Abstract
Malnutrition is one of the common complications of heart failure (HF). In recent years, malnutrition has been proven to be associated with an increased risk of cardiovascular death, therefore, assessing HF patients’ nutritional status is essential. To date, several nutritional indicators have been proven to have a predictive value in patients with HF and atrial fibrillation (AF), therefore, in this study, we aimed to investigate Controlling Nutritional Status (CONUT) score, Prognostic Nutritional Index (PNI), body mass index (BMI), and albumin in patients with HF and AF (HF–AF). Data from 570 consecutive patients diagnosed with HF–AF between January 2018 to December 2018 were collected. The primary endpoint was all-cause mortality. The subgroup analysis was done by analyzing the impact of different types of AF (paroxysmal AF, persistent AF, and long-standing AF) on nutritional indexes and long-term mortality. During a median follow-up of 1194 days, we discovered that the non-survivor group tended to have a higher CONUT score, lower BMI, lower PNI, and lower albumin level (all P < .05). Multivariate analysis was performed to assess the prognostic ability of available nutritional indicators, and we found that CONUT score ≥ 5 [hazard ratio [HR]: 2.139; 95% confidence interval [CI]: 1.598–2.863, P < .001], age > 65 years old [HR: 2.798; 95% CI: 1.647–4.752, P < .001], and urea [HR: 1.035; 95% CI: 1.015–1.057, P < .001] may serve as an independent prognostic biomarker in HF–AF patients. Malnutrition was associated with higher mortality in patients with HF–AF. Among the available nutritional risk stratification tools, CONUT score was a better prognostic tool compared to PNI, albumin, and BMI.
Keywords: albumin, atrial fibrillation, BMI, CONUT, heart failure, malnutrition, PNI
1. Introduction
Body mass index (BMI) owing to its strong correlation with body fat has been widely accepted as a tool for indicating nutritional status in adults. The World Health Organization recommends BMI as the most practical population-level measure of nutritional status, defining values of < 18.5 kg/m² as undernutrition, >25 kg/m² as overweight, and > 30 kg/m² as obesity for both sexes and all adult age groups.[1] Despite its simplicity and widespread use, BMI has notable limitations. It does not account for factors such as age, sex, bone structure, fat distribution, muscle mass, or fluid retention (e.g., edema), which can lead to significant misclassification of nutritional status.[2–4]
Recently, there has been a growing appreciation and attention that nutritional status may play a significant role in cardiovascular diseases (CVD). For example, the “obesity paradox” phenomenon, where overweight and obese have lower cardiac death,[5,6] while malnutrition was associated with a greater risk for CV death.[7] Heart failure (HF), a leading cause of morbidity and mortality with annual mortality rates approaching 24%,[8] is frequently complicated by malnutrition, which further worsens prognosis. This highlights the importance of identifying and managing nutritional status in HF patients.
Although simple indicators such as BMI and serum albumin have been used for nutritional assessment in elderly and general populations,[9] their inconsistent predictive value underscores the need for more robust and comprehensive tools. Recently, nutritional risk scores such as the Prognostic Nutritional Index (PNI)[10] and Controlling Nutritional Status (CONUT) score[11] have been proposed as alternatives. Both PNI and CONUT have shown prognostic value in patients with HF[12,13] and atrial fibrillation (AF).[14] However, it remains unclear whether these associations also hold in patients with concomitant HF and AF (HF–AF), a subgroup characterized by complex pathophysiology. Therefore, this study aimed to investigate the prognostic impact of several malnutrition indicators in patients with HF–AF.
2. Methods
2.1. Study population
Data from 614 consecutive patients diagnosed with HF–AF between January 2018 and December 2018 were retrospectively collected. The inclusion criteria were adults (≥18 years) with symptomatic HF, AF documented on admission electrocardiography (ECG), a prior diagnosis of HF–AF, significantly elevated brain natriuretic peptide or N-terminal pro-brain natriuretic peptide levels on admission, and evidence of structural or functional cardiac abnormalities on echocardiography. Exclusion criteria included congenital heart disease, a history of cryoballoon or radiofrequency ablation, metabolic disorders, septic shock, hypovolemic shock, and anaphylactic shock. The study was approved by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University and conducted in accordance with the principles of the Declaration of Helsinki. The requirement for individual informed consent was formally waived by the Ethics Committee due to the retrospective design and the minimal risk posed to participants.
2.2. Outcome definitions
All-cause mortality was defined as death from any cause. HF was defined as a clinical syndrome consisting of cardinal symptoms (breathlessness, ankle swelling, and fatigue) that may be accompanied by signs (elevated jugular venous pressure, pulmonary crackles, and peripheral edema) due to a structural and/or functional abnormality of the heart that results in elevated intracardiac pressures and/or inadequate cardiac output at rest and/or during exercise.[15] Furthermore, HF has been divided into distinct phenotypes based on the measurement of left ventricular ejection fraction, they are: first, heart failure with reduced ejection fraction with EF ≤ 40%. Second, HF with mildly reduced ejection fraction with EF between 41% and 49%. Third, HF with preserved ejection fracti6on with EF ≥ 50%.[15] The New York Heart Association classification categorizes patients into 4 functional classes: Class I, asymptomatic; Class II, mild symptoms with ordinary physical activity; Class III, marked limitation with less than ordinary activity; and Class IV, symptoms present even at rest.[16]
AF was defined as a supraventricular tachyarrhythmia characterized by uncoordinated atrial electrical activation resulting in ineffective atrial contraction. ECG features of AF include irregular R–R intervals, absence of distinct repeating P waves, and irregular atrial activation.[17] AF can be further classified into 4 types: paroxysmal AF, which terminates spontaneously or with intervention within 7 days of onset; persistent AF, defined as continuous AF sustained beyond 7 days, including episodes requiring pharmacological or electrical cardioversion after ≥ 7 days; long-standing persistent AF, defined as continuous AF lasting > 12 months in patients for whom a rhythm control strategy has been selected; and permanent AF, in which both the patient and physician accept the presence of AF and no further attempts to restore or maintain sinus rhythm are planned.[17]
2.3. Data collection and management
Demographic data and admission vital signs including admission heart rate, systolic blood pressure, and diastolic blood pressure were collected. ECG examination and cardiac echocardiography were performed within 24 hours after admission. Blood sample was collected immediately after admission with an EDTA-anticoagulated tube and was assayed at a central laboratory. AF stroke risk stratification and bleeding scores (CHA2DS2-VASc and HAS-BLED scores) were assessed immediately once the patient was diagnosed with AF. Follow-ups were conducted by 3 cardiologists at the hospital, with study endpoints obtained through the electronic medical record system and/or telephone interviews. All clinical outcomes were defined according to standardized diagnostic criteria, and cases with uncertain diagnoses were jointly reviewed by at least 2 cardiologists to ensure consistency and minimize misclassification bias.
