Variables associated with endogenous hyperinsulinism in hypoglycemia diagnosis. Could the 72-hour fasting test be shortened in low-risk patients?

Abstract

Background

The 72-hour fasting test remains the standard for the diagnosis of endogenous hyperinsulinism. We investigated which variables could identify patients at low risk for endogenous hyperinsulinism, in whom a shortening of the 72-hour fasting test could be considered.

Methods

This multicenter, retrospective study included 64 individuals (46 women, median age 45 years) without diabetes who underwent 72-hour fasting tests for the etiologic diagnosis of hypoglycemia. Pre- and intra-test variables were collected, including point-of-care glucose trajectories during the test. Testing was stopped before 72 h if symptomatic serum glucose <55 mg/dL or asymptomatic serum glucose ≤45 mg/dL occurred. Endogenous hyperinsulinism was diagnosed in individuals who had serum glucose <55 mg/dL, serum insulin ≥3.0 μU/mL, and serum C-peptide ≥0.6 ng/mL.

Results

Patients with endogenous hyperinsulinism (n = 10) had steeper descending point-of-care glucose trajectories (p < 0.001) than those without it. Older age and lower minimum pre-test serum glucose concentrations were independently associated with endogenous hyperinsulinism. A calculator for probability prediction of endogenous hyperinsulinism was developed including these variables and sex (AUC = 0.94). Older age, female sex, lower body mass index, and lower minimum point-of-care glucose during the first 24 h of fasting were independently associated with serum glucose <55 mg/dL after the first 24 h of fasting. A calculator for predicting probability of serum glucose <55 mg/dL after the first 24 h of fasting was developed including these variables (AUC = 0.84).

Conclusions

Pre- and intra-test variables can identify individuals at low risk for endogenous hyperinsulinism, in whom shortening the 72-hour fasting test could be considered.