Medical Financial Hardship and Food Security among Cancer Survivors in the United States

The number of cancer survivors in the United States (U.S.) continues to grow due to advances in detection and treatment, with the majority of survivors now living five years or longer after diagnosis (1). As the cancer-survivor population grows, so does concern about their long-term physical, psychosocial, and financial well-being (2, 3). Cancer survivors experience worse health and quality of life compared with individuals without a cancer history (4, 5), and many face financial toxicity (i.e., the objective financial burden and subjective financial distress of patients with cancer due to the costs of treatment; ref. 6). Because of cancer's financial toxicity, some survivors cut back household budgets, reduce household assets, or make other financial sacrifices (7). This financial fragility may increase the risk of food insecurity, which in turn is associated with negative health outcomes and foregone care among cancer survivors (8).

Food security, defined as “access by all people at all times to enough food for an active, healthy life,” (9) is recognized as an important social determinant of health (10). Food insecurity, a condition in which “households lack access to adequate food because of limited money or other resources,” (10) is associated with poor quality diet and low nutritional intake (11). Among U.S. households in 2020, 89.5% reported being food secure, whereas 10.5% reported being food insecure (9). Of those households that reported food insecurity, 6.6% reported low food security, and 3.9% reported very low household food security (9). Among general populations of adults, food insecurity is a risk factor for chronic disease (12, 13), depression and anxiety (14–16), and overweight or obese weight status (17), although these associations sometimes differ across age and gender (10, 18). Food insecurity is also a predictor of cancer (19). This is especially concerning for cancer survivors, many of whom are at risk of cancer recurrence and new primary malignancies (20, 21).

Although there is less research regarding food security among cancer survivors, the available research shows that cancer survivors who are younger, have lower incomes, and are in poorer health are at heightened risk of food insecurity (22). High levels of both nonmedical financial worry (i.e., very worried about retirement, standard of living, bills, and housing costs) and food insecurity have been found among younger cancer survivors, compared with individuals without a cancer history (23). Regardless of age, cancer survivors with lower incomes and more comorbidities reported higher levels of nonmedical financial worry and food insecurity, compared with those without a cancer history (23). On the basis of this limited body of evidence, nonmedical financial worry, lower income, and poorer health are associated with food insecurity among cancer survivors (22, 23).

Less is known about the relationship between medical financial hardship and food security among cancer survivors. Cancer survivors often report material (e.g., problems paying medical bills), psychological (e.g., worrying about paying medical bills), and behavioral (delaying/forgoing care due to cost) medical financial hardships (24–26). Medical financial hardship is more likely among cancer survivors compared with individuals without a cancer history (25). Survivors face medical financial hardship during cancer treatment due to the costs of cancer care (6) and after active treatment due to depleted assets, diminished ability to work, and the costs of follow-up and other health care (27–29). Among cancer survivors, medical financial hardship is associated with more emergency room visits, lower preventive services use, worse self-rated health, and nonmedical financial sacrifices (26, 30).

It is likely that cancer survivors experiencing medical financial hardship are at greater risk of food insecurity than those who are not experiencing such hardship; however, researchers have not examined this relationship or what domains of medical financial hardship (i.e., material, psychological, behavioral) are most strongly associated with food insecurity among cancer survivors. To fill this gap, this study explores whether medical financial hardship (material, psychological, and behavioral) is related to household food security among a population-based sample of cancer survivors.

We used cross-sectional data from the 2020 National Health Interview Survey (NHIS). The NHIS is an annual cross-sectional survey that gathers data on illness, disability, chronic impairments, health insurance, health care access, and health services use from a nationally representative sample of the civilian, non-institutionalized U.S. population. One adult ages ≥18 years is selected at random from randomly contacted households to provide demographic and health-related information (31). The study was considered exempt by the University of Arkansas for Medical Sciences Institutional Review Board (#274508).

In 2020, 31,568 adults completed NHIS interviews. Among them, 4,130 adults ages ≥18 years reported a cancer diagnosis from a health professional.