The CHA2DS2-VASc score is a stroke risk stratification tool for patients with AF. It includes the following components: congestive HF, hypertension, age ≥ 75 years, diabetes mellitus, prior stroke, vascular disease, age 65 to 74 years, and female sex. Two points are assigned for age ≥ 75 years and a history of stroke, transient ischemic attack, or thromboembolism. One point is assigned for age 65 to 74 years, diabetes mellitus, hypertension, HF or left ventricular systolic dysfunction, vascular disease (e.g., prior myocardial infarction, peripheral artery disease, or aortic plaque), and female sex. Stroke risk was categorized as low for a CHA2DS2-VASc score of 0, intermediate for a score of 1, and high for a score > 2. The HAS-BLED score is recommended for assessing major bleeding risk in AF patients receiving anticoagulation therapy. It includes hypertension, abnormal liver or renal function, prior stroke, history of bleeding, labile international normalized ratio, age > 65 years, and concomitant use of drugs (e.g., antiplatelets, NSAIDs) or alcohol, with each factor contributing 1 point. Bleeding risk was categorized as low (score = 0), intermediate (score 1–2), and high (score > 3).
2.4. Nutritional assessment and classifications
PNI was calculated as 10 × serum albumin (g/dL) + 0.005 × total lymphocyte count (per mm3). PNI was divided into 3 groups, with PNI < 35 at severe risk of malnutrition, PNI of 35 to 38 at moderate risk of malnutrition, and PNI ≥ 38 is normal.[10] CONUT score was determined by assessing 3 laboratory markers, they are serum albumin (g/dL), total cholesterol (TC) (mg/dL), and lymphocyte (/mL). CONUT score was divided into 2 groups, with a low score (0–4) indicating normal to mild risk of malnutrition, while a high score (5–12) indicating moderate to high risk of malnutrition.[11] BMI was classified into 4 categories, they are < 18.5 (undernutrition), 18.5 to 24.9 (normal weight), 25.0 to 29.9 (overweight), and ≥30.0 (obesity).[1] These data were assessed at hospital discharge.
2.5. Endpoint
The primary endpoint was all-cause mortality. The subgroup analysis was done by analyzing the impact of different types of AF (paroxysmal AF, persistent AF, and long-standing AF) on nutritional indexes and long-term all-cause mortality.
2.6. Statistical analysis
Data were expressed as median and interquartile boundary values or means ± standard deviation as appropriate. Comparison of variables in baseline characteristics was calculated with Pearson X2 test, or Mann–Whitney U test, as appropriate. Receiver operating characteristic curve analysis was performed to evaluate the predictive power of nutritional indexes alone, without including other clinical covariates. Kaplan–Meier curves were constructed by log-rank test. Univariate cox regression model was performed to assess the relationship between nutritional indexes and mortality. Multivariable linear regression analysis was conducted by stepwise forward method to evaluate the independence of nutritional indexes. Several covariates were included during the multivariable cox regression analysis, they are sex, age, vital signs (systolic blood pressure, BMI), laboratory parameters (urea, creatinine), comorbidities (coronary heart disease, primary hypertension, type 2 diabetes mellitus), and echocardiographic parameters (left atrium, right atrium, and ejection fraction). These covariates are well-established predictors of risk in patients with HF–AF.[18–20] The adjusted hazard ratio (HR) and their 95% confidence interval (CI) were calculated. Statistical significance was defined as a 2-sided P-value < .05. HR > 1.0 with a P-value < .05 indicated a deleterious association, while HR < 1.0 with a P-value < .05 indicated a protective association. Data were analyzed with SPSS version 25.0 (IBM, USA), GraphPad Prism 8.4.3 (GraphPad Software, Inc., San Diego), and MedCalc Statistical Software 19.2.6 (Ostend, Belgium).
3. Results
A total of 614 patients (mean age 73.36 ± 11.34 years, 46.9% male) with HF–AF were enrolled in this study. Among these, 44 patients were excluded due to loss of follow-up, and the remaining 570 patients were analyzed. Of all patients, 217 (35.3%) died during a median follow-up of 1194 (IQR 730–1332) days.
Table 1 presents the baseline demographics of survivors and non-survivors. Compared to survivors, non-survivors were older, had a higher proportion of patients with New York Heart Association functional classification class IV, a greater risk of bleeding as indicated by a higher percentage of patients with HAS-BLED > 3, and a lower BMI (all P < .001). Additionally, non-survivors exhibited higher CONUT scores (4.0 vs 3.0, P < .001), lower PNI values (39.07 vs 40.59, P < .001), lower BMI (22.93 vs 23.78, P < .001), and lower serum albumin levels (39.0 vs 40.0, P = .003) compared to survivors.
Table 1.
