Background: Food insecurity (FI) among individuals with diabetes has been associated with both hypoglycemia and worse glycemic control, but studies of elderly insured populations are lacking.

Methods: As part of its Medicare Annual Wellness Visit, Kaiser Permanente Colorado (KPCO) initiated a member survey in January 2012 that included a question about FI (“Do you always have enough money to buy the food you need?”). We combined this survey with electronic health record data and other data to conduct a nested case-control study of FI among KPCO members ≥ 65 years with diabetes. Of 32,666 members with diabetes, 10,052 (31%) completed a survey. Of those, the 742 (7.4%) who had FI were matched 1:3 based on date of the survey (± 7 days) to 2,226 individuals without FI.

Results: Individuals with FI were slightly younger (mean age ± SD of 73.2 ± 6.1 vs. 73.9 ± 6.8), and were more likely to be female (60.6% vs. 46.8%), non-white (36.8% vs. 21.8%), living without a partner (49.6% vs. 33.0%), have less than high school education (21.2% vs. 9.0%), and have Medicaid (16.4% vs. 3.2%). Individuals with FI reported higher frequency of fair or poor general health (29.2% vs. 16.7%) and quality of life (23.6% vs. 10.5%), and were more likely to eat ≤ 2 meals a day (19.0% vs. 8.9%). Rates of hypertension and chronic kidney disease were similar, but individuals with FI were more likely to have depression on PHQ2 screening (24.8% vs. 9.1%). Individuals with FI were more likely to be taking any diabetes medication (70.5% vs. 63.6%) or insulin (25.4% vs. 19.4%), have an emergency department (ED) visit (19.0% vs. 13.9%) or hospitalization in the prior year (8.6% vs. 7.0%), and have an A1c ≥ 9% (11.2% vs. 5.8%). Mean A1c for those with FI was 7.4 ± 1.4 and those without FI was 7.1 ± 1.1. ED visits or hospitalizations for hypoglycemia or hyperglycemia were infrequent (≤10) in both groups.

Conclusions: In this insured elderly population with diabetes, food insecurity was associated with higher health care utilization and A1c levels. Strategies to address FI in individuals with diabetes deserve further evaluation.


E.B. Schroeder: None. C. Zeng: None. A. Sterrett: None. T.K. Kimpo: None. A.R. Paolino: None. J. Steiner: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at