Effect of renal function on the prognostic importance of chloride in patients with heart failure

Abstract

Hypochloremia has recently gained interest as a potential marker of outcomes in patients with heart failure (HF). The exact pathophysiologic mechanism linking hypochloremia to HF is unclear but is thought to be mediated by chloride-sensitive proteins and channels located in kidneys. This analysis aimed to understand whether renal dysfunction (RD) affects the association of hypochloremia with mortality in patients with HF. Using data from a nationwide registry, 438 cases with complete data on serum chloride concentration and 1-year survival were included in the analysis. Patients with an estimated glomerular filtration rate of <60 mL/min/m2 at baseline were accepted as having RD. Hypochloremia was defined as a chloride concentration <96 mEq/L at baseline. For HF patients without RD at baseline, patients with hypochloremia had a significantly higher 1-year all-cause mortality than those without hypochloremia (41.6% vs 13.0%, log-rank p < 0.001) and the association remained significant after multivariate adjustment (odds ratio (OR): 2.55, 95% confidence interval (CI): 1.25–5.21). The evidence supporting the association was very strong in this subgroup (Bayesian Factor (BF)10: 48.25, log OR: 1.56, 95% CI: 0.69–2.43). For patients with RD at baseline, there was no statistically significant difference for 1-year mortality for patients with or without hypochloremia (36.3% vs 29.7, log-rank p = 0.35) and there was no evidence to support an association between hypochloremia and mortality (BF10: 1.18, log OR :0.66, 95% CI: −0.02 to 1.35). In patients with HF, the association between low chloride concentration and mortality is limited to those without RD at baseline.

Prognostic value of hypochloremia on mortality in patients with heart failure: a systematic review and meta-analysis

“…Our study is consistent with and extends the results of previous epidemiological studies highlighting the independent association of lower serum chloride levels with an increased long-term risk of adverse clinical outcomes in heart failure, irrespective of ejection fraction. [5][6][7][8][9][10][24][25][26][27][28][29][30][31][32][33] In particular, LVEF spanned across all categories, and the prognostic role of chloride was not modified by LVEF in the meta-regression analysis.…”

“…The study with the largest population included 4762 patients 24 whereas the smallest study included 161 patients. 25 Among the included studies, seven reported hazard ratio per unit change of serum chloride for eight patient cohorts, [6][7][8][26][27][28][29] five studies reported hazard ratio comparing hypochloremia and normochloremia for eight cohorts of patients, 5,24,25,30,31 and three studies reported both types of hazard ratio assessments. 10,32,33 Six studies included patients with acute heart failure, 6,8,10,26,29,31 seven studies included patients with chronic hazard ratios 5,7,24,27,28,32,33 and the remaining two studies included a mixed population of patients with acute and chronic heart failure.…”

“…10,32,33 Six studies included patients with acute heart failure, 6,8,10,26,29,31 seven studies included patients with chronic hazard ratios 5,7,24,27,28,32,33 and the remaining two studies included a mixed population of patients with acute and chronic heart failure. 25,30 Hypochloremia was defined as a serum level less than 96 mmol/l in eight studies, 5,7,8,10,25,29,30,32 less than 98 mmol/l in two studies, 6,24 less than 99 mmol/l in one study, 33 less than 104 mmol/l in one study 31 and the cut-off was not reported in three studies. [26][27][28] Baseline characteristics of the included patient populations are shown in Table 2.…”

“…Based on NOS assessment, all 15 studies were classified as good quality (Supplementary Table 2, http://links.lww.com/JCM/A652). The study with the largest population included 4762 patients 24 whereas the smallest study included 161 patients 25 . Among the included studies, seven reported hazard ratio per unit change of serum chloride for eight patient cohorts, 6–8,26–29 five studies reported hazard ratio comparing hypochloremia and normochloremia for eight cohorts of patients, 5,24,25,30,31 and three studies reported both types of hazard ratio assessments 10,32,33 .…”

“…com/JCM/A652). In particular, as in the majority of studies adjusted hazard ratios were reported and used for the analysis, and only one study for each predictor (hypochloremia/serum chloride) reported unadjusted hazard ratios, 25,28 which may have overestimated the association with the primary endpoint; leave-one-sensitivity analyses were used to confirm the robustness of the main analysis.…”