Reduced Early Sodium Variability Is Associated with Acute Brain Dysfunction in Critically Ill Patients

Abstract

Background

Dysnatremia is common in critically ill patients, but the neurological relevance of early sodium dynamics beyond admission sodium is unclear. We evaluated whether sodium variability during the first 72 hours of intensive care was associated with acute brain dysfunction (ABD).

Methods

This retrospective cohort study included 109 adults in a multidisciplinary intensive care unit. Sodium variability was expressed as the sodium variability index (SVI), the difference between maximum and minimum serum sodium during the first 72 hours. ABD was defined as delirium, seizures, and/or coma. The multivariable model included SVI, age, APACHE II score, mechanical ventilation, sedation, fluid balance, number of sodium measurements, and sodium-modifying therapy.

Results

ABD occurred in 71 of 109 patients (65.1%). Median SVI was lower with ABD, though the unadjusted difference was not significant (7 [4–10] vs. 5 [3–8.5] mmol/L; p = 0.063). In multivariable analysis (n = 108), lower SVI was independently associated with ABD (adjusted OR per 1-mmol/L, 0.870; 95% CI, 0.770–0.983; p = 0.025), consistent across models and sensitivity analyses. Higher APACHE II score, mechanical ventilation, and sedation were also independently associated with ABD; admission sodium and hyponatremia were not.

Conclusions

Lower early sodium variability was independently associated with ABD; individual components were not significant in exploratory analyses. Early sodium dynamics may add information beyond a single admission measurement and may mark neurological vulnerability. Given the retrospective design, findings are hypothesis-generating; prospective multicenter studies are needed.