American Association for Clinical Chemistry
Better health through laboratory medicine
August 2009 Clinical Laboratory News: Diagnostic Profiles

CLN Banner Logo

August 2009: Volume 35, Number 8

Hyponatremia in a Community Population Linked to Poor Prognosis

Hyponatremia is one of the most common electrolyte imbalances in hospitalized patients, and has been associated with poor prognosis in patients hospitalized with congestive heart failure, in survivors of ST-elevation myocardial infarction (STEMI) and non-STEMI, and in patients hospitalized with community-acquired pneumonia. However, the prognostic value of hyponatremia has not been studied extensively in community-based, mature adult populations. New research indicates that hyponatremia is an independent predictor of death and myocardial infarction in this population (Am J Med 2009;122:679–686).

The population-based study involved 671 subjects age 55 to 75 with no overt heart disease. However, more than half had two risk factors such as diabetes, smoking, or family history of cardiovascular disease; the others had one or no such risk factors. A battery of lab tests was performed, including serum sodium analysis.

The subjects were divided into three study groups based on serum sodium levels: one with levels ≤134 mEq/L (2.1% of subjects); the second with levels ≤137 mEq/L (9.2% of subjects); and the third with levels ≥138 mEq/L (90.8% of subjects). The commonly used cut point for hyponatremia is 134 mEq/L, but the investigators also used 138 mEq/L as a conservative definition based on new research. Composite endpoints studied were death and MI.

Hyponatremia persisted as an independent predictor of adverse outcomes through several iterations of Cox regression analysis, backward elimination models, and further exploratory analyses. The adjusted hazard ratio for the standard hyponatremia cut-off was 3.56. Subtle hyponatremia between 134 mEq/L and 137 mEq/L did not significantly predict the endpoints after controlling for conventional risk factors, but when diuretic users were excluded, the predictive value of subtle hyponatremia was significant, with a hazard ratio of 2.37.

The researchers considered numerous reasons for hyponatremia besides the use of diuretics in this community-based population, but determined that the mechanisms for the condition being independently associated with adverse outcomes “remain elusive.” The authors called for further research to understand the complicated dynamics behind and risks associated with hyponatremia.

Serial BNP Cited as Independent Predictor in ACS Shows

Recently published research indicates that serial B-type natriuretic peptide (BNP) levels in patients hospitalized for acute coronary syndrome (ACS) and at 7 weeks post-discharge predicts risk for mortality, readmission with ACS, and admission with congestive heart failure (CHF) after 10 months, and that it does so independently of traditional echocardiographic findings such as left ventricular hypertrophy (LVH), left ventricular systolic function, or left arterial diameter (Am Heart J 2009;158:133–40). The study builds on prior research demonstrating that BNP provides prognostic information for patients following hospitalization for decompensated heart failure, and for the spectrum of ACS. Likewise, BNP previously has been shown to have predictive power in ACS independent of left ventricular ejection fraction but not echocardiographic LVH, which is a common cardiovascular risk factor in patients with coronary artery disease.

The study involved 443 consecutively hospitalized patients with ACS who presented within 72 hours of ischemic discomfort, provided a clinical history, and had an electrocardiogram (ECG), bedside BNP analysis and echocardiography, along with other lab tests, including admission Hb, estimated glomerular filtration rate, and cardiac troponin T. Seven weeks after the ACS admission, participants had follow-up ambulatory ECG, routine lab tests including BNP, and transthoracic echocardiography. The researchers used a variety of mechanisms to determine all cause death and the cardiovascular events of readmission with ACS or admission with CHF at 10 months after the initial hospitalization.

Admission BNP levels ranged from 4 to 2,390 pg/mL with a mean of 272 pg/mL + 361 pg/mL. With a BNP cut-off of 80 pg/mL, higher admission and follow-up BNP levels significantly predicted both 10-month mortality and cardiovascular events. After multivariate analysis adjusting for significant clinical and echocardiographic predictors, a baseline BNP of >80 pg/mL was still associated with an almost 3-fold higher long-term risk of cardiovascular events (adjusted relative risk, 2.63; 95% CI=1.34-5.19). Elevated BNP levels remained an independent predictor of CV events even after adjusting for the presence of all echocardiographic abnormalities on multivariate analysis (relative risk, 4.12; 95% CI=1.58-10.72).

The authors speculate that continuous serial sampling of BNP might identify important cardiovascular events that would otherwise not be anticipated.