Food Insecurity Is Associated With Poorer Glycemic Control in Patients Receiving Free Versus Fee-Based Care

The institutional review board of Ohio University in Athens approved this study.

Adult patients with diabetes (n = 166) were recruited from University-based free and fee-for-service clinics in southeastern Ohio. These clinics provide service for individuals living in Athens, Hocking, Meigs, Morgan, Perry, Vinton, and Washington counties. Table 1 describes the counties from which the participants were drawn.

Participants were recruited and interviewed from July to December 2010. Each prospective participant received a formal briefing regarding the purposes, benefits, and risks of the study. Patients with type 1 or type 2 diabetes were eligible for inclusion; those with other conditions (e.g., gestational diabetes, prediabetes, or other endocrine conditions) were excluded. After the briefing, individuals voluntarily chose to participate in the study and signed the associated consent forms.

Patients were administered an 86-item survey. The survey included demographic questions and the 18-item USDA Food Security Survey Module (16). The Food Security Survey Module was used to determine household food security (HFS) and household adult food security (HAFS) status (16,17). Participants’ characteristics, including age and diabetes status, as well as medical record information, were collected independent of the survey administration. Medical information included height, weight, BMI, A1C, blood lipids (total, HDL, and LDL cholesterol), systolic and diastolic blood pressure, and type 1 or type 2 diabetes diagnosis. If participants’ charts were missing needed medical information, participants were not asked to undergo additional tests for the purpose of the study.

All surveys were analyzed using SPSS version 18.0 (IBM Corp., Chicago, Ill.). Mann-Whitney U tests were used to assess for differences between clinic groups for HFS and HAFS. t Tests were used to determine differences between clinic groups with regard to BMI; A1C; total, LDL, and HDL cholesterol; and systolic and diastolic blood pressure. Pearson r correlations were used to examine the relationship of HFS status and HAFS status to A1C. Furthermore, Mann-Whitney U tests were also used to determine the distribution of HFS and HAFS with regard to type of diabetes. Frequency functions distinguished the number of participating patients with type 1 versus type 2 diabetes, as well as the primary treatment approach to managing diabetes.

Participants (n = 166) were 53 ± 16 years of age and almost entirely Caucasian (91.9%). The majority of participants (65.8%) had type 2 diabetes and some education at the collegiate level or higher (52.8%). Demographic data are summarized in Table 2. Table 3 summarizes participant food security status. Overall, 43.5% of free clinic patients and 13.8% of fee-for-service patients lived in households characterized by low or very low food security.

Table 4 summarizes differences between free clinic and fee-for-service clinic participants. A1C was the only statistically significant discriminating variable, with free clinic users having worse glycemic control during the past 2–3 months compared to fee-for-service clinic users (A1C 8.71 vs. 7.83%, P =0.005). Regardless of the type of clinic used, A1C values increased as HFS (r = 0.293, P <0.001) and HAFS (r = 0.288, P = 0.001) worsened.

Results indicate that individuals using free clinics are less food secure and have worse glycemic control than their counterparts who use fee-for-service clinics. In addition, regardless of clinic type, as food insecurity worsens, blood glucose control also worsens.

Other studies have provided empirical evidence of a higher rate of household food insecurity among populations with diabetes versus those without diabetes (18). This study indicates that those with diabetes who seek free care are prone to food insecurity. In this rural, Appalachian Ohio sample of clients using fee-for-service clinics, HFS status (13.8%) was lower than both the national (14.5%) and Ohio (16.4%) estimates. However, household food insecurity among those using free clinics (43.5%) was about three times higher than their fee-for-service counterparts, as well as three times higher than national and state estimates. HAFS levels, a proxy for individual or “personal” food security status, were similar. The disparity in food security may also reflect income level, given that U.S. households with income at or below the poverty level had food insecurity levels of 40.2% in 2010, the time frame for this study (8).

A review of other studies underscored that food insecurity may precipitate or be associated with poorer glycemic control (18). These results support the association between food insecurity and poor blood glucose control over the past 2–3 months, also demonstrating how those seeking free care have poorer control than their fee-for-service counterparts. According to Tuerk et al. (19), programs encouraging improved patient self-management practices should be developed because patients’ relative self-management skills also account for variance in glycemic control. Yet, in the absence of adequate access to food, appropriate blood glucose control may prove to be difficult.

Limitations of this study include its small convenience sample drawn from only one rural region of the United States, which limits its generalizability. Participants at the free clinic were only accessible once per month, limiting time to recruit a comparison group. It is possible that the participants do not accurately represent individuals with diabetes in Appalachian Ohio or other rural regions of the United States. Furthermore, this study recruited participants only from a diabetes-specific free clinic, but other patients with diabetes possibly attended the primary care free clinic in the study region. We are unable to infer whether there were differences between these two populations. Finally, people had to qualify for the free clinic, and it is possible that they may have over-reported their food insecurity because of concern about possibly losing access to care.

Successful management of type 1 or type 2 diabetes is complicated by food insecurity. Thus, screening for food insecurity should be completed at all clinic visits (20). In addition, diabetes educators should consider that patients may not have access to adequate food for successful diabetes management.

This study found that long-term blood glucose control worsens in accord with the severity of food insecurity and thus should serve as a catalyst for improving diabetes education in Appalachian Ohio and similar rural areas. As rural communities grapple with system-level barriers to detecting and managing diabetes (e.g., high rates of poverty; limited access to insurance, specialty medical care, and emergency services; and minimal exposure to diabetes and nutrition education [21]), medical practitioners should articulate self-management practices to patients during potentially infrequent visits, including strategies for dealing with times when patients may lack access to food. Screening for food insecurity and referring clients to community resources may also be prudent.

Many individuals must make sacrificial decisions regarding medicine, food, or medical attention, placing them in a cycle of food insecurity and chronic disease. This is particularly acute among Medicare recipients—one of the fastest-growing medical populations in the United States. This study may increase awareness among health care providers in free and low-cost clinics about the importance of looking for innovative solutions to better help these high-risk patients. This is especially important given recent research finding that food insecurity coupled with diabetes increases health care expenditures (22). Future empirical research in both rural and metropolitan areas to analyze relationships between diabetes-related comorbidities and food insecurity would positively contribute to the dearth of literature available on this topic. In addition, interventions aimed at improving both diabetes management and food security are also warranted.

Funding for this study was provided by an Ohio University Diabetes Research Initiative Research Fellowship and an Ohio University Provost Undergraduate Research Fund Award, Ohio University, Athens, Ohio.

No potential conflicts of interest relevant to this article were reported.

D.H.H. and J.H.S. supervised and K.A.B. contributed to the study design and data collection and the writing and revision of the manuscript. D.H.H. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.