Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes

Diabetes mellitus is characterized by hyperglycemia due to defects in insulin secretion or action; type 1 diabetes accounts for about 10% of cases. Intensive insulin therapy, often via continuous subcutaneous insulin infusion, aims to achieve near-normal glycemia to prevent complications. Carbohydrates strongly affect postprandial glycemia, and rice is a common carbohydrate source globally and among people with diabetes. Starch in rice (amylose and amylopectin) becomes more digestible after cooking due to gelatinization. Cooling cooked starch induces retrogradation, forming double helices that resist amylase digestion (resistant starch), thereby reducing digestible carbohydrate content. While this could benefit glycemic control, the effect of retrograded starch on postprandial glycemia in people with type 1 diabetes had not been reported. This study evaluates whether cooling rice affects postprandial glycemia in adults with type 1 diabetes and explores impacts on hypoglycemia risk and subjective measures (hunger, satiety, desire to eat) and organoleptic qualities.

Background literature highlights that cooling cooked starch (RS3 formation) reduces enzymatic digestibility by promoting amylose and long amylopectin chain retrogradation into amylase-resistant double helices. Prior human studies in non–type 1 diabetes populations have shown mixed effects: some report reduced postprandial glycemia after cooled/reheated rice or increased resistant starch (e.g., Sonia et al., Ananda et al.), whereas others found no significant effect. Multiple heating/cooling cycles can further increase resistant starch content. Evidence on satiety effects of resistant starch is mixed, and palatability generally appears unaffected by cooling. Critically, there were no prior studies specifically assessing cooled starch products’ effects on postprandial glycemia in people with type 1 diabetes using intensive insulin therapy, motivating the present investigation.

Design: Randomized, single-blind, crossover clinical study. Ethical approval: Bioethics Committee of Poznan University of Medical Science, Poland (198/18; 1.02.2018). Participants: 32 adults with type 1 diabetes from the Department of Internal Medicine and Diabetology, Poznan University of Medical Sciences. Inclusion: T1D, age >18, intensive insulin therapy via personal insulin pump, BMI <30 kg/m², HbA1c <9%, written consent. Exclusion: pregnancy; other diabetes types; eating disorders; allergy/intolerance to meal components; celiac disease; autonomic neuropathy (including gastroparesis); pump use <3 months. Baseline assessments: Body composition by TANITA BC-418 MA; labs included HbA1c (turbidimetric inhibition immunoassay, Cobas 6000) and lipid profile (HDL, LDL, TC, triglycerides; enzymatic assays, Cobas 6000). Insulin therapy: All used pumps with bolus calculator. Basal rate testing between 1–5 pm prior to study; adjustments made if needed. No basal changes on test days. Procedures: Each participant consumed two standardized test meals on separate randomized days at 2 pm, 10 min eating duration. Pre-conditions: No hypoglycemia within 24 h before testing; pre-meal glucose 3.9–10 mmol/L; 5-h fast (water only); avoidance of vigorous activity from the day before through testing. Blinding: Single-blind to meal condition; cooled meal reheated to match temperature. Insulin dosing: Rapid-acting analog (Lispro/Aspart) bolus 10 min before the meal, based on carb exchange factor, correction factor, insulin sensitivity, via pump bolus calculator; no active insulin from prior bolus; stable glycemia confirmed by FreeStyle Libre prior to dosing and meal. Test meals: Long-grain white rice prepared on an induction plate (Silver Crest SIKP 2000 E2). Cooking: 70 g dry rice in 280 ml water, added to boiling water, cooked 18 min. Meal: 200 g rice (46 g carbohydrate) + 100 g plain tomato sauce (4 g carbohydrate) = 50 g carbohydrate total. Cooled condition: Pre-weighed portion cooled at 4 °C for 24 h, then reheated by immersion in 250 ml hot water for 3 min. Composition analyses: Energy, macronutrients, starch, fiber, water, ash per 100 g; resistant starch (AOAC 2002.02) measured in fresh and cooled (reheated) rice. Other analyses: Protein (Dumas), fat (NMR), starch (polarimetric), ash (dry method), carbohydrates by difference, fiber (enzymatic), water (oven). Glucose monitoring: FreeStyle Libre flash system at 5-min intervals for 180 min postprandially; sensors applied ≥2 days prior to first test. Breakfast standardization: Participants consumed the same breakfast on both test days; intake analyzed (energy, macronutrients, fiber) using Dietician 2014 with Polish food database; pre-breakfast insulin recorded. Hypoglycemia protocol: If sensor glucose <3.9 mmol/L or symptomatic hypoglycemia (confirmed by capillary glucose via OptiumXido), test stopped; 15–20 g fast-acting carbohydrate given; rechecked every 15 min; time recorded. Organoleptic assessment: Questionnaire rating taste, visual appeal, smell, and consistency. Subjective appetite measures: Visual Analogue Scale (0–10) for hunger, satiety, desire to eat at baseline and 30, 60, 120, 180 min post-meal. Statistical analysis: Shapiro–Wilk for normality; paired t-test for normally distributed paired variables; Wilcoxon signed-rank for non-normal continuous variables; McNemar test for hypoglycemia incidence. Results reported as medians (IQR) or n (%).

