Urinary and serum chloride phenotypes and readmission risk in acute heart failure

Background

Emerging evidence suggests that low serum chloride (sCl⁻) is independently associated with worse prognosis, impaired natriuretic response, and poorer decongestion outcomes in heart failure. More recently, urinary chloride (uCl⁻) has emerged as a potential marker of renal and systemic volume regulation; however, its prognostic significance remains unclear.

Purpose

Evaluate the association between uCl⁻ levels and long-term outcomes in acute heart failure (AHF), stratified by sCl⁻ levels.

Methods

We conducted a retrospective cohort study including 418 patients admitted for AHF. Clinical, laboratory, and point-of-care ultrasound data were collected at admission. Patients were stratified using the cohort median values of uCl⁻ (59 mmol/L) and sCl⁻ (100 mmol/L) and classified into four groups according to combined urinary and plasma chloride status. Patients were followed for up to 1 year, with censoring at death or end of follow-up. The primary endpoint was 1-year heart failure readmission. Survival was assessed using Kaplan–Meier analysis and compared with the log-rank test. The association between chloride handling and clinical outcomes was evaluated using Cox proportional hazards regression models, in which uCl⁻ and sCl⁻ were analyzed both as continuous variables and as categorical variables (four-group phenotype), adjusting for relevant clinical and biochemical covariates

Results

Median age was 87 years (IQR 82–90), and 67% were women. The low uCl/low sCl group showed the highest congestion burden, lowest urinary sodium (uNA), and the greatest 1-year readmission risk. Chloride phenotypes were significantly associated with clinical status and outcomes, with combined low uCl/low sCl identifying the most vulnerable subgroup.

During follow-up, 156 patients (37.3%) were readmitted. Kaplan–Meier plots showed a higher readmission risk in patients with low uCl⁻/low sCl⁻, with no clear differences among the remaining categories (log-rank p=0.029). (Figure1). In multivariable Cox regression, uCl⁻ analyzed as a continuous variable, but not uNa+ (p=0.204), was independently associated with readmission risk (HR 0.99 per mmol/L; 95% CI 0.98–0.99; p=0.026).(Figure2) When the four-level chloride phenotype was included in the model (reference: low uCl⁻/low sCl⁻), the other categories showed progressively lower 1-year readmission risk (p=0.019).

Conclusion

In patients hospitalized for AHF, uCl⁻ is an independent predictor of 1-year readmission, outperforming uNA and adding complementary information to sCl⁻. A combined low uCl⁻/low sCl⁻ profile identifies a high-risk congestion phenotype with significantly worse outcomes. These findings highlight the potential value of incorporating chloride-based markers into the routine assessment of decongestion and risk stratification in acute heart failure.For image description, please refer to the figure legend and surrounding text.For image description, please refer to the figure legend and surrounding text.