Continuous glucose monitoring evidence of celiac disease in type 1 diabetes.

Quantitative glycemic metrics are needed to identify undiagnosed celiac disease in type 1 diabetes and reduce delays in celiac diagnosis. Celiac enteropathy drives malabsorption that increases the risk of prandial insulin-glucose mismatch and hypoglycemia. We assessed if children with type 1 diabetes and celiac disease have lower post-prandial glucose levels preceding celiac diagnosis vs. those without celiac disease, leveraging continuous glucose monitoring (CGM) data and a computational meal annotation algorithm. In this retrospective cohort study, children with type 1 diabetes <12 months duration using CGM, positive celiac serologies and biopsy confirmed celiac disease (n=16) were matched 1-to-4 to those with negative celiac serologies (n=60). Meals were computationally annotated in the 30-day window before serologies. Differences in post-prandial trough glucose and other prandial glycemic outcomes were assessed via mixed models. Undiagnosed celiac disease was associated with a lower glucose rise from meal start to peak vs. no celiac disease (-8.9 %, 95 % CI, -14.9--2.5 %, p=0.009) and, during the first meal of the day, a lower fall from peak to trough (-9.3 %, 95 % CI, -16.5 %--1.5 %, p=0.02). There was no significant association between celiac disease and trough glucose, meal hypoglycemia or time hypoglycemic. Computational analysis revealed that blunted prandial glycemic trajectories, not hypoglycemia, are associated with undiagnosed celiac disease in children with type 1 diabetes using CGM. These findings challenge current guidelines, and future studies should validate and integrate these glycemic biomarkers into a CGM-based model for real-time prediction of celiac disease in type 1 diabetes.