Glycemic Impact of Low Glycemic Index Biscuits on Postprandial Glucose Response

The study addresses whether glycaemic index (GI) differences within the same low-GI category meaningfully affect postprandial glycaemic responses and explores potential second-meal effects. While many studies compare high- versus low-GI foods, little is known about differential impacts among foods that are all classified as low GI but have different GI values. The authors aimed to compare a basic low-GI biscuit (GI ~54) with a modified lower-GI biscuit (GI ~24) on postprandial glucose and insulin responses in healthy, young, non-diabetic males, and to examine second-meal effects across subsequent meals.

Prior research shows that postprandial blood glucose is a strong predictor of long-term health outcomes, and low-GI foods can improve glycaemic control. Most studies compare low versus high GI categories, leaving a gap regarding differences within the low-GI range. The second-meal effect—where one meal influences the glycaemic response to a subsequent meal—has been documented, with low-GI meals generally reducing later glycaemic excursions more than high-GI meals. However, second-meal effects between foods both classified as low GI have not been well studied. Literature also suggests roles for soluble fiber, fat type, and protein source in modulating glycaemic responses, with evidence that functional fibers and plant-based proteins may attenuate postprandial glycaemia, and that certain fats (e.g., coconut oil/MCTs) may reduce glycaemic responses via delayed gastric emptying and amylose–lipid complex formation.

Design: Randomized, controlled, single-blind, cross-over trial with two sessions per participant, each spanning three consecutive days, separated by ≥3-day washout. Participants: Healthy, young Asian Chinese males (n=13 analyzed; age 21–40 y; BMI 18.5–25 kg/m²; non-smokers; normal blood pressure; exclusion of metabolic disease, medications affecting glycaemia, food intolerances/allergies, competitive/endurance athletes, intentional food restriction, fasting glucose >6 mmol/L). Study approved by NHG DSRB (2018/01066) and registered (NCT04115579). Interventions: Two biscuit treatments:

• LGI 1: basic low-GI biscuit (GI 54.4 ± 6.3) made with all-purpose flour, butter, sugar, vanilla, baking soda, egg, salt.
• LGI 2: modified lower-GI biscuit (GI 23.8 ± 3.3) with plain flour + soluble fiber + plant-based protein (soy-derived), coconut oil replacing butter, and partial replacement of sugar with a low-GI sweetener. Dosing: Breakfast portion provided 50 g available carbohydrate; snack portion provided 25 g available carbohydrate. Meals: Standardized lunch (spaghetti with chicken sauce + fruit cocktail) and dinner (teriyaki chicken with rice + Milo drink) identical across sessions; only the biscuit treatment differed. Participants consumed only study foods and water; avoided alcohol and excessive activity. Procedures: Day 0 afternoon CGM insertion; Day 1 fasting arrival (~08:30–09:00), rest 10 min, IV cannula insertion, baseline blood, consume assigned breakfast biscuits within ~12–15 min with 250 mL water. Venous blood at 0, 30, 60, 90, 120, 150, 180 min for insulin. Standardized lunch consumed in 20 min; snack biscuits at 16:00 (within 15 min) at home; standardized dinner at 19:00 (within 20 min) at home. CGM removed Day 2 morning. Outcomes:
• Primary: Postprandial glycaemic response over 24 h via continuous glucose monitoring (iPro2, Medtronic). CGM calibrated four times daily. Data at 5-min intervals; baseline defined as average CGM readings over 30 min fasting on Day 1. Glycaemic response expressed as incremental area under the curve (iAUC) using trapezoidal rule; areas below baseline excluded.
• Secondary: Serum insulin response during breakfast over 180 min; iAUC calculated by trapezoidal rule (areas below baseline excluded). Insulin measured on Roche Cobas e411; intra-/inter-assay CVs <5% and <6%. Statistical analysis: Normality assessed by Shapiro–Wilk and Q–Q plots of differenced values. Paired t-tests compared mean iAUC between treatments; non-parametric tests for CV comparisons. Significance p<0.05. Power calculation indicated ≥8 subjects adequate to detect 15% change in 24-h AUC (power 0.85, alpha 0.05). Data presented as mean ± SEM or median (IQR) for CV.

