Estimate incidence and predictive factors of pediatric central diabetes insipidus in a single-institute study

Abstract

Objective

This study provided a rough estimate incidence of primary pediatric central diabetes insipidus (CDI) and examines the diagnostic factors between pediatric CDI and primary polydipsia (PP).

Methods

We collected 27 patients with chief complaints of polyuria and/or polydipsia from January 2014 to December 2018 in the Department of Pediatrics, Our University Hospital.

Results

We diagnosed type 1 diabetes mellitus (T1DM) in 16 patients, CDI in 5, PP in 5, and nocturnal enuresis in 1. The rough estimate incidence rate of pediatric CDI was 0.71/100,000/year. The diagnostic factors were a body mass index (BMI), urine gravity in the morning, urine volume and intake volume over 24 h, and bright spots in the posterior pituitary in a magnetic resonance image (MRI). The cutoff value of urine gravity in the morning for CDI was <1.010, with a sensitivity of 100 % and specificity of 100 %. The cutoff value of urine volume over 24 h for CDI was >2299 mL/m², with a sensitivity of 100 % and specificity of 85.7 %. No pediatric CDI patients had the bright spot in the posterior pituitary of their MRI, using a type 1–weighted image; however, only 1 out of 4 PP patients did not show the bright spot.

Conclusion

The rough estimate incidence of pediatric CDI with polydipsia and polyuria under the limited condition was 0.71/100,000/year, which was very low. Distinguishable factors between CDI patients and PP patients were BMI, urine and intake volumes over 24 h, and a bright spot on an MRI. Further studies with multiple institutes and more patients are required to confirm these findings.