Food insecurity and medication scrimping among U.S. adults with diabetes

Objective: Many American households experience food insecurity, lacking consistent access to sufficient, nutritionally suitable food.  According to the USDA (2012), approximately 15 percent (14.9 %) of American households were food insecure at one point during 2011.  Further, one in every 17 food insecure households (5.7%) had very low food security, meaning that one or more household members altered their normal meal schedule or was forced to reduce food intake at various times during the year due to the lack of financial resources.  Our research examined the prevalence of food insecurity among adults with self-reported diabetes, and whether food insecurity was associated with cutting back on prescribed medications.

Research Design and Methods:  We conducted a cross-sectional analysis of the 2011 National Health Interview Survey, a national representative household survey.  Adults were considered to have diabetes if they reported current use of insulin or “diabetic pills” (n=3,242).  Food insecurity was determined with a 10-item scale; we categorized respondents as food secure, marginally food secure, or food insecure.  “Scrimping” was defined as a “yes” response to one of four questions pertaining to reduced, delayed or avoided medication use.  Socio-demographic variables were included in adjusted analysis.  All analyses reflected the complex survey design.

Results: Approximately one in every six adults with diabetes reported food insecurity (17.0%) and an additional 8.9% were marginally food secure.  Overall, 18.9% of persons with diabetes reported one or more types of medication scrimping; 11.7% among food secure individuals, 27.7% among marginally food secure, and 45.6% among food insecure respondents.  In adjusted analysis, marginal food security (OR 2.66) and food insecurity (OR 5.89) remained strongly associated with medication scrimping.  Other factors positively associated with scrimping after controlling for food insecurity included age less than 65, mid-range income (above 100% of the poverty level but below 300%), being uninsured, living in the South, a BMI of 30 or more, and a high number of office visits.  Black or Hispanic race-ethnicity had a modest negative relationship to scrimping after adjusting for food security.  Similarly, education less than high school was associated with reduced odds of scrimping after adjustment.   

Conclusions: A quarter of adults with diabetes experience some level of food insecurity; a substantial subset of these individuals fail to obtain or take prescribed medications.  Practitioners may miss both problems unless targeted questions are included in each clinical encounter.  Clinicians also need to establish referral links to community food resources.