The Relationship Between Household Food Insufficiency and Development of Type 2 Diabetes over 10 Years in a Sample of U.S. Black and White Adults

Abstract Objectives Household food insufficiency (HFI) is associated with type 2 diabetes (T2D) but longitudinal studies are scarce. Our objectives were to test the relationship of 1) HFI during young-to-middle adulthood with risk of T2D 10 years (y) later; and 2) transient (1 time) or persistent (2 times) HFI with odds of T2D. Methods We used data from 3672 black and white adults (32–49y) from Year 15, 20, and 25 follow-up exams of the Coronary Artery Risk Development in Young Adults (CARDIA) Study (2000–2011). Participants with missing HFI or T2D data, prevalent T2D at Year 15 (and Year 20 for objective 2), or reported pregnancy were excluded (n = 720), for a total of n = 2952. At each exam, a screener question captured HFI over the past 12 months. Participants reporting that the food they had to eat was not always what they wanted, sometimes not enough, or often not enough were categorized with HFI. Participants with enough of the kinds of food they wanted were categorized with household food sufficiency. T2D was defined as fasting serum glucose ≥126 mg/dL or hemoglobin A1c ≥6.5%. Multivariable-adjusted repeated-measures and logistic regression models tested the relationships between HFI at Year 15 with risk of T2D at Year 25, and between transient or persistent HFI from Years 15 to 20 with odds of T2D at Year 25, respectively. Results T2D prevalence was 4.7% at Year 20 and 9.7% at Year 25. At Year 15, 13.2% reported HFI, and 14.6% and 5.2% were categorized with transient or persistent HFI, respectively. Adjusted for age, sex, and race, HFI at Year 15 was associated with 36% higher risk of T2D by Year 25 (Risk Ratio (95% Confidence Interval (RR (95% CI)): 1.36 (1.02, 1.80); P < 0.05), though the relationship was attenuated after adjustment for education (1.28 (0.96, 1.71)). Compared to no HFI at Years 15 and 20, participants with persistent, but not transient, HFI had higher odds of T2D at Year 25 in fully-adjusted models (Odds Ratio (OR (95% CI)): persistent HFI: 2.06 (1.09, 3.67), P < 0.05; transient HFI: 1.26 (0.76, 2.01)). Conclusions Transient or persistent exposure to HFI in young-to-middle adulthood may be a risk factor for developing T2D among black and white adults. Preventing and reducing HFI may be a strategy to reduce socioeconomic and racial disparities in T2D prevalence in the U.S. Funding Sources SDSU University Grants Program; NIH NHLBI K01.