Baseline characteristics of study participants.
| Baseline characteristics | Survivors | Non-survivors | P-value |
|---|---|---|---|
| N = 353 | N = 217 | ||
| Demographics, n (%) | |||
| Age (yr) | 71.84 ± 11.29 | 76.77 ± 10.22 | <.001 |
| Male (%) | 157 (44.5) | 106 (48.8) | .309 |
| SBP (mm Hg) | 125.5 (112.0–143.75) | 127.0 (112.0–145.0) | .844 |
| DBP (mm Hg) | 77.0 (67.0–88.0) | 74.0 (66.0–84.0) | .505 |
| HR (bpm) | 81.0 (71.0–91.0) | 78.0 (69.0–88.0) | .523 |
| Nutritional indexes | |||
| BMI (kg/m2) | 23.78 ± 3.94 | 22.93 ± 4.69 | <.001 |
| CONUT | 3.0 (2.0–4.0) | 4.0 (3.0–5.0) | <.001 |
| PNI | 40.59 (37.2–43.6) | 39.07 (35.9–42.1) | <.001 |
| Albumin (g/L) | 40.0 (37.0–43.0) | 39.0 (35.0–42.0) | .003 |
| NYHA (%) | <.001 | ||
| ≤II | 157 (44.4) | 66 (30.4) | |
| III | 148 (41.9) | 87 (40.1) | |
| IV | 48 (13.7) | 64 (29.5) | |
| CHA2DS2-VASc (%) | .410 | ||
| ≤2 | 72 (20.4) | 31 (14.3) | |
| 3–5 | 217 (61.5) | 139 (64.1) | |
| ≥6 | 64 (18.1) | 47 (21.6) | |
| HAS-BLED (%) | <.001 | ||
| <3 | 282 (79.9) | 151 (69.6) | |
| ≥3 | 71 (20.1) | 66 (30.4) | |
| Medical history (%) | |||
| Primary hypertension | 205 (58.1) | 120 (55.3) | .516 |
| Type 2 diabetes mellitus | 99 (28.0) | 54 (24.9) | .408 |
| CHD | 163 (46.2) | 106 (48.8) | .535 |
| PCI | 41 (11.6) | 19 (8.8) | .280 |
| COPD | 45 (12.7) | 34 (15.7) | .327 |
| Stroke | 40 (11.3) | 37 (17.1) | .052 |
bpm = beats per minute, BMI = body mass index, CHD = coronary heart disease, CONUT = Controlling Nutritional Status, COPD = chronic obstructive pulmonary disease, DBP = diastolic blood pressure, NYHA = New York Heart Association Classification, PCI = percutaneous coronary intervention, PNI = Prognostic Nutritional Index, SBP = systolic blood pressure.
Table 2 compares laboratory parameters, echocardiographic indices, and medication use between the 2 groups. Non-survivors group exhibited higher levels of urea, creatinine, and uric acid, but lower levels of lymphocytes, hemoglobin, TC, low-density lipoprotein (LDL), and apolipoprotein B (all P < .05). In terms of treatment, non-survivors were less likely to receive angiotensin-converting enzyme inhibitors/angiotensin receptor blockers, statins, beta-blockers, and warfarin, but showed a higher frequency of digoxin use (all P < .05). In Figure 1, we assessed the sensitivity and specificity of each of the nutritional indexes to predict long-term mortality. Based on our data, we found that PNI has the highest predictive value (AUC: 0.649; 95% CI: 0.602–0.696, P < .001), followed by CONUT (AUC: 0.647; 95% CI: 0.600–0.694 P < .001), albumin (AUC: 0.603; 95% CI: 0.555–0.651, P < .001), and BMI (AUC: 0.567, 95% CI: 0.517–0.617, P = .001). Furthermore, in Figure 2, we analyzed the Kaplan–Meier curves for each of the nutritional indexes, and we found that BMI < 18.5, PNI < 35, albumin < 40 g/L, and CONUT score ≥ 5 presented with a higher risk of death (all log-rank P < .05).
Table 2.
Comparison of laboratory parameters, echocardiographic findings, and medication use.
| Survivors | Non-survivors | P-value | |
|---|---|---|---|
| N = 353 | N = 217 | ||
| BNP (pg/mL) | 409.0 (188.5–823.0) | 497.5 (242.0–1150.0) | .115 |
| D-dimer (mg/L) | 450.0 (134.5–1140.0) | 680.5 (317.5–1342.5) | .352 |
| Leukocyte (×103/µL) | 6.48 (5.33–8.04) | 6.63 (4.97–8.60) | .470 |
| Lymphocyte (×103/µL) | 12.5 (9.0–16.3) | 9.7 (6.1–13.4) | <.001 |
| Hemoglobin (g/L) | 131.0 (118.0–142.0) | 124.0 (106.5–137.0) | <.001 |
| Platelet (×103/µL) | 167.0 (118.0–142.0) | 165.0 (126.0–213.0) | .835 |
| APTT (s) | 30.5 (26.8–36.6) | 33.25 (28.3–40.2) | .761 |
| INR | 1.14 (1.04–1.43) | 1.18 (1.06–1.38) | .239 |
| Urea (mmol/L) | 7.0 (5.6–8.8) | 8.1 (5.93–12.68) | <.001 |
| Creatinine (µmol/L) | 82.0 (69.0–101.0) | 89.5 (70.0–121.8) | <.001 |
| Uric acid (µmol/L) | 379.5 (316.25–466.75) | 433.0 (338.25–550.75) | <.001 |
| TC (mg/dL) | 136.18 ± 46.74 | 125.64 ± 44.89 | .008 |
| HDL-C (mmol/L) | 1.13 ± 0.38 | 1.07 ± 0.43 | .173 |
| LDL-C (mmol/L) | 2.14 ± 0.87 | 1.92 ± 0.83 | .003 |
| Apo A (g/L) | 1.21 (1.04–1.40) | 1.08 (0.88–1.32) | .221 |
| Apo B (g/L) | 0.73 (0.58–0.90) | 0.69 (0.54–0.88) | .008 |
| Lp (a) (mg/L) | 74.0 (33.25–189.75) | 90.0 (39.5–206.5) | .565 |
| hsCRP (mg/L) | 2.69 (0.9–7.38) | 5.70 (1.77–15.88) | .002 |
| HbA1C (%) | 6.10 (5.7–6.6) | 6.0 (5.7–6.4) | .667 |
| Echocardiography | |||
| Right atrium (mm) | 39.53 ± 14.85 | 41.85 ± 15.67 | .076 |
| Left atrium (mm) | 38.39 ± 14.51 | 39.47 ± 13.78 | .380 |
| Left ventricle (mm) | 45.66 ± 17.06 | 47.95 ± 17.8 | .127 |
| Ejection fraction (%) | 48.2 ± 19.47 | 46.73 ± 19.29 | .380 |
| Medications (%) | |||
| Warfarin | 166 (47.0) | 79 (36.4) | .013 |
| NOAC | 127 (36.0) | 68 (31.3) | .258 |
| ACEI/ARB | 182 (51.6) | 82 (37.8) | .001 |
| Statin | 191 (54.1) | 91 (41.9) | .005 |
| Beta-blockers | 248 (70.3) | 130 (59.9) | .011 |
| Non-DHPs CCB | 9 (2.5) | 7 (3.2) | .636 |
| Digoxin | 9 (2.5) | 13 (6.0) | .038 |
| Aspirin | 55 (15.6) | 34 (15.7) | .978 |
| Clopidogrel | 67 (19.0) | 51 (23.5) | .196 |
| Ticagrelor | 1 (0.3) | 0 | .433 |
ACEI/ARB = angiotensin-converting enzyme inhibitor/ angiotensin receptor blocker, APTT = activated partial thromboplastin time, BNP = brain natriuretic peptide, HDL-C = high-density lipoprotein-cholesterol, hsCRP = high sensitivity C-reactive protein, INR = international normalized ratio, LDL-C = low-density lipoprotein-cholesterol, non-DHPs CCB = non-dihydropyridine calcium channel blocker, TC = total cholesterol.
