AS737 - Defining the clinical characteristics of inflammatory and insulinemic dietary patterns using feeding study dietary data and blood biomarkers

Investigator Names and Contact Information

Introduction/Intent

Abstract:

Background: The empirical dietary index for hyperinsulinemia (EDIH)[1, 2] and empirical dietary inflammatory pattern (EDIP)[3, 4] are two empirical hypothesis-oriented dietary indices developed to assess the ability of the diet to contribute to chronic insulin hypersecretion and chronic systemic inflammation, respectively. Both indices were developed in the Nurses’ Health Study (NHS) using food frequency questionnaire (FFQ) data and plasma markers of insulin response (C-peptide) for EDIH and inflammation (CRP, IL-6, TNFα-R2 and adiponectin) for EDIP. Both indices have been calculated only from FFQ data and have shown strong associations with disease endpoints in multiple studies including the Women’s Health Initiative.[5, 6] However, EDIH and EDIP have never been estimated from food records, 24-hour diet recalls (24HR) or feeding study dietary data (gold standard), dietary assessment instruments typically used in clinical studies. It is therefore important to determine if EDIH and EDIP, calculated from these instruments, will retain their respective constructs of dietary insulinemic or inflammatory potential.

Aim: We will calculate EDIH and EDIP scores from feeding study dietary data, 4-day food records (4DFR) and 24-HR and determine if the indices from these dietary assessment instruments are measuring what we expect them to measure, that is, predict insulinemia and inflammation respectively. In other words, we will assess the construct validity of the indices against plasma C-peptide for EDIH, and CRP, IL6, TNFA-R2 and adiponectin for EDIP.

Methods: We will use dietary data and plasma biospecimens from the WHI Nutrition and Physical Activity Assessment Study Observational Study (NPAAS-OS: AS218) and Feeding Study (NPAAS-FS: AS272). NPAAS-OS recruited 450 WHI-OS women to examine the measurement properties of dietary self-report data against some nutrient biomarkers. NPAAS-FS was a controlled-feeding study in which 4DFR from 153 women were used to compose study diets for a 2-week feeding study to assess the impact of habitual diet on dietary biomarkers. We have funds to assess new plasma biomarker data in NPAAS FS and OS. For our construct validation objective, we will determine associations of the dietary indices with their construct biomarkers in multivariable-adjusted linear regression models. We will request fasting plasma samples for 441 women in NPAAS-OS (visit 2), 90 women in the NPAAS-OS reliability study and 152 in NPAAS-FS (visits 1 and 2 = 304); for a total of 835 plasma samples (20µL per sample).

Specific Aims:

Aim: To conduct a construct validation study of EDIH and EDIP using plasma biomarkers from the NPAAS (FS, n=152 and OS, n=441), and reliability study (n=90). We hypothesize that: (a) EDIH and EDIP significantly predict concentrations of the plasma biomarkers representing the constructs of the two dietary indices in consumed food data and in self-reported food data from 4DFR, 24HR and FFQ; with the strength of associations decreasing in the following order: consumed dietary data < 4DFR < 24HR < FFQ. (b) EDIH and EDIP as mechanisms-based dietary patterns, are more predictive of the biomarkers compared to traditional dietary patterns such as Healthy Eating Index (HEI)-2015.