CONUT Score as a Predictor of Mortality Risk in Acute and Chronic Heart Failure: A Meta-Analytic Review

Abstract

Heart failure (HF) is a major global health burden and a leading cause of morbidity and mortality. Nutritional status has emerged as an essential factor influencing outcomes in HF, with the Controlling Nutritional Status (CONUT) score gaining attention as a simple, objective marker derived from serum albumin, total cholesterol, and lymphocyte count. This meta-analysis evaluated the prognostic value of the CONUT score in predicting all-cause mortality in patients with acute and chronic heart failure. A systematic search was conducted in the PubMed, MEDLINE, Google Scholar, and Cochrane Library databases for the past ten years, using combinations of keywords such as "heart failure", "CONUT score", "malnutrition", and "mortality". Studies were included if they reported hazard ratios (HRs) for all-cause mortality in relation to CONUT score categories in adult HF populations. Eight eligible studies comprising 15,761 patients were included. Pooled analysis showed that higher CONUT scores were significantly associated with increased all-cause mortality (pooled HR = 1.47; 95% CI: 1.30-1.66). Despite substantial heterogeneity (I² = 80%), the direction of effect was consistent across studies. The CONUT score is a useful prognostic marker in acute and chronic heart failure patients. Further research should explore the effects of targeted nutritional interventions in high-risk HF patients identified by elevated CONUT scores and efforts to standardize malnutrition cut-offs in clinical practice.

Keywords: CONUT score; heart failure; malnutrition; mortality; nutritional screening; nutritional status; risk stratification.

Figure 1

PRISMA diagram [21]. PRISMA flowchart depicting the study selection process. In total, 243 records were identified through electronic databases (PubMed, MEDLINE, Cochrane Library, and Google Scholar). After removal of duplicates, 233 articles were screened, of which 200 were excluded based on title and abstract. Then, 33 articles were assessed for eligibility, with 25 excluded due to irrelevant outcomes or being review articles. Ultimately, eight studies were included in the final qualitative synthesis.

Figure 2

Forest Plot 1 showing pooled HR for all-cause mortality in HF patients with a high CONUT score [9,14,24,25,26,27,28,29]. Hazard ratios (HRs) indicating a statistically increased risk. Study heterogeneity was substantial (I² = 80%, p < 0.01).

Figure 3

Subgroup analysis: HF_a and HF_c. Hazard ratios with corresponding 95% confidence intervals for studies within the HF_a and HF_c subgroups, analyzed using a random-effects model [9,14,24,25,26,27,28,29]. The table also includes Z-values, p-values, and a forest plot illustrating effect sizes and heterogeneity across studies. Red squares represent the hazard ratio (HR) estimates for individual studies, with square size proportional to study weight. Horizontal red lines denote 95% confidence intervals (CI) for each study. The vertical dashed line indicates the line of no effect (HR = 1). Black diamonds represent the pooled HR with the width corresponding to the 95% CI.

Figure 4

Funnel Plot 2 for publication bias [9,14,24,25,26,27,28,29]. The funnel plot does not indicate a potential publication bias. Dashed lines indicate the 95% confidence limits around the summary effect, helping to assess symmetry, and the Red vertical line represents the null hypothesis (Hazard Ratio = 1), against which publication bias is visually assessed. The horizontal axis represents the hazard ratio (HR), and the vertical axis represents the standard error of the log HR.