Prognostic utility of the prognostic nutritional index combined with serum sodium level in patients with heart failure

Abstract

Background and aimsBoth malnutrition and hyponatremia (serum sodium <135 mmol/L) can be induced by the impaired absorption function of the edematous intestinal wall caused by heart failure (HF) and are prognostic factors of mortality in HF. However, little is known about the interrelationship of nutritional status and hyponatremia in mortality risk prediction in HF. Methods and resultsThis study enrolled 2882 HF patients admitted to the HF care unit of Fuwai Hospital, Beijing, China from 2008 to 2018; 71.3% were male and the mean age was 56.64 ± 15.96 years. Nutritional status was assessed by prognostic nutritional index (PNI), calculated as serum albumin (g/L) + 5 × total lymphocyte count (109/L). Lower PNI indicates worse nutritional status. Patients were divided into 8 groups based on baseline PNI quartiles (Q1: <43.6, Q2: 43.6–48.55, Q3: 48.55–63.25, Q4: >63.25) and sodium level (normal sodium and hyponatremia). After adjustment, patients in the PNI Q1 associated with hyponatremia had a 2.12-fold higher risk of all-cause death (95% confidence interval [CI]: 1.67–2.70) compared with those in the PNI Q4 with normal sodium. A refinement in risk prediction was observed after adding PNI quartile and serum sodium category to the original model (ΔC-statistic = 0.018, 95% CI: 0.007–0.025; net re-classification index = 0.459, 95% CI: 0.371–0.548; integrated discrimination improvement = 0.025, 95% CI: 0.018–0.032). ConclusionHF patients with both the lowest PNI quartile and hyponatremia are at higher risk of all-cause mortality. The combination of PNI and serum sodium level enhanced the predictive value for all-cause mortality in hospitalized HF patients. Clinical trial registrationURL: ClinicalTrials.gov; Unique Identifier: NCT02664818.