Content area
Abstract
Background: Type 2 diabetes mellitus (T2DM) is a public health threat that jeopardizes the health of current and future generations in the United States. Reduced insulin sensitivity, also called insulin resistance, is a major contributing factor in the development of glucose intolerance and T2DM. Insulin sensitivity is impacted by numerous dietary and non-dietary factors. In dietary intervention studies, emerging research has found that individuals do not respond the same way to standardized dietary interventions, suggesting that factors other than the primary dietary exposure of interest influence an individual’s response.
Aims: We conducted a secondary data analysis using primary data from the DAIRY randomized dietary intervention trial (Schmidt et al., Submitted for publication) to assess the extent to which alternative dietary factors that were not controlled for (i.e. healthy eating index independent of dairy (HEI-ID), fiber intake and added sugar intake) might at least partly explain group or interindividual differences in the impact of a dairy intervention on insulin sensitivity in individuals with the metabolic syndrome. Aim 1: Assess the impact of adjusting for (a) baseline or (b) changes in HEI-ID, fiber intake, and added sugar intake on the differential effect of the dietary interventions on the Matsuda-DeFronzo insulin sensitivity index (ISI, primary) and homeostatic model assessment of insulin resistance (HOMA-IR, secondary). Aim 2: Investigate whether interindividual differences in the Matsuda ISI (primary) and HOMA-IR (secondary) are explained by (a) baseline or (b) changes in HEI-ID, fiber intake, and added sugar intake.
Methods: We created a database of participant-level dietary intake data from 24-hour dietary recalls, using reports generated from the Nutrition Data System for Research (NDSR) software and database. Participant-level data was broken out as “baseline” (average of 2 recalls) and “intervention” (average of 3 recalls) for all variables. To address Aim 1, we conducted both per protocol and intent-to-treat repeated measures analysis of variance (RM-ANOVA) for Matsuda ISI (primary) and HOMA-IR (secondary), adjusting for (a) baseline or (b) changes in HEI-ID, fiber intake, and added sugar intake. To address Aim 2, we conducted both per protocol and intent-to-treat multiple linear regression analyses for Matsuda ISI (primary) and HOMA-IR (secondary), including dietary intervention group, as well as (a) baseline or (b) changes in HEI-ID, fiber intake, and added sugar intake, as additional independent variables.
Results: Adjusting both per protocol and intent-to-treat analyses for baseline and changes in HEI-ID, fiber intake or added sugar intake between wash-in and intervention diet periods did not impact the significant decreases in the Matsuda ISI and increases in HOMA-IR observed in the low-fat and full-fat dairy intervention groups in the main study. With the exception of one weakly significant association between change in HEI-ID and change in HOMA-IR , neither baseline, nor changes in HEI-ID, fiber intake or added sugar intake between wash-in and intervention diet periods, predicted interindividual differences in the Matsuda ISI or HOMA-IR that resulted from the dairy dietary intervention.
Conclusions: These results demonstrate that the impact of low-fat and full-fat dairy on insulin sensitivity in individuals with the metabolic syndrome, as observed in the primary study analysis, was independent from dietary changes in fiber, added sugar, or the HEI-ID that resulted from the incorporation of dairy into study participants’ diets. These effects occurred independent of changes in fat mass. Aside from the one weakly significant association noted above, these results also reveal that HEI-ID, fiber intake and added sugar intake did not predict changes in insulin sensitivity in the context of this intervention study.
