Outcomes of Cooking Matters for Diabetes: A 6-week Randomized, Controlled Cooking and Diabetes Self-Management Education Intervention

Abstract

Background: Diabetes self-management education and support is the cornerstone of diabetes care, yet only 1 in 2 adults with diabetes attain hemoglobin A1c (HbA1c) targets. Food insecurity makes diabetes management and HbA1c control more difficult.

Objective: Our aim was to test whether a cooking intervention with food provision and diabetes self-management education and support improves HbA1c and diabetes management.

Design: This was a waitlist-controlled, randomized trial.

Participants/setting: Participants were 48 adults with type 1 or type 2 diabetes.

Intervention: Cooking Matters for Diabetes was adapted from Cooking Matters and the American Diabetes Association diabetes self-management education and support intervention into a 6-week program with weekly food provision (4 servings).

Main outcome measures: Surveys (ie, Summary of Diabetes Self-Care Activities; Medical Outcomes Study Short Form Health Survey, version 1; Diet History Questionnaire III; 10-item US Adult Food Security Survey Module; and Stanford Diabetes Self-Efficacy Scale) were administered and HbA1c was measured at baseline, post intervention, and 3-month follow-up.

Statistical analysis: Mixed-effects linear regression models controlling for sex and study wave were used.

Results: Mean (SD) age of participants was 57 (12) years; 65% identified as female, 52% identified as White, 40% identified as Black, and 19 (40%) were food insecure at baseline. Intervention participants improved Summary of Diabetes Self-Care Activities general diet score (0 to 7 scale) immediately post intervention (+1.51; P = .015) and 3 months post intervention (+1.23; P = .05), and improved Medical Outcomes Study Short Form Health Survey, version 1, mental component score (+6.7 points; P = .025) compared with controls. Healthy Eating Index 2015 total vegetable component score improved at 3 months (+0.917; P = .023) compared with controls. At baseline, food insecure participants had lower self-efficacy (5.6 vs 6.9 Stanford Diabetes Self-Efficacy Scale; P = .002) and higher HbA1c (+0.77; P = .025), and demonstrated greater improvements in both post intervention (+1.2 vs +0.4 Stanford Diabetes Self-Efficacy Scale score; P = .002, and -0.12 vs +0.39 HbA1c; P = .25) compared with food secure participants.

Conclusions: Cooking Matters for Diabetes may be an effective method of improving diet-related self-care and health-related quality of life, especially among food insecure patients, and should be tested in larger randomized controlled trials.

Trial registration: ClinicalTrials.gov NCT04152811.

Keywords: Cooking; Diabetes self-management education and support; Food insecurity; Social determinants of health; Type 2 diabetes mellitus.

Figure 1.

Cooking Matters for Diabetes intervention administration and data collection schedule. Arms 1 and 3 are intervention arms; arms 2 and 4 are waitlist-control arms. A1c = hemoglobin A1c.

Figure 2.

Mean hemoglobin A1c in Cooking Matters for Diabetes participants over time by food security and intervention/control group status. Mean and SE hemoglobin A1c for control and intervention participants with and without food insecurity at baseline. Interpretation: Food insecure intervention participants experienced the numerically largest decrease in hemoglobin A1c (interaction not significant). Food insecure participants started with higher hemoglobin A1c than food secure participants (P < .05). A1c = hemoglobin A1c.