Impact of change in serum sodium concentration on mortality in patients hospitalized with heart failure and hyponatremia.

BACKGROUND Hyponatremia is a common electrolyte abnormality among patients hospitalized with heart failure and it is a marker for increased short-term and long-term mortality. However, little is known about the time course of hyponatremia and whether changes in serum sodium levels affect clinical outcomes. METHODS AND RESULTS Patients (n=322) hospitalized with decompensated heart failure and serum sodium <135 mmol/L were evaluated. After hospital discharge, the first sodium value obtained within a 60- to 270-day period was recorded, and patients were classified into 3 groups, based on whether the serum sodium value increased (≥2 mmol/L), decreased (≤2 mmol/L), or remained unchanged (±1 mmol/L) relative to the baseline value. Kaplan-Meier survival curves were constructed to illustrate mortality as a function of change in sodium concentration over time, and a Cox-proportional hazards model was constructed to determine if change in serum sodium concentration predicted mortality after adjusting for relevant covariates. The mean age of the population was 66 years, 45% were women, and 55% were white. The mean baseline sodium level was 131 mmol/L and the mean ejection fraction was 32.5%. Two hundred twenty-two patients (68.9%) exhibited an increase in sodium during follow-up; in 57 patients (17.7%) the level was unchanged and in 43 patients (13.4%) there was a decrease in sodium level. During a median follow-up of 610 days, there was a strong positive association between change in sodium level and survival (P for trend <0.001); that is, increased sodium was associated with decreased mortality. In multivariable analysis, change in sodium concentration and higher blood urea nitrogen were the strongest predictors of mortality (both P<0.0001). CONCLUSIONS Among patients hospitalized with heart failure and hyponatremia, change in serum sodium concentration over time is a strong predictor of long-term survival.