- Nutrient composition (per 100 g cooked rice): 435.34 kJ (104 kcal), protein 2.47 g, fat 0.1 g, carbohydrates 23 g, starch 21.7 g, fiber 0.95 g, water 74.4 g, ash 0.1 g. Resistant starch: fresh 7.52 ± 0.05 g/100 g; cooled 11.96 ± 0.04 g/100 g. - Postprandial glycemia (cooled vs fresh): • Maximum glycemia: 9.9 (9.4–10.9) vs 11 (10.3–11.7) mmol/L; p=0.0056. • Maximum glycemic increase: 2.7 (1.5–3.6) vs 3.9 (2.5–4.7) mmol/L; p<0.0001. • Incremental AUC (0–180 min): 135 (34.3–283.9) vs 336 (123.9–486.9) mmol/L × 180 min; p<0.0001. • Time to peak glycemia: 35 (28–43) vs 45 (35–55) min; p=0.031. - Hypoglycemia within 180 min: 12 events (38%) after cooled rice vs 3 (9%) after fresh; p=0.0039; median time of hypoglycemia: cooled 110 (92.5–160) min vs fresh 110 (90–130) min; p=0.18. - Insulin dosing: No significant differences between conditions in pre-meal bolus [5.5 (5.0–6.1) vs 5.4 (4.9–6.0) units; p=0.92] or basal dose [15.8 (13.0–19.9) vs 15.8 (13.0–19.9) units; p=0.66]. - Pre-test breakfasts and pre-breakfast insulin were similar between days (no significant differences in energy, macronutrients, fiber, or insulin). - Organoleptic ratings (taste, visual appeal, smell, consistency) did not differ between fresh and cooled rice. - Hunger, satiety, and desire to eat ratings up to 180 min post-meal did not differ between conditions.

Cooling and reheating rice increased resistant starch content and reduced digestible carbohydrate, leading to significantly attenuated postprandial glycemic excursions in adults with type 1 diabetes using insulin pumps. This directly answers the research question, showing that a simple culinary modification can lower the maximum glycemia, the incremental rise, and the overall glycemic exposure (AUC), while also shortening time to peak. The shorter time to peak may better align with rapid-acting insulin pharmacodynamics, potentially improving matching of insulin action to glucose appearance. However, using the same carbohydrate-count–based insulin dose for both fresh and cooled rice increased hypoglycemia risk after cooled rice, likely because effective available carbohydrate was lower. This suggests practical implications: insulin dosing should be reduced when consuming cooled rice to mitigate hypoglycemia. The study controlled for confounders, including standardized meal composition, identical breakfast macronutrients and insulin doses, stable basal insulin, and fixed timing, strengthening the causal interpretation. Findings are consistent with several non–type 1 diabetes studies showing reduced glycemic responses with cooled starch, and extend these observations to people with type 1 diabetes. Palatability and subjective appetite were unchanged, supporting feasibility of adopting cooled rice without compromising eating experience.

Cooling rice before consumption reduces postprandial glycemic rise in adults with type 1 diabetes but increases hypoglycemia risk if insulin dosing is not adjusted. Practical takeaway: consider reducing prandial insulin for meals with cooled rice. The cooling process did not alter organoleptic properties or subjective hunger/satiety. Future research should define insulin dose adjustment algorithms for cooled starch meals, explore multiple heating/cooling cycles to optimize resistant starch while ensuring safety, and use patient-blinded CGM to minimize potential bias. Larger, diverse cohorts and longer follow-up could assess generalizability and longer-term metabolic effects.

- Single-blind design with FreeStyle Libre providing real-time glucose could allow participants to infer meal condition; patient-blinded CGM could reduce this bias. - Same insulin dosing was used for both conditions based on carbohydrate counting, likely contributing to increased hypoglycemia after cooled rice; dose-adjustment strategies were not tested. - Short postprandial monitoring window (180 min). - Single-center study with a modest sample size of young adults with T1D on pump therapy; generalizability to other populations or therapies may be limited. - Only one 24-h cooling and reheating cycle was evaluated; other cooling protocols were not examined.