An ordinal household food security measure was included in the 2020 NHIS dataset. It was determined on the basis of responses to 10 questions measuring household food security over the past 30 days (e.g., being worried food would not last until there is money to buy more, not being able to afford to eat balanced meals, cutting or skipping meals; ref. 32). Household food security was categorized as one of four categories as defined by the U.S. Department of Agriculture Economic Research Service (33). These included: (i) high food security: No problems or anxiety about consistently accessing adequate food; (ii) marginal food security: Those who had problems or anxiety about accessing adequate food, but the quality, variety, and quantity of their food intake were not substantially reduced; (iii) low food security: Those households who have reduced the quality, variety, and desirability of their diets, but the quantity of food intake and normal eating patterns were not substantially disrupted; or (iv) very low food security: Households where eating patterns of one or more household members were disrupted and food intake reduced because the household lacked money and other resources for food.

Medical financial hardship was assessed by three domains: Material (problems paying or unable to pay medical bills); psychological (worry about paying medical bills); and behavioral (delayed or foregone medical care due to cost). Material medical financial hardship was assessed with the question, “In the past 12 months, did you or anyone have problems paying or were unable to pay any medical bills?” with a dichotomized response of yes/no. Psychological medical financial hardship was evaluated with the question, “If you get sick or have an accident, how worried are you that you will be able to pay your medical bills?” Responses to this question were categorized as not at all worried, somewhat worried, and very worried. Finally, behavioral medical financial hardship was measured using two questions: “During the past 12 months, have you delayed getting medical care because of the cost?” and “During the past 12 months, was there any time when you needed medical care but did not get it because of the cost?”, each with a dichotomized response of yes/no.

Cancer-related characteristics included cancer type and time since diagnosis. The NHIS asked participants who were ever told by a health professional they had cancer to report their cancer type. Up to three different kinds of cancers could be mentioned by the sample adult. Participants who reported more than one cancer were put in their own analysis category to avoid issues such as double-counting. The final categories were female breast, female reproductive (cervical, ovarian, and uterine), male reproductive (prostate), digestive system (esophageal, gallbladder, liver, pancreatic, stomach, colorectal), melanoma skin cancer, non-melanoma skin cancer and skin cancer of unknown type, other cancers (bladder, blood, bone, brain, leukemia, lung, lymphoma, thyroid, head and neck, and other), and multiple diagnoses.

The NHIS survey asks participants to provide the age(s) at which they were diagnosed with cancer. Time since diagnosis was measured as age at the time of survey minus self-reported age at the time of the diagnosis. Time since diagnosis was categorized as recently diagnosed (most recent diagnosis <2 years ago) and previously diagnosed (most recent diagnosis ≥2 years ago). Participants who reported more than one cancer diagnosis were categorized on the basis of their most recent diagnosis.

Individual-demographic and health covariates were chosen on the basis of prior research on factors associated with food security among the general population and cancer survivors specifically (22, 23, 34–36). Demographic and health characteristics included age at the time of the survey (<65, ≥65), sex (male, female), race/ethnicity (White, Black, Hispanic, Asian, American Indian/Alaska Native, other/multiple racial groups), marital status (married/partnered, single/divorced/widowed), education level (less than high school/GED, HS/GED, some college/associate degree/bachelor's degree, above bachelor's degree), family income level as a percentage of the federal poverty level (≥400% FPL, ≥300%–399% FPL, ≥200%–299% FPL, <100%–199% FPL), Supplemental Nutrition Assistance Program (SNAP) recipient in the past 12 months (yes, no), health insurance (no, yes), employment (not employed in the past week or past year, employed in the past week or past year), nativity (no, yes), U.S. region of residence (South, Midwest, West, Northeast), area of residence (large central metropolitan, large fringe metropolitan, medium and small metropolitan, nonmetropolitan), self-reported health (excellent, very good, good, fair/poor), body mass index (underweight, normal, overweight, obese), and sum of comorbid conditions.

Weighted percentages and 95% confidence intervals (CI) for categorical variables and weighted means and standard errors (SE) for continuous variables were computed to describe the study population.

