385-P: A More Accurate Method to Diagnose Impaired Awareness of Hypoglycemia (IAH) in People with Type 1 Diabetes (T1D)

Traditional questionnaires used to define IAH have been criticized for high variability, recall bias, and lack of sensitivity. This study in T1D subjects sought to compare awareness of hypoglycemia determined with self-reported questionnaires (Gold Score (GS)) with actual symptom scores (SS) assessed during hyperinsulinemic, stepped-hypoglycemic clamps (blood glucose (BG) lowered to 100, 65, 55, and 45 mg/dL). Neurogenic and neuroglycopenic symptoms were assessed at each BG level as were counterregulatory hormone responses (CRR). There was no statistically significant correlation (R2=0.27, p=0.37) between GS and SS assessed during the hypoglycemic clamp (Figure 1A) indicating that self-reported hypoglycemia awareness questionnaires do not accurately reflect symptoms of hypoglycemia. Receiver-operating curves indicated a cutoff value for peak SS of < 25 as accurately identify IAH in T1D subjects (sensitivity 88%, specificity 83%)(Figure 1B). Epinephrine levels were correlated with SS (R2=0.57, p<0.0001) but not GS (R2=0.02, p=0.59) suggesting SS, but not GS, appropriately identified subjects with impaired CRR. It is concluded that 1) hypoglycemia questionnaires have major limitations in accurately identifying T1D subjects with IAH and impaired CRR, and 2) though more labor intensive, accurate diagnoses of IAH should be based on low peak SS obtained during a hypoglycemic clamp.