A within-subjects cross-over trial investigated the effects of low versus high GI meals on blood glucose (BG). Obese youths between the ages of 7–16 years, diagnosed with T2D or IGT were recruited through a pediatric endocrinology clinic from June-August of 2006. Eligible youths were identified by their health care provider and referred to the study. Informed parent consent and child assent were obtained. The study protocol was approved by the National Institute of Child Health and Human Development Institutional Review Board.
Subjects were served a full day of high glycemic index (GI) and a full day of low GI mixed meals in a supervised clinic setting, matched for macronutrient content, in counterbalanced order ( and ). Standard washout dinners (sandwich and chips) and snacks (cereal with milk) were consumed at home the evening prior to both conditions. The two conditions were separated by one day, during which participants consumed their typical breakfast and lunch at home, followed by the same wash-out dinner and evening snack. Subjects’ BG levels were assessed using the Medtronic MiniMed Continuous Glucose Monitoring System. Continuous glucose monitoring (CGM) began in the fasting state and continued throughout the study period.
Meals served for high glycemic index dietary conditions
Meals served for low glycemic index dietary conditions
Daytime (8:00 am to 8:00 pm) and nighttime (8:01 pm to 7:59 am) mean BG were calculated. Because the extent of BG variability may be relevant in addition to overall BG (6
), a measure of intra-day glycemic variability, four-hour continuous overlapping net glycemic action (CONGA4) was calculated. CONGA4 is defined as the standard deviation of the differences in BG for every 4-hour pair of values; higher values indicate greater BG variability (7
). Paired-sample t-tests were used to test for treatment condition differences in mean BG and BG variability for each time period. Data analyses were conducted using SPSS version 13.0. Previous data on the effect of a low GI diet on BG in adults with T2D (8
) indicated that 8 sets of paired data provide 85% power to detect a 15% difference in BG. Because we were unable to recruit 8 subjects for each diagnosis (T2D & IGT), data were combined for primary analyses, with subgroup analyses conducted for descriptive purposes only.