EARLY METABOLIC IMBALANCE IS A RISK FACTOR FOR INCIDENT PRE-DIABETES: CARDIA 30-YEAR FOLLOW-UP

Abstract

Therapeutic Area: Other: Pre-Diabetes risk factors Background: Early metabolic imbalance (EMI) is an overlooked condition characterized by compensated insulin resistance in the face of normal fasting glucose, triglycerides, hemoglobin A1c, and high-density lipoprotein cholesterol (HDL-C). Thus, individuals with EMI elude routine screening for diabetes and cardiovascular risk. Previously, we reported that EMI is an independent risk factor for type 2 diabetes (T2D) and atherosclerotic cardiovascular disease (CVD). Hypothesis: EMI is an independent risk factor for prediabetes (PreD). Methods: This was a retrospective cohort analysis of CARDIA: Coronary Artery Risk Development in Young Adults. The parent study began in 1985, enrolling 5,114 young adults ages 18-30, with study visits every five years for 30 years. For this retrospective analysis, the baseline exclusion criteria were fasting hyperglycemia, hypertriglyceridemia, low HDL-C, pregnancy, diabetes, or CVD; n=3,292. Cox proportional hazards regression analysis was performed using JMP Pro 17.2 (SAS Institute) and Stata 18.0 (StataCorp). The primary exposure was EMI: baseline homeostatic model assessment of insulin resistance v.2 (HOMA2-IR) at/above the median cutpoint; the baseline co-variates included canonical T2D risk factors. The primary outcome was time-to-incident PreD, defined as ≥2 blood glucose or A1c values consistent with PreD, without a prior history and treatment for overt T2D. Results were reported as a Cox hazard ratio (HR) with 95% confidence interval (CI) and p-value. Results: Over the 30-year follow-up period, 870 cases of incident PreD occurred: cumulative incidence 26.4%. Cox Model 1 included a categorical variable (high/low) for HOMA2-IR at baseline (Table 1). Model 2 included a categorical variable for the interaction between HOMA2-IR, sex, and race. Model 3 included the 3-way categorical variable from model 2, as well as 5 additional covariates including canonical T2D risk factors. Overall, the hazard ratios for EMI (high vs. low HOMA2-IR) were modified by race and sex (Table 1). Conclusions: Young adults in CARDIA with EMI had an increased risk for future prediabetes, even after adjusting for known risk factors. The hazard ratios were modified by sex and race. These findings imply that EMI is associated with variable degrees of pancreatic beta cell decline over time.