Diabetes RElated to Acute Pancreatitis and Its Mechanisms: Metabolic Outcomes Using Novel CGM Metrics

The primary objective of the DREAM-ON study is to determine if continuous glucose monitoring (CGM) metrics can predict the incidence of prediabetes mellitus (PDM) and diabetes mellitus (DM) after an episode of acute pancreatitis (AP). Secondary objectives of the DREAM-ON study include determining if CGM metrics predict the need for insulin therapy in participants who develop diabetes mellitus after AP, and if CGM metrics correlate with measures of insulin secretion and insulin resistance. The specific aims of the DREAM-ON study are as follows: Aim 1: To test whether standard CGM metrics predict incident DM. The investigators will perform blinded CGM in DREAM-ON participants at their scheduled visits at months 3, 12, 24 and subsequent annual visits. The investigators will test whether standard CGM metrics (mean glucose, time in tight range 70-140, time in range 70-180, time above 180 mg/dL, time above 250 mg/dL and glucose CV) predict incident DM determined by fasting plasma glucose (FPG), HbA1c, oral glucose tolerance testing (OGTT) and clinical report. Aim 2: To test whether CGM metrics predict need for insulin therapy in patients who develop DM after AP. From blinded CGM, we will test whether standard CGM metrics (mean glucose, time in tight range 70-140, time in range 70-180, time above 180 mg/dL, time above 250 mg/dL and glucose CV) as well as other indices of glucose variability, including mean amplitude of glycemic excursions (MAGE), predict need for long-term insulin therapy. Aim 3: To determine whether CGM metrics correlate with measures of insulin secretion and insulin resistance. The investigators will test whether standard and advanced CGM metrics correlate with measures of insulin secretion and insulin resistance derived from the OGTT, the mixed meal tolerance test (MMT) and the frequently sampled intravenous glucose tolerance test (FSIGTT). The investigators also will test whether these metrics can be used as a surrogate to predict diabetes subtype (i.e., insulin deficient vs. insulin resistant).