Water deprivation test in children: challenging but still necessary for the differential diagnosis of polyuria-polydipsia syndrome.

The differential diagnosis within polyuria-polydipsia syndrome, especially in the pediatric population, remains challenging. Despite its limited accuracy, the water deprivation test (WDT) is the reference test in pediatrics. We retrospectively analyzed WDT performed in 65 pediatric patients (mean age 8.2 yr; 0-17.9). We evaluated the accuracy and safety of the WDT, investigated possible baseline variables that could predict outcomes, and reevaluated interruption and interpretation criteria. Among the 83 WTDs performed, 47 (56.6%) were conclusive (accuracy = 53%). The simplified 2-h WDT was inconclusive in 47.6% of the sample. A baseline urinary osmolality (uOsm) greater than 279 mosmol/kgH2O excluded arginine vasopressin deficiency (AVP-D) and/or AVP resistance (AVP-R) in all patients (sensitivity = 100%; specificity = 74%; P < 0.01). A uOsm greater than 533 mOsm/kgH2O at the time of WTD discontinuation also excluded the diagnosis of AVP disorders in all subjects (sensitivity = 100%; specificity = 100%; P < 0.01). Therefore, a baseline uOsm > 279 mOsm/kgH2O and a peak uOsm > 533 mOsm/kgH2O excluded AVP disorders. These new cutoffs can improve WDT accuracy and should be validated in prospective studies. No serious adverse effects (SAEs) were observed. The lower limit of 3% weight loss was the criterion for early WDT interruption in nine conclusive tests. The demonstrated safety encourages us to use weight losses greater than 5% as stopping intervals and also tolerate a more extended test duration when necessary. In conclusion, WDT in children and adolescents is safe and still necessary until better diagnostic tools are proven helpful in pediatric patients with polyuria-polydipsia syndrome.NEW & NOTEWORTHY This large pediatric cohort study confirms the safety of the water deprivation test but reveals its limited diagnostic accuracy. Updated diagnostic criteria can improve performance and reduce unnecessary testing. A baseline urinary osmolality (uOsm) > 279 mOsm/kgH2O excluded arginine vasopressin (AVP) disorders with 100% sensitivity and 74% specificity and an interrupt uOsm > 533 mOsm/kgH2O excluded them with 100% sensitivity and specificity. A weight loss >5% was a safe threshold. Prospective, multicenter studies are needed to generalize these findings.