Figure 1.
ROC analysis of each nutritional indexes on all-cause mortality. ROC = receiver operating characteristic.
Figure 2.
K–M curves of each nutritional indexes on all-cause mortality. K–M curves = Kaplan–Meier curve.
Cox regression analysis was performed to evaluate the association between nutritional risk indexes and long-term all-cause mortality. In the univariate analysis, CONUT score ≥ 5 [HR: 2.244; 95% CI: 1.680–2.998, P < .001), albumin < 40 g/L [HR: 1.730; 95% CI: 1.279–2.339, P < .001], and PNI < 35 [HR: 1.608; 95% CI: 1.131–2.286, P = .008] were associated with higher all-cause mortality. However, in the multivariate analysis, only a CONUT score ≥ 5 [HR: 2.139; 95% CI: 1.598–2.863, P < .001] remained independently associated with the primary outcome, along with age > 65 years [HR: 2.798; 95% CI: 1.647–4.752, P < .001] and urea level [HR: 1.035; 95% CI: 1.015–1.057, P < .001] (Table 3).
Table 3.
Univariate and multivariate Cox regression analyses of the primary endpoint.
| Univariate (HR, 95% CI) |
P-value | Multivariate (HR, 95% CI) |
P-value | |
|---|---|---|---|---|
| CONUT ≥ 5* | 2.244 (1.680–2.998) | <.001 | 2.139 (1.598–2.863) | <.001 |
| PNI < 35† | 1.608 (1.131–2.286) | .008 | ||
| Albumin < 40 g/L‡ | 1.730 (1.279–2.339) | <.001 | ||
| Age > 65 yr old§ | 3.250 (1.918–5.508) | <.001 | 2.798 (1.647–4.752) | <.001 |
| Male | 1.158 (0.872–1.537) | .312 | ||
| Body mass index (kg/m2) | 0.967 (0.932–1.004) | .076 | ||
| Systolic blood pressure (mm Hg) | 1.003 (0.997–1.009) | .329 | ||
| Coronary heart disease | 1.171 (0.882–1.555) | .276 | ||
| Primary hypertension | 0.974 (0.731–1.298) | .857 | ||
| Type 2 diabetes mellitus | 0.948 (0.688–1.306) | .743 | ||
| Urea (mmol/L) | 1.046 (1.025–1.067) | <.001 | 1.035 (1.015–1.057) | <.001 |
| Creatinine (µmol/L) | 1.002 (1.001–1.003) | <.001 | ||
| Right atrium (mm) | 1.009 (0.999–1.020) | .079 | ||
| Left atrium (mm) | 1.005 (0.994–1.015) | .393 | ||
| LVEF (%) | 0.997 (0.990–1.004) | .424 |
CI = confidence interval, CONUT = Controlling Nutritional Status, HR = hazard ratio, LDL-C = low-density lipoprotein-cholesterol, LVEF = left ventricular ejection fraction, PNI = Prognostic Nutritional Index.
CONUT score < 5 used as reference.
PNI ≥ 35 used as reference.
Albumin ≥ 40 g/L used as reference.
Age ≤ 65 years old used as reference.
Subgroup analysis (Table 4) was performed to evaluate the impact of AF types on nutritional indexes and long-term all-cause mortality. In the univariate regression analysis, BMI was not significantly associated with mortality across all AF types (P > .05). In contrast, PNI < 35 was associated with higher mortality in patients with persistent AF [HR: 2.209; 95% CI: 1.012–4.824, P = .047] and long-standing AF [HR: 2.853; 95% CI: 1.386–5.873, P = .004]. Similarly, albumin < 40 g/L was linked to increased mortality in patients with long-standing AF [HR: 2.771; 95% CI: 1.455–5.275; P = .002]. Notably, a CONUT score ≥ 5 was consistently associated with higher mortality across all AF types, including paroxysmal AF [HR: 1.773; 95% CI: 1.056–2.977; P = .030], persistent AF [HR: 2.129; 95% CI: 1.339–3.386; P = .001], and long-standing AF [HR: 3.085; 95% CI: 1.797–5.294; P < .001].
Table 4.