To examine unadjusted associations between medical financial hardship and household food security among cancer survivors, we first used the Rao–Scott χ² test of independence that accounts for the NHIS complex survey design (37).

Before estimating the regression model, multicollinearity was assessed between all independent variables based on the variance inflation factor (VIF). Because the outcome variable was ordinal, ordinal logistic regression was used to examine adjusted associations between medical financial hardship and food security. To determine whether the parallel assumption for proportional-odds was not violated, the Brant Test was conducted (38, 39). The Brant test indicated that the assumption of parallel regression was not violated [χ²_{(df = 42)} = 57.29; P = 0.058]. Therefore, we proceeded with ordinal logistic regression. All four medical financial hardship variables were included in the same ordinal logistic regression model, which controlled for all individual characteristic variables.

Descriptive and regression analyses were all weighted using sampling weights variables so that estimates are representative of the U.S. study population and to obtain correct SEs. Estimates were obtained from complete case analyses, because missing data were <5% (40). STATA 17 was used with the svy prefix command to fit statistical models for NHIS complex survey data (41). Statistical significance was a priori determined at an alpha level α = 0.05.

Finally, we also investigated whether the association between medical financial hardship and food security would differ between cancer survivors (n = 4,130) and those without a cancer history (n = 27,404). To that end, we conducted a moderation analysis on the whole sample of 31,568 adults (i.e., adults with and without a cancer history, excluding those who did not provide a response to this question and pregnant women). This analysis explored cancer history's effect modification by including an interaction product term between the dichotomous cancer history variable and each of the medical financial hardship variables controlling for individual characteristics. A post hoc adjusted Wald test was conducted to test for moderation.

The deidentified data underlying the results presented in this study may be made available upon request from the corresponding author, E. Hallgren, at [email protected]. The data are not publicly available in accordance with funding requirements and participant privacy.

Demographic, cancer-related, and health-related characteristics of the sample are summarized in Table 1. About one in 10 (9.6%) adults surveyed by NHIS reported a history of cancer. The majority of cancer survivors (88.5%) reported high household food security, 4.8% reported marginal household food security, 3.6% reported low household food security, and 3.1% reported very low household food security (Fig. 1). Figure 2 depicts medical financial hardship among cancer survivors. More than one in 10 cancer survivors (13.2%) had problems paying or were unable to pay medical bills in the past 12 months. More than a quarter (26.2%) was somewhat worried about their ability to pay medical bills, and 12.4% were very worried. Under 10% of respondents reported delaying medical care or foregoing medical care due to cost (6.5% and 5.5%, respectively).

The demographic, cancer-related, and health-related characteristics of the study population across the four categories of household food security are depicted in Table 2.

All four medical financial hardship variables demonstrated significant bivariate associations with household food security (Table 2). The VIF of the independent variables ranged between 1.02 and 2.04, with a mean VIF of 1.36, indicating no multicollinearity among the independent variables. After adjusting for the influence of the other covariates, we found the odds of being in a lower food security category (i.e., high vs. marginal, marginal vs. low, or low vs. very low) increased by 1.73 (95% CI, 1.06–2.82) times for cancer survivors who had problems paying or were unable to pay their medical bills compared with those who did not. The odds of being in a lower food security category (i.e., high vs. marginal, marginal vs. low, or low vs. very low) were 2.88 times higher (95% CI, 1.64–5.07) for cancer survivors who were very worried about paying their medical bills compared with those who were not at all worried. The odds of being in a lower food security category (i.e., high vs. marginal, marginal vs. low, or low vs. very low) were 2.56 times higher (95% CI, 1.29–5.09) for cancer survivors who delayed medical care due to cost compared with those who did not. The association between forgone medical care due to cost and food security status was not significant in the adjusted model (Table 3).