• Fasting glucose before breakfast did not differ between treatments (p=0.61).
• Breakfast glycaemia: LGI 2 significantly reduced incremental glucose peak and iAUC at 0–1 h, 0–2 h, and 0–3 h compared with LGI 1 (all p<0.05). Example from table: incremental glucose peak at breakfast 0.4 ± 0.1 mmol/L (LGI 2) vs 1.0 ± 0.2 mmol/L (LGI 1), p=0.008.
• Breakfast insulin: LGI 2 led to a significantly lower incremental insulin response over 180 min than LGI 1 (iAUC 0–180 min, p=0.02), corresponding to a ~45% reduction.
• Magnitude: At breakfast (50 g available CHO), LGI 2 produced ~56.4% lower glucose response than LGI 1.
• Snack (25 g available CHO): LGI 2 showed a 24% reduction in glucose response versus LGI 1; not statistically significant but potentially physiologically relevant.
• Lunch: No significant second-meal effect at standard lunch; results unchanged after controlling lunch iAUC120 for breakfast iAUC120.
• Dinner: Significant second-meal effect observed; total iAUC120 at dinner was significantly lower for the LGI 2 treatment than LGI 1 (p<0.05).
• 24-h glycaemia: No significant difference in 24-h iAUC between treatments (median 24-h iAUC: LGI 2 393.0 [IQR 596.2] vs LGI 1 511.9 [IQR 280.4]; p=0.51).
• Glycaemic variability (CV): Significantly lower CV at breakfast with LGI 2 vs LGI 1 (p=0.04); borderline lower CV at snack (p≈0.05); small absolute CV values overall.
• Fasting insulin before breakfast did not differ (p=0.25).

The findings demonstrate that even within the low-GI category, lower absolute GI can substantially attenuate postprandial glycaemia and insulinaemia in healthy individuals. The modified lower-GI biscuit (LGI 2) markedly reduced early postprandial glucose and insulin at breakfast, with additional attenuation at dinner suggesting a delayed second-meal effect later in the day but not at lunch. The diurnal pattern may relate to higher evening insulin resistance and the role of plasma free fatty acid suppression in second-meal physiology. Potential mechanisms include the combined effects of functional ingredients in LGI 2: soluble fiber increasing chyme viscosity to slow carbohydrate digestion/absorption; coconut oil (rich in MCTs) delaying gastric emptying and forming amylose–lipid complexes that increase resistant starch; partial replacement of sucrose with a low-GI sweetener; and plant-based protein potentially enhancing glycaemic regulation relative to animal protein. The reduced glycaemic variability with LGI 2 suggests fewer glucose fluctuations, which may have long-term benefits for insulin sensitivity and cardiometabolic risk. The cross-over design and CGM under free-living conditions enhance ecological validity, showing that modest formulation changes to a common food can impact glycaemic profiles across a day.

Within the low-GI range, absolute GI value significantly influences postprandial glycaemic and insulin responses. A novel lower-GI biscuit (GI ~24) produced substantially lower breakfast glycaemia and insulinaemia than a basic low-GI biscuit (GI ~54), a modest attenuation at snack, and a significant second-meal effect at dinner, without altering 24-h iAUC. Reformulating staple products with functional ingredients (soluble fiber, plant-based fats, plant proteins, low-GI sweeteners) is a practical strategy to improve postprandial glycaemic control, potentially benefiting individuals at risk of type 2 diabetes and cardiovascular disease. Future research should evaluate broader populations (e.g., prediabetes), include additional metabolic biomarkers (e.g., lipids, satiety hormones), and delineate mechanisms underlying time-of-day second-meal effects.

• Small sample size (n=13) of healthy, young Chinese males limits generalizability to other ages, sexes, ethnicities, and to individuals with dysglycaemia or diabetes.
• Exploratory design; only glucose and insulin assessed; no measurements of plasma lipids or satiety/appetite biomarkers.
• The intervention differs only in biscuit formulation; while this isolates biscuit effects, it may not capture interactions with varied habitual diets.