What question did this study set out to answer?

The study aims to create a risk prediction model for impaired awareness of hypoglycemia in diabetes and analyze its associations with clinical factors.

March 26, 2026Open Access

Construction of a Risk Prediction Model and Associative Path Analysis for Impaired Awareness of Hypoglycemia in People with Diabetes

Puntos clave

The study aims to create a risk prediction model for impaired awareness of hypoglycemia in diabetes and analyze its associations with clinical factors.
Cross-sectional design involving 280 hospitalized diabetes patients
Defined impaired awareness of hypoglycemia using Gold Scale score≥4
Utilized binary logistic regression for model construction
Employed structural equation modeling for path analysis
Conducted sensitivity analysis based on diabetes types
Impaired awareness of hypoglycemia was found in 60.4% of participants
Lower total cholesterol levels were identified as an independent correlate (aOR=0.819)
The model exhibited moderate discriminative ability (AUC=0.630) with high sensitivity (87.0%)
Path analysis indicated age impacts hypoglycemia risk indirectly through glomerular filtration rate and peripheral neuropathy

Resumen

Purpose: We aimed to develop a risk prediction model for impaired awareness of hypoglycemia (IAH) in diabetes and explore the associations with clinical variables, providing a basis for targeted screening and personalized intervention. Patients and Methods: This cross-sectional study included 280 hospitalized people with diabetes (73.2% type 2 and 42.9% female) who experienced hypoglycemia between October 2023 and February 2024. We defined IAH as a Gold Scale score ≥ 4. We used binary logistic regression to construct the prediction model and structural equation modeling to analyze the associated pathways. Sensitivity analysis stratified by diabetes type verified the robustness of the results. Results: IAH affected 60.4% of the participants. Multivariate analysis revealed lower total cholesterol (TC) levels (aOR=0.819; 95% CI: 0.684– 0.981; P=0.030) as independent correlates. The model, which incorporated BMI, TC level, serum albumin level, peripheral neuropathy (PN) incidence, glomerular filtration rate (GFR), age, sex, and insulin use, showed moderate discriminative ability (AUC=0.630, P< 0.001), with high sensitivity (87.0%) and good calibration (Hosmer–Lemeshow test: P=0.743). Path analysis revealed that age indirectly influenced IAH risk (β=0.243) by decreasing the GFR (β=− 0.419; P< 0.001) and increasing PN risk (β=0.352; P< 0.001) and that a lower GFR (β=− 0.133; P=0.024) and TC level (β=− 0.131; P=0.024) were directly correlated with IAH risk. The serum albumin level was maintained by preserving the GFR (β=0.203; P< 0.001) and reducing PN risk (β=− 0.139; P=0.012). Conclusion: Higher BMI and lower TC are key correlates of IAH in hospitalized patients with diabetes. Age, GFR, serum albumin, and PN interact through direct and indirect pathways to affect the risk of IAH. This model, based on routine clinical indicators and with high sensitivity, represents a practical screening tool. Clinicians should perform early risk assessment and intervention in diabetic patients with higher BMI, lower TC, advanced age, malnutrition, renal insufficiency, or PN. Keywords: diabetes mellitus, hypoglycemia, impaired awareness of hypoglycemia, risk prediction model, path analysis

Preguntar a la IA

Me gusta

Guardar

Ver artículo completo