Hypernatremia in diabetic ketoacidosis

✍ Scribed by Monroe, K. W. ;King, W. D. ;Nichols, M. H. ;Atchison, J. A.

• Publisher: John Wiley and Sons
• Year: 1997
• Tongue: English
• Weight: 375 KB
• Volume: 14
• Category: Article
• ISSN: 1357-8170
• DOI: 10.1002/pdi.1960140606

✦ Synopsis

Abstract

The elevated blood glucose in DKA creates an osmotic gradient which typically results in hyponatremia. Many patients present with elevated serum sodium concentrations in DKA and very little is known about these patients. We questioned if these patients represent a ‘sicker’ population than those with normal or low serum sodium concentrations.

The objective of this study was to characterise those patients who develop elevated corrected serum sodium levels in diabetic ketoacidosis (DKA). A retrospective medical record review was done of all episodes of DKA at The Children's Hospital of Alabama between January 1992 and June 1994. All cases of DKA were in patients less than 17 years of age.

Among the 79 study entries which consisted of 52 separate individuals 46% female (n=36) 54% male (n=43), 11 patients were admitted for more than one episode of DKA, 27% (n=21) were patients with new onset diabetes and 73% (n=58) were patients with a prior history of diabetes mellitus. Of the corrected serum concentrations of sodium, 85% (n=67) were less than 149 mmol/L. In the patients with corrected serum sodium concentrations greater than 149 mmol/L (n=12), 35% (n=4) were African‐American and 65% (n=8) Caucasian, which was representative of the overall study population race distribution. Sixty seven per cent (n=8) of the hypernatremic patients were male, 33% (n=4) female.

Patients with new onset diabetes were more likely to present with elevated concentrations of sodium in serum than known diabetic patients (t=5.52, p=0.02); 33% (n=7) of the patients with new‐onset diabetes were hypernatremic. No association was found between corrected sodium and age (t=0.31, p=0.76), duration of symptoms (t=0.42, p=0.68) or serum urea nitrogen concentration which was used as a marker for level of hydration (t=1.48, p=0.14). Duration of symptoms was found to be longer among patients with new onset diabetes (Z~T~=4.31, p < 0.001) and patients presenting with elevated serum concentration of sodium (Z~T~=2.20, p=0.028).

We conclude that patients with new onset diabetes are more likely to present with elevated sodium values unrelated to age or level of hydration.