1005-P: Patients with Overweight and Obesity at High-Risk for Type 2 Diabetes: A Predictive Model

Obesity is the most common chronic disease in the U.S. Patients with obesity are at high risk for type 2 diabetes mellitus (T2DM); however, many of their risk factors for T2DM are modifiable. It is important to identify patients with obesity at high risk for T2DM to appropriately direct treatment and resources. We conducted an observational study of EMR data of 399,711 adults with BMI ≥ 25 kg/m2 and without baseline T2DM between 2000 and 2019. We first identified a literature-based pool of risk factors for T2DM onset and conducted variable selection by applying LASSO penalized Cox regression with ten-fold cross-validation on an 80% training dataset. We estimated a regular Cox model on the training dataset with selected variables and evaluated model performance on the remaining 20% dataset. Over a mean of 7.4 years of follow-up, 38,898 (9.8%) study patients developed T2DM. The predictive model achieved a Harrell’s C-statistic of 0.81. The greatest risk of T2DM was associated with gestational diabetes (HR 2.53; 95% CI 2.12-3.01), Asian and Hispanic race/ethnicity (HRs 2.38 and 1.94; 95% CIs 2.24-2.53 and 1.86-2.02, respectively), family history of DM (HR 2.00; 95% CI 1.95-2.04) and PCOS (HR 1.84; 95% CI 1.63-2.07). Female sex (HR 0.77; 95% CI 0.75-0.79), having a partner (HR 0.90; 95% CI 0.88-0.92) or commercial insurance (HR 0.86; 95% CI 0.84-0.88) were associated with a lower risk of T2DM. Characteristics found in >5% of patients that carried additional risk included proteinuria (5.0% of patients; HR 1.53; 95% CI 1.47-1.59), hypertension (23.9%; HR 1.21; 95% CI 1.18-1.25), high cholesterol (11.9%; HR 1.19; 95% CI 1.15-1.23), and history of smoking (40.5%; HR 1.18; 95% CI 1.15-1.21). Each standard deviation increase in BMI (5.2 kg/m2) was associated with a 1.51-fold (95% CI 1.49-1.52) increase in T2DM onset. It is feasible to use predictive modeling to identify patients with overweight and obesity at high risk of T2DM. This approach could be utilized to guide population management and clinical treatment decisions. Disclosure A. Turchin: Consultant; Self; Proteomics International, Research Support; Self; AstraZeneca, Eli Lilly and Company, Novo Nordisk, Pfizer Inc., Sanofi, Stock/Shareholder; Self; Brio Systems. F. J. Morrison: None. M. Shubina: None. S. Shinde: Employee; Self; Eli Lilly and Company, Stock/Shareholder; Self; Eli Lilly and Company. N. Ahmad: Employee; Self; Eli Lilly and Company. H. Kan: Employee; Self; Eli Lilly and Company, Stock/Shareholder; Self; Bristol-Myers Squibb Company, GlaxoSmithKline plc. Funding Eli Lilly and Company