Subgroup analysis of atrial fibrillation types and their impact on long-term mortality.
| Parameters | Paroxysmal AF (HR, 95% CI) |
P-value | Persistent AF (HR, 95% CI) |
P-value | Long-standing AF (HR, 95% CI) |
P-value |
|---|---|---|---|---|---|---|
| BMI (kg/m2) | ||||||
| <18.5 | 1.308 (0.454–3.774) | .619 | 1.672 (0.560–4.992) | .357 | 2.428 (0.490–12.035) | .277 |
| 18.5–24.9 | 0.437 (0.182–1.047) | .063 | 1.370 (0.545–3.442) | .503 | 1.460 (0.352–6.051) | .602 |
| 25.0–29.9 | 0.662 (0.252–1.743) | .404 | 0.889 (0.323–2.446) | .820 | 0.647 (0.130–3.206) | .594 |
| ≥30 | 1 | – | 1 | – | 1 | – |
| PNI | ||||||
| <35 | 0.973 (0.352–2.690) | .959 | 2.209 (1.012–4.824) | .047 | 2.853 (1.386–5.873) | .004 |
| 35–38 | 1.290 (0.403–4.127) | .668 | 1.075 (0.391–2.952) | .888 | 2.142 (0.663–6.924) | .203 |
| >38 | 1 | – | 1 | – | 1 | – |
| Albumin (g/L) | ||||||
| <40 | 1.292 (0.776–2.153) | .325 | 1.593 (0.993–2.556) | .053 | 2.771 (1.455–5.275) | .002 |
| >40 | 1 | – | 1 | – | 1 | – |
| CONUT | ||||||
| 0–4 | 1 | – | 1 | – | 1 | – |
| ≥5 | 1.773 (1.056–2.977) | .030 | 2.129 (1.339–3.386) | .001 | 3.085 (1.797–5.294) | <.001 |
AF = atrial fibrillation, BMI = body mass index, CI = confidence interval, CONUT = Controlling Nutritional Status, HR = hazard ratio, PNI = Prognostic Nutritional Index.
4. Discussion
The findings of the present study can be summarized as follows: first, a high risk of malnutrition was associated with increased mortality in patients with HF–AF. Second, among the nutritional indexes evaluated, low BMI, elevated CONUT scores, low PNI scores, and reduced albumin levels were all associated with higher mortality; however, after multivariate adjustment, only a CONUT score ≥ 5 remained an independent prognostic biomarker in patients with HF–AF. Third, subgroup analysis revealed that different AF types were associated with varying mortality risks, with the CONUT score being the only nutritional index consistently identified as an independent predictor.
Malnutrition is a common complication in HF, with a reported prevalence of 46%.[21] To date, there is no clear explanation of how HF is associated with the loss of muscles, fat, and bone mass. Several studies suggest that multiple factors contribute to this wasting process, including activation of neurohormonal and proinflammatory cytokines,[22] imbalances between anabolic and catabolic pathways,[23] dietary deficiencies,[24] gut malabsorption,[25] and decreased bowel perfusion.[26] Moreover, as HF progresses, patients frequently experience depression, anxiety, and other adverse psychological states that may further exacerbate wasting. Malnutrition is a significant concern because it has been linked to increased mortality in acute coronary syndromes,[27] AF,[14] and HF,[21] highlighting the importance of nutritional risk assessment and early intervention. While surrogate markers such as serum albumin, lymphocyte counts, and low BMI have been shown to reflect nutritional status and predict poor prognosis,[28,29] their limited sensitivity and specificity underscore the need for more sophisticated risk stratification tools.
BMI owing to its simplicity is widely used for defining body size and indicating nutritional status, to the extent of assessing the risk of mortality.[30] However, its predictive ability is inconsistent and prone to errors and bias. For example, Schneider et al compared waist circumference, BMI, waist-to-height ratio, and waist-to-hip ratio in predicting cardiovascular, and all-cause mortality among obese patients, finding that BMI had the lowest predictive value.[31] Moreover, BMI fails to distinguish between components of body composition, fat distribution, fluid retention, and absolute weight gain.[32] Although BMI has been recognized as an independent prognostic factor in HF patients, our findings demonstrated that it had the lowest predictive value compared to albumin, PNI, and CONUT scores. This aligns with results from Yoshihisa et al,[33] who similarly reported that BMI performed worse than PNI, CONUT, and albumin in predicting mortality [AUC: 0.64; 95% CI: 0.58–0.70, P < .01].
Serum albumin has traditionally been used to quantify the number of plasma-circulating proteins, which is thought to reflect the nutritional status.[34] Moreover, significant loss of muscle mass is often observed in patients with hypoalbuminemia, further supporting the use of albumin as an indicator of malnutrition.[35] The association between hypoalbuminemia with increased all-cause mortality is well established.[36] Gibbs et al[37] demonstrated that serum albumin concentration is a strong predictor of mortality and morbidity in surgical patients. Similarly, a meta-analysis by El Iskandarani et al[38] found that hypoalbuminemia in HF patients is associated with significantly higher in-hospital and long-term mortality. Horwich et al[39] and Sardar et al[40] also showed that hypoalbuminemia is not only associated with increased mortality but also serves as an independent prognostic biomarker in HF patients. However, Yoshihisa et al[33] reported that albumin alone has only modest predictive value in HF patients, performing worse than PNI and CONUT, which is consistent with our findings.
