Serum Chloride Level Predicts Mortality in Acute Heart Failure

By Van Selby, MD

Assistant Professor of Medicine, UCSF Cardiology Division, Advanced Heart Failure Section, San Francisco

Dr. Selby reports no financial relationships relevant to this field of study.

SOURCE: Grodin JL, et al. Prognostic role of serum chloride levels in acute decompensated heart failure. J Am Coll Cardiol 2015;66:659-666.

The association between serum sodium level and outcomes in acute decompensated heart failure (ADHF) is well-established. Serum chloride levels are also routinely obtained with basic chemistry panels, but the clinical significance of hypochloremia in ADHF has not been studied.

Grodin and colleagues reviewed data from 1318 patients hospitalized at the Cleveland Clinic with a discharge diagnosis of ADHF. They examined the association between the admission serum chloride level and all-cause mortality, and compared the prognostic significance of serum chloride and sodium levels. The relationship between serum chloride, sodium, and mortality was also evaluated in a validation cohort of 876 patients admitted to the Hospital of the University of Pennsylvania for heart failure (HF). Patients in both cohorts had predominantly systolic HF.

In univariate analyses, each unit increase in chloride was associated with a 6% decrease in mortality (hazard ratio [HR] per unit increase, 0.94; 95% confidence interval [CI], 0.92-0.95; P < 0.001). Admission serum sodium level was also associated with mortality (HR per unit increase, 0.95; 95% CI, 0.93-0.97; P < 0.001). In multivariate analyses adjusted for other clinical variables, the chloride level remained predictive of mortality (HR per unit increase, 0.93; P < 0.001), whereas serum sodium level was no longer independently associated with mortality (HR per unit increase, 1.03; P = 0.18). Of note, mortality risk was strongly associated with changes in serum chloride levels < 96 mEq/L, but did not vary significantly at values > 96.

The findings were similar in the cohort from Pennsylvania; serum chloride remained predictive in multivariate models adjusted for other clinical predictors (HR, 0.95; P = 0.01), while serum sodium was not (HR, 0.99; P = 0.58). The authors conclude that serum chloride levels measured during hospitalization for ADHF are independently and inversely associated with long-term mortality, independent of serum sodium levels.

COMMENTARY

The authors should be commended for challenging the long-held assumption that sodium is the key electrolyte in the pathophysiology of heart failure. Furthermore, in an era where the search for novel biomarkers often leads to increasingly complex and costly laboratory techniques, they chose to study a parameter that has been measured as part of standard care for decades. The pathophysiological role of chloride in heart failure is not well understood, mainly because it has not been studied in detail. Many of the same maladaptive responses that reduce serum sodium in HF also impact chloride levels. These include elevated levels of both arginine vasopressin and angiotensin II. Yet, these similarities alone cannot explain why chloride would be an even stronger prognostic marker than sodium. The authors suggest chloride plays a broader homeostatic role, serving as a buffer for cations and helping the kidney eliminate salt and water. This argument is further supported by the important role of chloride in a wide variety of other bodily functions, including maintenance of osmotic balance and muscle function. Regardless of the mechanism, the strong statistical association between chloride levels and mortality suggests further investigation may yield new insights that help shape our understanding of HF.

This study does have several limitations. The cohorts studied were obtained from two large, tertiary academic centers, and not necessarily representative of the broader ADHF population. The multivariate models did not include many well-known prognostic markers in HF, such as systolic blood pressure and natriuretic peptide levels. Furthermore, it must be emphasized that adding chloride levels to the multivariate predictive model had a very modest effect on the overall predictive ability. The C-statistic, a marker of discriminative ability, only increased from 0.68 to 0.69 with the addition of chloride level. The cohorts studied in this analysis were comprised of patients with primarily systolic HF, so we cannot conclude that chloride levels maintain their prognostic utility in those with preserved ejection fraction (unlike sodium, whose association with mortality has been proven in both preserved and reduced ejection fraction).

There is still a lot of work to be done on this topic. Future studies will need to validate the prognostic role of chloride in broader HF cohorts and elucidate the pathophysiological role of chloride in HF. Eventually, interventions that specifically target hypochloremia may improve the care of patients with ADHF, a condition for which we have yet to identify a therapy that meaningfully improves outcomes. Despite the limitations and further work needed, for now we have a new prognostic marker that is widely available and possibly more useful than serum sodium level for estimating prognosis in ADHF.