Development and validation of empirical indices to assess the insulinaemic potential of diet and lifestyle

• Published in: British journal of nutrition 2016-11, Vol.116 (10), p.1787-1798
• Main Authors: Tabung, Fred K., Wang, Weike, Fung, Teresa T., Hu, Frank B., Smith-Warner, Stephanie A., Chavarro, Jorge E., Fuchs, Charles S., Willett, Walter C., Giovannucci, Edward L.
• Format: Article
• Language: English
• Subjects: Cholesterol; Diet; Dietary Surveys and Nutritional Epidemiology; Food; Food groups; Glycemic index; Insulin; Lifestyles; Nutrition research
• ISSN: 0007-1145; 1475-2662
• DOI: 10.1017/S0007114516003755

The glycaemic and insulin indices assess postprandial glycaemic and insulin response to foods, respectively, which may not reflect the long-term effects of diet on insulin response. We developed and evaluated the validity of four empirical indices to assess the insulinaemic potential of usual diets and lifestyles, using dietary, lifestyle and biomarker data from the Nurses’ Health Study (NHS, n 5812 for hyperinsulinaemia, n 3929 for insulin resistance). The four indices were as follows: the empirical dietary index for hyperinsulinaemia (EDIH) and the empirical lifestyle index for hyperinsulinaemia (ELIH); the empirical dietary index for insulin resistance (EDIR) and the empirical lifestyle index for insulin resistance (ELIR). We entered thirty-nine FFQ-derived food groups in stepwise linear regression models, and defined indices as patterns most predictive of fasting plasma C-peptide, for the hyperinsulinaemia pathway (EDIH and ELIH), and of theTAG:HDL-cholesterol ratio, for the insulin-resistance pathway (EDIR and ELIR). We evaluated the validity of indices in two independent samples from NHS-II and Health Professionals Follow-up Study (HPFS) using multivariable-adjusted linear regression analyses to calculate relative concentrations of biomarkers. The EDIH is comprised of eighteen food groups; thirteen were positively associated with C-peptide and five were inversely associated. The EDIR is comprised of eighteen food groups; ten were positively associated with TAG:HDL-cholesterol and eight were inversely associated. Lifestyle indices had fewer dietary components, and included BMI and physical activity as components. In the validation samples, all indices significantly predicted biomarker concentrations – for example, the relative concentrations of the corresponding biomarkers comparing extreme index quintiles in the HPFS were EDIH, 1·29 (95 % CI 1·22, 1·37); ELIH, 1·78 (95 % CI 1·68, 1·88); EDIR, 1·44 (95 % CI 1·34, 1·55); and ELIR, 2·03 (95 % CI 1·89, 2·19); all P trend