In recent years, growing evidence linking malnutrition to higher all-cause mortality has drawn increased attention to the role of nutritional status as a potential biomarker in CVD. The most commonly used nutritional risk stratification tools were PNI and CONUT scores. The CONUT score was originally developed as an objective screening tool to identify undernutrition in hospitalized patients.[11] Subsequent studies demonstrated its utility in assessing protein reserves, caloric depletion, and immune function in patients with chronic heart failure.[41] In contrast, the PNI was initially designed to estimate operative risk in gastrointestinal surgical patients[42] and has since been recognized for its association with clinical outcomes in patients with HF.[43]
Several studies have demonstrated the use of CONUT and PNI scores in patients with HF, with most of the studies coming to the same conclusion, in which high CONUT score (≥5) and low PNI (<35) score were independently associated with increased risk of mortality.[12,13,33] In line with the previous findings, we also found that high CONUT (≥5) and low PNI (<35) were linked to higher mortality; however, after multivariate adjustment, only a high CONUT score (≥5) remained an independent prognostic biomarker in patients with HF–AF. Since TC level is a component of the CONUT score and 49.5% of our patients were receiving statin therapy, we conducted a subgroup analysis excluding statin users to avoid potential overestimation of CONUT scores. This analysis confirmed that CONUT score ≥ 5 remained significantly associated with higher mortality [HR: 2.556; 95% CI 1.74–3.739; P < .001].
It is noteworthy that although PNI demonstrated the highest predictive ability among the nutritional indexes in our study, only the CONUT score remained an independent prognostic biomarker after multivariate analysis in patients with HF–AF. This finding contrasts with the study by Yoshihisa et al,[33] which reported that PNI was superior to CONUT, albumin, and BMI. Several factors may explain this discrepancy. First, our study population consisted primarily of patients with HF–AF, who may exhibit a more pronounced inflammatory response compared to those in Yoshihisa cohort. Second, we observed that patients with severe malnutrition (PNI < 35) in our study had lower serum albumin levels (2.4 vs 2.6 g/dL) and total lymphocyte counts (694.78 vs 884.2/mm³) than those reported by Yoshihisa et al.[33] Third, the CONUT score incorporates TC levels in its algorithm, and lower TC has been associated with both a higher risk of AF[44] and increased mortality in HF.[45] In our cohort, patients with moderate to severe malnutrition risk (CONUT ≥ 5) had significantly lower TC levels (109.0 vs 143.3 mg/dL) than those in Yoshihisa study. These differences suggest that the CONUT score may be a more suitable prognostic marker than other nutritional indices in patients with HF–AF.
In the present study, we found that the non-survivor group had higher levels of urea and creatinine but lower levels of lymphocytes, hemoglobin, apolipoprotein B, TC, and LDL. Several mechanisms may underlie these findings. For instance, HF is known to contribute to bone marrow dysfunction and renal impairment, which can affect hematopoiesis and metabolic homeostasis.[46,47] Additionally, apolipoprotein B, TC, and LDL have long been recognized as independent risk factors for CVD.[48] We also noted that the non-survivor group was less likely to receive angiotensin-converting enzyme inhibitors/angiotensin receptor blockers and beta-blockers during hospitalization. Although these therapies may potentially influence nutritional status in HF patients,[49,50] they remain essential and irreplaceable components of evidence-based HF management.
In the subgroup analysis, we found that different types of AF may influence the prognostic value of nutritional indexes. To date, the underlying mechanisms for this observation remain unclear; however, it may be related to the duration and progression of the disease. Notably, our findings showed that patients with long-standing AF had worse outcomes compared to those with paroxysmal or persistent AF.
Our present study has several clinical implications. First, malnutrition is a common complication in HF, and its association with increased mortality in both HF and AF patients is well established, particularly when it progresses to cardiac cachexia, which carries an even poorer prognosis. Second, although BMI is the most convenient tool for nutritional risk stratification, it has shown inconsistent results and limited predictive value in HF and AF patients. This underscores the need to emphasize alternative algorithms, such as the CONUT and PNI scores, which may provide more accurate risk assessment. Third, our findings demonstrate that malnutrition is clearly associated with higher mortality in patients with HF–AF, highlighting the potential value of interventions aimed at improving nutritional status to improve patient outcomes.
5. Limitations
The present study has some limitations. First, it is a single-center retrospective study, which introduces inherent selection bias and potential confounding factors that may affect the generalizability of the results. The study population reflects the demographic and clinical characteristics of patients from our hospital, which may limit applicability to other regions or healthcare settings with different patient profiles. Second, due to the relatively long follow-up period, we were unable to thoroughly analyze potential complications arising from disease progression or changes in medication over time, which may influence nutritional status and outcomes. Third, we only used admission BMI, PNI, and CONUT for the present study, while nutritional and inflammatory status can fluctuate rapidly in HF–AF patients. Serial follow-up assessments and additional nutritional markers could provide more comprehensive insight. Fourth, factors such as patients’ functional status (e.g., activities of daily living), marital status, income, and education level: which could also influence mortality risk were not included as covariates in this analysis and should be considered in future studies. Moreover, receiver operating characteristic curve analysis evaluated the predictive value of nutritional indexes alone without adjusting for other covariates, which may limit its ability to account for potential confounding effects. Lastly, as studies on nutritional risk stratification in HF–AF patients remain limited and our sample size was relatively small, further multicenter studies with larger cohorts are warranted to validate these findings.
6. Conclusions
Malnutrition was associated with higher mortality in patients with HF–AF. Among the available nutritional risk stratification tools, CONUT score was a better prognostic tool compared to PNI, albumin, and BMI.
Author contributions
Conceptualization: Xianya Chen, Huachao Hu, Kai Lu.
Data curation: Zhen Xia, Xianya Chen.
Formal analysis: Zhen Xia, Xianya Chen.
Supervision: Hua Xiao.
Validation: Siyuan Xie, Kai Lu, Hua Xiao.
Visualization: Siyuan Xie.
Writing – original draft: Siyuan Xie.
Abbreviations:
- AF
- atrial fibrillation
- BMI
- body mass index
- CI
- confidence interval
- CONUT
- Controlling Nutritional Status
- CVD
- cardiovascular disease
- ECG
- electrocardiography
- HF
- heart failure
- HF–AF
- heart failure and atrial fibrillation
- HR
- hazard ratio
- LDL
- low-density lipoprotein
- PNI
- Prognostic Nutritional Index
- TC
- total cholesterol
All methods were carried out in accordance with relevant guidelines and regulations. The study was approved by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University and conducted in accordance with the principles of the Declaration of Helsinki. The requirement for individual informed consent was formally waived by the Ethics Committee due to the retrospective design and the minimal risk posed to participants.
