Author + information
- Justin L. Grodin, MD,
- Jeffrey M. Testani, MD, MTR and
- W.H. Wilson Tang, MD∗ ()
- ↵∗Department of Cardiovascular Medicine, Heart and Vascular Institute, Cleveland Clinic, 9500 Euclid Avenue, Desk J3-4, Cleveland, Ohio 44195
We appreciate the comments from Dr. Lim regarding our analysis investigating the prognostic role of serum chloride levels in acute decompensated heart failure (ADHF). As discussed, chloride levels can be affected by a diversity of mechanisms in ADHF. Indeed, hypochloremia and hyponatremia can develop by maladaptive neurohormonal responses in ADHF that limit free water excretion. However, we posit that association of hypochloremia and mortality is likely a result of a chloride’s broader homeostatic role in heart failure than sodium and even acid-base status. As shown in the baseline characteristics of our analysis, chloride levels were positively correlated with sodium levels and negatively correlated with serum bicarbonate levels (1). These findings highlight 2 of chloride’s roles: 1) that of an extracellular solute complementary to sodium; and 2) that of an acid buffer to cations. Because both factors can confound the risk relationship of chloride with mortality, they were a priori included in the multivariable model. Neither was significantly associated with mortality after adjustment for chloride levels, thus implying that that it is actually the chloride levels that play the larger prognostic role.
An approximation of the strong ion difference (SID) can be estimated as the difference between sodium and chloride concentrations (2). Indeed, the SID was also associated with mortality in our cohort (hazard ratio [HR]: 1.07; 95% confidence interval [CI]: 1.04 to 1.10; p < 0.001). However, although the SID remains significant when adjusting for sodium (HR: 1.08; 95% CI: 1.05 to 1.10; p < 0.001), the association disappeared when adjusting for chloride (HR: 1.01; 95% CI: 0.98 to 1.05; p = 0.39), whereas the association of chloride and mortality remains significant. Nevertheless, we caution the incorporation of sodium and chloride levels within the same statistical model as the SID due to collinearity. Taken together, although the framework of interpreting the SID is mechanistically insightful, chloride levels are simpler to interpret and seem to be the main prognostic role player.
Diuretic therapy plays an important pathophysiological role in the generation of hypochloremia in heart failure, especially because it may be a more sensitive marker of a depletional state (2). Although inclusion in our cohort required loop diuretic use, there were limited granular data on diuretic dosing during hospitalization in our dataset. Therefore, we were unable to draw any specific mechanistic conclusions regarding the contribution of diuretics to the generation of hypochloremia.
Please note: Dr. Testani is supported by National Institutes of Health (NIH) grants K23HL114868 and L30HL115790. Dr. Tang is supported by NIH grant R01HL103931. Dr. Grodin has reported that he has no relationships relevant to the contents of this paper to disclose.
- 2015 American College of Cardiology Foundation