|KRISHNAN, SRIDEVI - University Of California|
|HEMBROOKE, TARA - University Of California|
Submitted to: Journal of Nutrition and Metabolism
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/13/2012
Publication Date: 3/29/2012
Citation: Krishnan, S., Newman, J.W., Hembrooke, T.A., Keim, N.L. 2012. Variation in metabolic responses to meal challenges differing in glycemic index in healthy women: Is it meaningful?. Journal of Nutrition and Metabolism. 9(1):26.
Interpretive Summary: In human nutrition research, it is common to observe a large degree of individual variation in response to specific dietary interventions or challenges. New approaches to data analysis could be useful as the nutritional sciences field moves toward the goal of personalizing nutritional needs and dietary recommendations. We applied two statistical tools, range-scaling and principal component analysis, to enable the identification of typical and atypical metabolic responders to meals with low glycemic index and high glycemic index. Despite the fact that all of our research participants were healthy and most responded as expected to the meal challenges, we were able to identify a small subset with sub-clinical insulin resistance and another subset with higher concentrations of leptin in their blood. These distinct sub-groups may have increased risk for developing metabolic diseases, like diabetes, and dietary guidance may need to be adjusted to offset this risk. Follow-up study is needed to confirm disease susceptibility, using this approach.
Technical Abstract: Background: Metabolic phenotyping has potential utility as a diagnostic tool. While clinical parameters are commonly used in disease diagnosis, tools that enhance diagnosis of metabolic dysfunctions are needed. Objective: To identify typical and atypical metabolite temporal patterns in response to paired meal challenge tests. Design: Metabolic responses to high and low glycemic index (GI) meals were tested in 24 healthy pre-menopausal women, aged 18-35y, with BMI of 20-30Kg/m2 using a cross-over design. On test days, blood glucose, insulin, leptin and non-esterified fatty acids were measured at fasting, and for 8hr following test meal consumption. The data were range scaled and analyzed to assess the presence of distinct response phenotypes to the meal challenge tests. Results: Participants showed higher circulating glucose and insulin in response to the high GI compared to the low GI meal challenge. Using range-scaling and Principal Component Analysis three distinct groups were identified based on differential responses to the paired challenges. Members of the most populated phenotype (n=18) displayed little deviation from the expected response to the two meal challenges. Two minor phenotypes (n=3/group) with distinct responses were observed, one consistent with sub-clinical insulin resistance, and the other suggestive of hyperleptinemia. Conclusions: The differential responses of glucose, insulin and leptin to low and high glycemic test meals revealed three distinct phenotypes. Future efforts should evaluate whether these phenotypic sub-groups have differential susceptibility to disease and or response to nutritional interventions.