The authors have no conflicts of interest to disclose.
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
How to cite this article: Xie S, Xia Z, Chen X, Hu H, Lu K, Xiao H. Prognostic value of CONUT, PNI, albumin, and BMI in patients with heart failure and atrial fibrillation: A retrospective cohort study. Medicine 2025;104:40(e44439).
Contributor Information
Siyuan Xie, Email: 2297337537@qq.com.
Zhen Xia, Email: 873755751@qq.com.
Xianya Chen, Email: 455047610@qq.com.
Huachao Hu, Email: 1224631417@qq.com.
Kai Lu, Email: lukai2013@foxmail.com.
References
- [1].WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet. 2004;363:157–63. [DOI] [PubMed] [Google Scholar]
- [2].Rolland-Cachera MF, Cole TJ, Sempe M, Tichet J, Rossignol C, Charraud A. Body mass index variations: centiles from birth to 87 years. Eur J Clin Nutr. 1991;45:13–21. [PubMed] [Google Scholar]
- [3].Gallagher D, Visser M, Sepulveda D, Pierson RN, Harris T, Heymsfield SB. How useful is body mass index for comparison of body fatness across age, sex, and ethnic groups? Am J Epidemiol. 1996;143:228–39. [DOI] [PubMed] [Google Scholar]
- [4].National Academies of Sciences, Engineering, and Medicine; Health and Medicine Division; Food and Nutrition Board; Roundtable on Obesity Solutions; Callahan EA. editor. Translating Knowledge of Foundational Drivers of Obesity into Practice: Proceedings of a Workshop Series. National Academies Press (US); 2023:10. [PubMed] [Google Scholar]
- [5].Hansel B, Roussel R, Elbez Y, et al. ; Investigators RR. Cardiovascular risk in relation to body mass index and use of evidence-based preventive medications in patients with or at risk of atherothrombosis. Eur Heart J. 2015;36:2716–28. [DOI] [PubMed] [Google Scholar]
- [6].Chen Y, Koirala B, Ji M, et al. Obesity paradox of cardiovascular mortality in older adults in the United States: a cohort study using 1997–2018 national health interview survey data linked with the national death index. Int J Nurs Stud. 2024;155:104766. [DOI] [PubMed] [Google Scholar]
- [7].Yang Y, Gao F, Liu T, et al. Impact of malnutrition on in-hospital outcomes in patients with acute myocardial infarction and no standard modifiable cardiovascular risk factors. J Cardiol. 2025;S0914-5087:00115–7. [DOI] [PubMed] [Google Scholar]
- [8].Emmons-Bell S, Johnson C, Roth G. Prevalence, incidence and survival of heart failure: a systematic review. Heart. 2022;108:1351–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Lai KY, Wu TH, Liu CS, et al. Body mass index and albumin levels are prognostic factors for long-term survival in elders with limited performance status. Aging (Albany NY). 2020;12:1104–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Wei W, Wu X, Jin C, et al. Predictive significance of the prognostic nutritional index (PNI) in patients with severe COVID-19. J Immunol Res. 2021;2021:9917302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Ignacio de Ulibarri J, Gonzalez-Madrono A, de Villar NG, et al. CONUT: a tool for controlling nutritional status. First validation in a hospital population. Nutr Hosp. 2005;20:38–45. [PubMed] [Google Scholar]
- [12].Sze S, Pellicori P, Kazmi S, et al. Prevalence and prognostic significance of malnutrition using 3 scoring systems among outpatients with heart failure: a comparison with body mass index. JACC Heart Fail. 2018;6:476–86. [DOI] [PubMed] [Google Scholar]
- [13].Chien SC, Lo CI, Lin CF, et al. Malnutrition in acute heart failure with preserved ejection fraction: clinical correlates and prognostic implications. ESC Heart Fail. 2019;6:953–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Zhu S, Zhao H, Zheng M, Peng J. The impact of malnutrition on atrial fibrillation recurrence post ablation. Nutr Metab Cardiovasc Dis. 2021;31:834–40. [DOI] [PubMed] [Google Scholar]
- [15].McDonagh TA, Metra M, Adamo M, et al. ; ESC Scientific Document Group. 2021 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021;42:3599–726. [DOI] [PubMed] [Google Scholar]
- [16].Rohde LE, Zimerman A, Vaduganathan M, et al. Associations between New York Heart Association Classification, objective measures, and long-term prognosis in mild heart failure: a secondary analysis of the PARADIGM-HF trial. JAMA Cardiol. 2023;8:150–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Hindricks G, Potpara T, Dagres N, et al. ; ESC Scientific Document Group. 2020 ESC guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for cardio-thoracic surgery (EACTS): the task force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J. 2021;42:373–498. [DOI] [PubMed] [Google Scholar]
- [18].Schnabel RB, Sullivan LM, Levy D, et al. Development of a risk score for atrial fibrillation (Framingham Heart Study): a community-based cohort study. Lancet. 2009;373:739–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Sieweke JT, Hagemus J, Biber S, et al. Echocardiographic parameters to predict atrial fibrillation in clinical routine-the EAHsy-AF risk score. Front Cardiovasc Med. 2022;9:851474. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Lawson CA, Zaccardi F, Squire I, et al. Risk factors for heart failure: 20-year population-based trends by sex, socioeconomic status, and ethnicity. Circ Heart Fail. 2020;13:e006472. [DOI] [PubMed] [Google Scholar]
- [21].Lv S, Ru S. The prevalence of malnutrition and its effects on the all-cause mortality among patients with heart failure: a systematic review and meta-analysis. PLoS One. 2021;16:e0259300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Mann DL. Inflammatory mediators and the failing heart: past, present, and the foreseeable future. Circ Res. 2002;91:988–98. [DOI] [PubMed] [Google Scholar]
