Project Abstract
Almost half of adults have either diabetes mellitus (DM) or prediabetes (preDM), but many of those have
undiagnosed conditions. Current DM diagnosis and risk prediction are based on single “snapshot”
measurements, including fasting blood glucose, postprandial glucose, and hemoglobin (Hb)A1c. However,
some individuals that do not fall into high-risk categories by traditional DM biomarkers have occasional
glycemic excursions similar to individuals with preDM or even DM. Understanding the determinants of these
glycemic patterns and whether they confer risk in developing DM will elucidate the pathophysiology
responsible for the progression toward DM and could improve DM risk prediction. We will use continuous
glucose monitoring (CGM) to measure glycemic patterns in 2700 adults (mean age 58 years) from the
community-based Framingham Heart Study Third Generation cohort and Omni 2, a multi-ethnic cohort. We
aim to describe normative glycemic patterns in a large sample of healthy individuals (most of whom do not
have DM), exploring how standard clinical measures (body mass index, fasting blood glucose, HbA1c), blood
metabolites, gut microbiome, dietary patterns, physical activity, family history of DM, and polygenic risk score
for DM relate to CGM-derived glycemic variables. The CGM-derived variables we will explore include time
spent in glycemic ranges (e.g. 70-140 mg/dL), hyper/hypoglycemic episodes, mean glucose and glucose
variability. Furthermore, we will examine whether CGM-derived glycemic variables predict development of DM
over 2-3 years follow-up (through an annual self-reported health history) and relate to prevalence of
cardiovascular disease (cross-sectionally). Our study will provide information that could improve the prediction
of developing DM and DM complications. Our findings may also lead to new discoveries that will tailor and
target prevention of DM.