Exploring Corresponding Cutpoints for Hypoglycaemia Awareness Questionnaire ( HypoA ‐Q) Impaired Awareness Subscale in Relation to the Gold and Clarke Measures: A Post Hoc Analysis From the Hypo‐ METRICS Study

Abstract

ABSTRACT Aims The Hypoglycaemia Awareness Questionnaire–Impaired Awareness subscale (HypoA‐Q/IA) was developed to enhance impaired awareness of hypoglycaemia (IAH) assessment, addressing methodological deficiencies of the established Gold and Clarke measures. An IAH cutpoint (HypoA‐Q/IA ≥ 12) was identified from a clamp study in Type 1 diabetes (T1D). We explored cutpoints in relation to Gold and Clarke measures, and their co‐occurrence, in the larger, real‐world observational Hypo‐METRICS study, in people with T1D (pwT1D) and insulin‐treated Type 2 diabetes (pwT2D). Material and Methods Hypo‐METRICS recruited pwT1D or pwT2D with ≥ 1 hypoglycaemic event in the previous 3 months. IAH was defined as Gold ≥ 4, Clarke ≥ 4 and/or HypoA‐Q/IA ≥ 12. Bidirectional ROC analysis was performed between HypoA‐Q/IA and Gold/Clarke cutpoints. Results Questionnaires were completed by 232 pwT1D and 285 pwT2D (male 47% vs. 62%, p < 0.001; median age 49 vs. 63 years, p < 0.001; HbA1c 7.3% vs. 7.4%, p = 0.03). IAH prevalence was 24% by Gold ≥ 4, 16% by Clarke ≥ 4 and 7% by HypoA‐Q/IA ≥ 12. ROC analysis yielded HypoA‐Q/IA cutpoints of ≥ 7 against Gold ≥ 4 (AUC 0.87) and ≥ 9 against Clarke ≥ 4 (AUC 0.89), and cutpoints of Gold ≥ 4 (AUC 0.87) and Clarke ≥ 3 (AUC 0.93) against HypoA‐Q/IA ≥ 12. Of those with IAH by either measure, 24% were identified by both HypoA‐Q/IA ≥ 12 and Gold ≥ 4, and 32% by both HypoA‐Q/IA ≥ 12 and Clarke ≥ 4, with lower overlap in T2D than T1D. Conclusions HypoA‐Q/IA cutpoints corresponding to legacy IAH cutpoints (≥ 7–9) were lower than the clamp‐derived threshold (≥ 12), suggesting ≥ 12 may identify a subgroup with more severe IAH. These cutpoints enable cross‐study comparison; prospective validation against severe hypoglycaemia risk is needed.