Mild hyponatremia carries a poor prognosis in community subjects

Abstract

Objective: Hyponatremia has been shown to predict adverse outcome in congestive heart failure and pneumonia among other common clinical entities, but its significance in the general population is elusive.

Methods: The population-based Copenhagen Holter Study included 671 men and women aged 55 to 75 years with no history of cardiovascular disease, stroke, or cancer. Baseline evaluation included 48-hour ambulatory electrocardiogram monitoring, blood tests, and a questionnaire. Hyponatremia was defined as s-Na < or = 134 mEq/L or s-Na < or = 137 mEq/L according to previously accepted definitions. An adverse outcome was defined as deaths or myocardial infarction. Median follow-up was 6.3 years.

Results: Fourteen subjects (2.1%, group A) had s-Na < or = 134 mEq/L, and 62 subjects (9.2%, group B) had s-Na < or = 137 mEq/L. No subject had s-Na < 129 mEq/L. An adverse outcome occurred in 43% of group A, 27% of group B, and 14% of subjects with s-Na >137 mEq/L (controls) (P < .002). Adjusted hazard ratio for adverse outcome was 3.56 (95% confidence interval [CI], 1.53-8.28, P < .005) in group A compared with controls and 2.21 (95% CI, 1.29-3.80, P < .005) in group B after controlling for age, gender, smoking, diabetes, low-density lipoprotein cholesterol, and blood pressure. The hazard ratios were robust for additional adjusting for variables showing univariate association to hyponatremia (ie, beta-blocker and diuretic use, heart rate variability, creatinine, C-reactive protein, and NT-pro brain natriuretic peptide). By excluding diuretic users (18% of subjects), the adjusted hazard ratio for adverse outcome was 8.00 (95% CI, 3.04-21.0, P < .0001) in group A and 3.17 (95% CI, 1.76-5.72, P = .0001) in group B compared with controls.

Conclusion: Hyponatremia is an independent predictor of deaths and myocardial infarction in middle-aged and elderly community subjects.