Eating vegetables first may lower risk of diabetic retinopathy and nephropathy

Methodology

This cross-sectional observational study recruited diabetic patients attending outpatient services at the Ise Red Cross Hospital in Ise City, Japan during two time periods: June-August 2023 and June-August 2024. Eligibility criteria included patients with diabetes (type 1 or type 2) who were able to complete questionnaires independently. Exclusion criteria were alcoholism, severe mental illness, and inability to cooperate with the survey. Of 929 patients screened, 832 were included in the final analysis after excluding 97 due to missing eating order questionnaire or complication data.

Participants completed a validated self-administered questionnaire about their usual eating order, choosing one of four patterns: eating vegetables first, eating protein first, eating carbohydrates first, or no particular eating order. Diabetic complications were diagnosed according to Japan Diabetes Society clinical guidelines. Diabetic retinopathy was diagnosed by an ophthalmologist. Diabetic neuropathy was assessed through Achilles tendon reflex, vibratory sensation at the medial ankle, subjective symptoms, or nerve conduction test abnormalities. Diabetic nephropathy was defined as urinary albumin/creatinine ratio ≥30 mg/g or eGFR <30 mL/min/1.73 m². Cardiovascular disease included ischemic heart disease, chronic heart failure, and cerebrovascular disease.

Logistic regression analysis calculated odds ratios for each complication based on eating order, using the "no order" group as reference. Models were adjusted for age, sex, BMI, diabetes classification, diabetes duration, HbA1c, exercise habits, smoking, alcohol consumption, loneliness (measured with a validated Japanese Loneliness Scale), hypertension, dyslipidemia, and diabetes medications. Statistical significance was set at p<0.05 using two-sided tests. The study was approved by the Ise Red Cross Hospital Ethical Review Board and conducted according to Helsinki Declaration principles. Written informed consent was obtained from all participants.

Interventions

The intervention of interest was the self-reported dietary habit of eating vegetables first at meals. This behavioral pattern involves consuming vegetable dishes before eating protein, carbohydrates, or other meal components. The practice is commonly recommended in Japan as a strategy to improve postprandial glucose control and is culturally facilitated by traditional Japanese meal structure ("Ichiju-sansai" - one soup and three dishes) where vegetables are often served as separate side dishes.

Participants were classified into four eating order groups based on their self-reported usual pattern: (1) vegetables first - eating vegetable dishes before other foods; (2) protein first - eating protein foods including milk and yogurt before other foods; (3) carbohydrates first - eating rice, bread, noodles, or other carbohydrate-rich foods before other foods; or (4) no particular order - no consistent eating pattern. The primary comparison was between the vegetables first group and the no order group. This was not a controlled dietary intervention but rather an observational comparison of existing dietary behaviors.

The proposed mechanism by which eating vegetables first may reduce diabetic complications involves multiple physiological effects. Dietary fiber in vegetables consumed early in the meal delays gastric emptying and slows carbohydrate absorption, thereby reducing postprandial glucose spikes. This eating pattern has been shown to enhance secretion of incretin hormones (GLP-1 and GIP) and reduce postprandial insulin levels. By improving postprandial glycemic control and reducing glycemic variability throughout the day, the vegetables first pattern may reduce oxidative stress and inflammatory cytokine production, potentially protecting against microvascular damage.

Key Findings

Among 832 diabetic patients, 52.6% reported habitually eating vegetables first, 24.5% had no particular eating order, 15.8% ate carbohydrates first, and 7.1% ate protein first. After comprehensive adjustment for demographic, clinical, and lifestyle factors, the habit of eating vegetables first was significantly associated with lower prevalence of diabetic retinopathy (adjusted OR 0.64, 95% CI 0.41-0.98, p=0.043) and diabetic nephropathy (adjusted OR 0.47, 95% CI 0.30-0.71, p<0.001) compared to having no particular eating order. The association was particularly strong for nephropathy, suggesting a 53% reduction in odds. No significant associations were found between eating vegetables first and diabetic neuropathy (adjusted OR 0.80, p=0.289) or cardiovascular disease (adjusted OR 1.28, p=0.351). Neither the protein first nor carbohydrates first patterns showed significant associations with any complications.

Comparison with other Studies

These findings align with and extend previous research on meal sequencing and metabolic outcomes in diabetes. Several prior studies have demonstrated that eating vegetables before carbohydrates reduces postprandial glucose excursions and glycemic variability in diabetic patients. A 2013 study by Imai et al. showed that this eating pattern improved postprandial glucose levels, while a 2022 retrospective cohort study found sustained glycemic benefits over 5 years. Research by Shukla et al. (2015) demonstrated that food order significantly impacts postprandial glucose and insulin levels in type 2 diabetes. A 2023 randomized controlled trial confirmed that eating vegetables first reduces postprandial glucose and insulin regardless of eating speed.

However, this appears to be the first study to specifically examine associations between eating order and diabetic complications. The differential findings—significant associations for retinopathy and nephropathy but not neuropathy or cardiovascular disease—are consistent with literature suggesting that postprandial hyperglycemia and glycemic variability are particularly strongly associated with microvascular complications. Previous studies have shown inconsistent associations between glycemic variability and macrovascular disease or neuropathy, which may explain why eating vegetables first was not associated with these outcomes. The cross-sectional design limits comparison with longitudinal intervention studies, and the Japanese population may have unique dietary patterns that facilitate this eating habit compared to Western populations.

Journal Reference

Imataka K, Ida S, Tanaka T, Azuma K, Murata K. The Habit of Eating Vegetables First Is Linked with a Lower Risk of Diabetic Retinopathy and Nephropathy: Cross-Sectional Study in Japanese Diabetic Patients. Diabetes Metab Syndr Obes. 2026;19:555812. doi:10.2147/DMSO.S555812