Anthropometric features are not predictive of 72-hour fast duration in insulinomas

Objective: The 72-hour fast is used to document Whipple's triad and understand the mechanism of hypoglycemia. Although hypoglycemia develops within 24 hours in the majority of fasts, identifying possible determinants of fast duration may help to predict the need for admission. Therefore, we determined the relation between anthropometric features on fast duration and assessed end of fast parameters on maximal tumor size, extent of disease, or tumor recurrence. Methods: A retrospective analysis of patients with insulinoma in the past 25 years who underwent a 72-hour fast was conducted. Electronic medical records were reviewed to obtain anthropometric patient data and tumor characteristics. Results: A total of 233 patients underwent the 72-hour fast. The mean age at diagnosis was 50 ± 16 years, with a body mass index (BMI) of 29 ± 7 kg/m2, and 66% (153 of 233) were female. Duration of fast was not associated with gender (P = .2), age (P = .3), or BMI (P = .7). A shorter fast duration was inversely related to end of fast C-peptide (P = .0075) but not insulin (P = .13) or proinsulin (P = .28) concentration. End of fast C-peptide was associated with increased tumor size (P = .036) and multiplicity (P =.01). Proinsulin was associated with increased tumor size (P<.01) and malignancy (P = .018). Conclusion: Duration of fast was not significantly related age, gender, weight, or BMI, although end-of-fast C-peptide and proinsulin may provide some information regarding tumor characteristics. Consequently, the duration of fast cannot be predicted a priori and should be allowed to run for the planned length unless hypoglycemia develops.