Diets of low glycemic load may dampen the postprandial glycemic response, thereby avoiding high blood glucose concentrations that could be detrimental to health (1). We compared blood glucose profiles of two nondiabetic subjects (A and B) consuming meals of high (baguette, strawberry jam, and maltose), medium (baked potato, cheese, and Coca Cola), or low (chickpeas, tuna, vinegar, and oil) glycemic load at regular intervals throughout the day.
Three test meals contained the same calorie content but different glycemic load for each subject (A: glycemic load = 92, 49, and 19; B: glycemic load = 115, 66, and 24). For each glycemic load category, three full portions of the test meal with a 4-h interval in between and six half-portions with a 2-h interval in between were consumed on 2 different days. For each subject, six 12-h blood glucose profiles deduced from the interstitial glucose in subcutaneous abdominal tissue measured by MiniMed continuous glucose monitoring system were obtained.
A relatively stable blood glucose pro-file was observed throughout the day with low–glycemic load meals for both subjects (Fig. 1). Consumptions of high–and medium–glycemic load meals were usually followed by peaks of blood glucose. However, there did not appear to be an obvious dose-response effect between the actual glycemic load and the height of the peaks (either full portion versus half portion or high glycemic load versus medium glycemic load), suggesting a possible “threshold” effect (2). Nibbling diets with small frequent meals may only help avoid hyperglycemia when the meal glycemic load is below a certain threshold level. Compared with the glucose response of the first meal, some of those triggered by each subsequent but identical meal appeared to be lower. This apparently greater “breakfast” glycemic response may be due to higher ACTH and glucocorticoid levels before awakening. Since glycemic index values of food are derived in the fasting state, the glycemic load formula may give better prediction of the postprandial glucose response for breakfast than those for lunch or dinner.
Calculated meal glycemic load may deviate from the actual glycemic response of food combinations. Potential limitations of the continuous glucose monitoring system also need to be considered when interpreting our glycemic profiles (3). In this study, 25% of the sensor-deduced blood glucose concentrations deviated by ≥15% from the corresponding fingerstick glucometer values (for calibration) among 48 paired values.
Our pilot study suggests that a stable blood glucose profile can be maintained by consuming a low–glycemic load diet. However, meal glycemic load may need to be below a certain threshold to be of benefit. Identical meals may produce different blood glucose responses at different times of the day, indicating that the glycemic load formula may not predict the postprandial glucose response for meals eaten in the nonfasting state.
Twelve-hour glucose profiles reported by continuous glucose monitoring system on the 6 experimental days for subject A consuming high–, medium–, and low–glycemic load (GL) meals as full portions (A; three meals) or as half portions (B; six meals) and subject B consuming high–, medium–, and low–glycemic load (GL) meals as full portions (C; three meals) or as half portions (D; six meals).
Twelve-hour glucose profiles reported by continuous glucose monitoring system on the 6 experimental days for subject A consuming high–, medium–, and low–glycemic load (GL) meals as full portions (A; three meals) or as half portions (B; six meals) and subject B consuming high–, medium–, and low–glycemic load (GL) meals as full portions (C; three meals) or as half portions (D; six meals).