In addition, as a Supplementary Analysis (Supplementary Tables S1 and S2), we explored whether having a cancer history had a moderating effect on the relationship between the three domains of medical financial hardship (material, psychological, and behavioral) and food security. In Supplementary Table S1, we provide descriptive statistics for the population without a cancer history. The estimates of interaction terms between cancer history and two of the domains of medical financial hardship, namely the material and behavioral domains, were non-significant (Supplementary Table S2). For psychological medical financial hardship, the estimate of the “somewhat worried” category was significant (P = 0.034); however, an adjusted Wald test further showed that association between the predictor psychological financial hardship and food security was non-significant [F_{(2, 569)} = 2.46; P = 0.09]. In other words, the association between material, psychological, and behavioral medical financial hardships and food security status did not significantly differ between adults with and without a cancer history in our study population. Nevertheless, within the cancer survivor population, our findings indicated a significant and independent effect of each of the four financial hardship variables on food security (Associations between medical financial hardship and food security among cancer survivors).

The purpose of this study was to examine the relationship between medical financial hardship and household food security among cancer survivors and explore what domains of medical financial hardship (i.e., material, psychological, and behavioral) are related to household food security among this population.

The majority of cancer survivors in the sample reported high household food security (88.5%), with 4.8% reporting marginal, 3.6% reporting low, and 3.1% reporting very low household food security. These rates are somewhat lower than the rates of low household food security (6.6%) and very low household food security (3.9%) reported by U.S. households overall in 2020 (9). It is possible that a disproportionate number of patients with cancer with low or very low food security passed away prematurely or were unable to participate in the NHIS survey. Although food insecurity was reported by a relatively small proportion of cancer survivors, it has important clinical and non-clinical implications for those who are affected. Food insecurity likely resulted in poorer diet quality for these survivors (11). Food insecurity is also a predictor of cancer (19), a particular concern for cancer survivors who may be at risk of cancer recurrence or new primary malignancies (20, 21). Food insecurity likely adds to distress and reduced quality of life in survivorship as well.

About 13% had problems paying or were unable to pay medical bills in the past 12 months, and more than one third of participants were somewhat or very worried about paying medical bills. Less than 10% of respondents reported delaying or foregoing medical care due to cost. These rates of material, psychological, and behavioral medical financial hardships among cancer survivors are lower than those reported by Zheng and colleagues (25) that used 2013 to 2016 NHIS data. Zheng and colleagues (25) included more measures for each domain of medical financial hardship (material, psychological, and behavioral) than the current study, which likely accounts for the higher rates of medical financial hardship among cancer survivors they reported. Our results are similar to the rates of material and psychological medical financial hardships among cancer survivors reported by Han and colleagues (26) but substantially lower than the rate of behavioral medical financial hardship they reported. Han and colleagues (26) used Medical Expenditures Panel Survey data and included more and different measures of medical financial hardship compared with the current study. These differences in results suggest the number and types of measures used to assess medical financial hardship will affect the overall estimates of material, psychological, and behavioral medical financial hardships among cancer survivors.

Our results showed that material, psychological, and behavioral medical financial hardships were each associated with an increased risk of lower household food security among cancer survivors. These results align with a growing body of evidence demonstrating that medical financial hardship is associated with a host of negative outcomes for cancer survivors, including more emergency room visits, lower preventive services use, worse self-rated health, and nonmedical financial sacrifices (26, 30). The strongest association was observed between the psychological domain of medical financial hardship (i.e., very worried about paying medical bills) and food security status. It may be that those who reported being very worried about paying their medical bills already reduced their food intake or struggled to afford food. Previous research suggests that food-insecure households routinely make trade-offs or choose between paying for food and other essential costs such as housing and medical care (42).

Our results align with the literature showing survivors with lower income and poorer health are at greater risk of food insecurity (22, 23). Taken together, our findings add to a body of work identifying a sub-set of particularly vulnerable cancer survivors who face a combination of interrelated challenges, including medical financial hardship, low income, poor health, and food insecurity, all of which may greatly decrease quality of life among survivors.

