Mortality and causes of death in patients hospitalized with hyponatremia - a propensity matched cohort study.

Abstract

Chronic hyponatremia is common in hospitalized patients and associated with high mortality. To what extent this reflects a causal effect remains uncertain. This study was based on the Stockholm Sodium Cohort, a longitudinal laboratory data repository covering 1.6 million individuals from 2005 to 2018. Using 1:1 propensity score matching, we explored mortality rates and causes of death in patients with mild (130-134 mmol/L), moderate (125-129 mmol/L), profound (120-124 mmol/L) or very profound (<120 mmol/L) hyponatremia compared to patients with normal (135-145 mmol/L) sodium concentrations on admittance to medical wards. In total, 283 837 individuals fulfilled inclusion criteria, 79 407 of which had hyponatremia. Of these, 66 941 (52.7 % women) were successfully matched to counterparts with normal sodium concentrations. Thirty-day mortality rates were higher in patients with hyponatremia (HRs from 1.35 [95 % CI 1.28-1.42] in mild to 3.38 [95 % CI 2.16-5.28] in very profound hyponatremia). One-year mortality rates were marginally elevated in patients with mild hyponatremia (HR 1.04 [95 %CI 1.01-1.07]), but higher with more pronounced hyponatremia (HRs 1.18 [95 %CI 1.09-1.27] to 1.38 [95 %CI 1.11-1.69]). Excess 30-day mortality in mild, moderate, and profound hyponatremia was largely driven by malignant and gastrointestinal diseases. Excess mortality with hyponatremia is proportional to the sodium disturbance but attenuates over time. However, causes of death suggest that residual confounding from imbalanced severity of underlying diseases is the main cause of increased mortality.