among nonelderly adults in a nationally representative sample. Am J Public Health 2015;105(10):e48–e59.
    27. Berkowitz SA, et al. Treat or eat: food insecurity, cost-related medication underuse, and unmet needs. Am J Med 2014;127(4):303–10; 310.e1-e3.
    28. Ziliak JP, Gundersen C. The state of senior hunger in America: 2014 report. Chicago: Feeding America 2016.
    29. Sharkey JR. Risk and presence of food insufficiency are associated with low nutrient intakes and multimorbidity among homebound older women who receive home-delivered meals. J Nutr 2003;133(11):3485–91.
    30. Zhang JX, et al. The differential rates in cost-related non-adherence to medical care by gender in the US adult population. J Med Econ 2017;20(7):752–9.
    31. Zivin K, et al. Factors influencing cost-related nonadherence to medication in older adults: a conceptually based approach. Value Health 2010;13(4):338–45.
    32. Bakk L. Medicare Part D coverage gap: race, gender, and cost-related medication nonadherence. Soc Work Public Health 2015;30(6):473–85.
    33. Kaiser Family Foundation. An overview of the Medicare Part D prescription drug benefit. Washington, DC; 2019 Nov 13.
    34. Flores HL, Amiri A. Addressing food insecurity in vulnerable populations. Am J Nurs 2019;119(1):38–45.
    35. Gundersen C, Ziliak JP. Food insecurity and health outcomes. Health Aff (Millwood) 2015;34(11):1830–9.
    36. Hebert LE, et al. Alzheimer disease in the United States (2010–2050) estimated using the 2010 census. Neurology 2013;80(19):1778–83.

    cost-related medication nonadherence; food insecurity; food supply; older adults; patient compliance

    Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.