Previous research by Zheng and colleagues (23) showed that nonmedical financial worry was associated with food insecurity among younger cancer survivors. Our findings demonstrate that medical financial worry, along with other forms of medical financial hardship, is associated with risk of food insecurity among cancer survivors, with odds of being in a lower food security category decreasing with age. Together, these findings suggest both nonmedical and medical financial worry and food insecurity are most strongly related among younger cancer survivors.

Medical financial hardship is more likely among cancer survivors compared with individuals without a cancer history (25). Given the risk of food insecurity among cancer survivors with medical financial hardship, it is crucial for cancer specialists and other healthcare providers to screen for financial hardship, food insecurity, and other social determinants of health during patient encounters. Furthermore, providers should be made aware that younger cancer survivors are often at particular risk of both financial hardship and food insecurity and conduct appropriate social needs screenings for this group.

To avoid the potential for cancer survivors to feel stigmatized by food insecurity screenings, providers should consider framing these screenings in the importance of healthy eating and adequate nutrition during and after cancer treatment (43). Furthermore, providers should engage oncology social workers, cancer navigators, and financial counselors to assist cancer survivors and their families in obtaining food assistance through federal and local programs. As prior work has suggested, cancer survivors may cancel appointments or ration medication due to financial concerns related to housing and food procurement (8, 43); therefore, healthcare teams may need to reach out to these survivors specifically to offer assistance.

Finally, a Supplementary Analysis found that the association between medical financial hardship and food security did not significantly differ between adults with and without a cancer history. Nonetheless, within the cancer survivor population, our findings indicated a significant and independent effect of each of the four financial hardship variables on food security. This suggests that medical financial hardship and food security are related among both cancer survivors and those without a cancer history. Nonetheless, because of the added risk of cancer-related financial toxicity, special attention should be paid to the potential for medical financial hardship and food insecurity among cancer survivors.

These results should be interpreted with some limitations in mind. First, this study used cross-sectional data to examine the association between the three domains of medical financial hardship (material, psychological, and behavioral) and household food security; as such, causality cannot be inferred. Second, NHIS response rate was 50.7%. Non-response bias might have affected the data and results if respondents and non-respondents differed in responses to the variables under study. However, this response rate is higher than most national surveys (44). Sample sizes for some racial/ethnic minority groups are small, and results for these groups should be interpreted with caution. Furthermore, determination of medical financial hardship was based on one or two questions for each domain. These questions may not have been sensitive enough to capture more nuanced levels of medical financial hardship or distress. Finally, household food security and material, psychological, and behavioral medical financial hardship were self-reported. Self-report measures may be prone to social desirability and recall biases. Despite these limitations, this study is based on a representative sample of the U.S. civilian population of cancer survivors, which facilitates generalizability of findings at the population level and the formulation of public health policies tailored to this vulnerable population.

Our findings point to the importance of additional research in the area of cancer and food security. Future qualitative research could help illuminate the trade-offs cancer survivors make in terms of their financial needs (e.g., how they decide to pay medical bills vs. buy nutritious food). Longitudinal research is needed to help establish how financial and material needs change over the course of patients' treatment and into a longer-term survivorship. Additional research on the development of interventions to mitigate medical financial hardship and food insecurity among cancer survivors is also needed, particularly among the highly vulnerable subset of cancer survivors who may face medical financial hardship, poor health, and food insecurity simultaneously.

T. Thompson reports grants from American Cancer Society during the conduct of the study. No disclosures were reported by the other authors.

E. Hallgren: Formal analysis, validation, writing–original draft, writing–review and editing. M.-R. Narcisse: Formal analysis, validation, writing–original draft, writing–review and editing. J.A. Andersen: Writing–original draft, writing–review and editing. D.E. Willis: Formal analysis, validation, writing–original draft, writing–review and editing. T. Thompson: Writing–review and editing. G. Bryant-Smith: Writing–review and editing. P.A. McElfish: Writing–review and editing.

This research was supported by University of Arkansas for Medical Sciences Translational Research Institute funding awarded through the National Center for Advancing Translational Sciences of the National Institutes of Health (NIH; UL1 TR003107). P.A. McElfish (Co-Investigator) received this grant.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.