- [23].Inui A. Cancer anorexia-cachexia syndrome: are neuropeptides the key? Cancer Res. 1999;59:4493–501. [PubMed] [Google Scholar]
- [24].Mustafa I, Leverve X. Metabolic and nutritional disorders in cardiac cachexia. Nutrition. 2001;17:756–60. [DOI] [PubMed] [Google Scholar]
- [25].King D, Smith ML, Lye M. Gastro-intestinal protein loss in elderly patients with cardiac cachexia. Age Ageing. 1996;25:221–3. [DOI] [PubMed] [Google Scholar]
- [26].Krack A, Sharma R, Figulla HR, Anker SD. The importance of the gastrointestinal system in the pathogenesis of heart failure. Eur Heart J. 2005;26:2368–74. [DOI] [PubMed] [Google Scholar]
- [27].Raposeiras Roubin S, Abu Assi E, Cespon Fernandez M, et al. Prevalence and prognostic significance of malnutrition in patients with acute coronary syndrome. J Am Coll Cardiol. 2020;76:828–40. [DOI] [PubMed] [Google Scholar]
- [28].Lourenco P, Silva S, Frioes F, et al. Low prealbumin is strongly associated with adverse outcome in heart failure. Heart. 2014;100:1780–5. [DOI] [PubMed] [Google Scholar]
- [29].Seko Y, Kato T, Morimoto T, et al. Association between body mass index and prognosis of patients hospitalized with heart failure. Sci Rep. 2020;10:16663. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Nuttall FQ. Body mass index: obesity, BMI, and health: a critical review. Nutr Today. 2015;50:117–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [31].Schneider HJ, Friedrich N, Klotsche J, et al. The predictive value of different measures of obesity for incident cardiovascular events and mortality. J Clin Endocrinol Metab. 2010;95:1777–85. [DOI] [PubMed] [Google Scholar]
- [32].Neeland IJ, Poirier P, Despres JP. Cardiovascular and metabolic heterogeneity of obesity: clinical challenges and implications for management. Circulation. 2018;137:1391–406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Yoshihisa A, Kanno Y, Watanabe S, et al. Impact of nutritional indices on mortality in patients with heart failure. Open Heart. 2018;5:e000730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Cabrerizo S, Cuadras D, Gomez-Busto F, Artaza-Artabe I, Marin-Ciancas F, Malafarina V. Serum albumin and health in older people: review and meta analysis. Maturitas. 2015;81:17–27. [DOI] [PubMed] [Google Scholar]
- [35].Visser M, Kritchevsky SB, Newman AB, et al. Lower serum albumin concentration and change in muscle mass: the health, aging and body composition study. Am J Clin Nutr. 2005;82:531–7. [DOI] [PubMed] [Google Scholar]
- [36].Biancucci M, Barbiero R, Pennella B, et al. Hypoalbuminaemia and heart failure: a practical review of current evidence. Eur J Heart Fail. 2025;27:293–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [37].Gibbs J, Cull W, Henderson W, Daley J, Hur K, Khuri SF. Preoperative serum albumin level as a predictor of operative mortality and morbidity: results from the national VA surgical risk study. Arch Surg. 1999;134:36–42. [DOI] [PubMed] [Google Scholar]
- [38].El Iskandarani M, El Kurdi B, Murtaza G, Paul TK, Refaat MM. Prognostic role of albumin level in heart failure: a systematic review and meta-analysis. Medicine (Baltim). 2021;100:e24785. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Horwich TB, Kalantar-Zadeh K, MacLellan RW, Fonarow GC. Albumin levels predict survival in patients with systolic heart failure. Am Heart J. 2008;155:883–9. [DOI] [PubMed] [Google Scholar]
- [40].Sardar MR, Rab Z, Salamon J, Saeed W, Zolty R. Hypoalbuminemia is an independent predictor of long-term mortality in patients with heart failure with preserved ejection fraction (HFpEF). J Card Fail. 2011;17:S82–3. [Google Scholar]
- [41].Nakagomi A, Kohashi K, Morisawa T, et al. Nutritional status is associated with inflammation and predicts a poor outcome in patients with chronic heart failure. J Atheroscler Thromb. 2016;23:713–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Buzby GP, Mullen JL, Matthews DC, Hobbs CL, Rosato EF. Prognostic nutritional index in gastrointestinal surgery. Am J Surg. 1980;139:160–7. [DOI] [PubMed] [Google Scholar]
- [43].Cheng YL, Sung SH, Cheng HM, et al. Prognostic nutritional index and the risk of mortality in patients with acute heart failure. J Am Heart Assoc. 2017;6:e004876. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].Lee HJ, Lee SR, Choi EK, Han KD, Oh S. Low lipid levels and high variability are associated with the risk of new-onset atrial fibrillation. J Am Heart Assoc. 2019;8:e012771. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [45].Rauchhaus M, Clark AL, Doehner W, et al. The relationship between cholesterol and survival in patients with chronic heart failure. J Am Coll Cardiol. 2003;42:1933–40. [DOI] [PubMed] [Google Scholar]
- [46].Westenbrink BD, Voors AA, de Boer RA, et al. Bone marrow dysfunction in chronic heart failure patients. Eur J Heart Fail. 2010;12:676–84. [DOI] [PubMed] [Google Scholar]
- [47].Szlagor M, Dybiec J, Młynarska E, Rysz J, Franczyk B. Chronic kidney disease as a comorbidity in heart failure. Int J Mol Sci . 2023;24:2988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [48].Simons LA, Simons J, Friedlander Y, McCallum J. Cholesterol and other lipids predict coronary heart disease and ischaemic stroke in the elderly, but only in those below 70 years. Atherosclerosis. 2001;159:201–8. [DOI] [PubMed] [Google Scholar]
- [49].Gammone MA, Efthymakis K, D’Orazio N. Effect of third-generation beta blockers on weight loss in a population of overweight-obese subjects in a controlled dietary regimen. J Nutr Metab. 2021;2021:5767306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [50].Mitchell CS, Premaratna SD, Bennett G, et al. Inhibition of the renin-angiotensin system reduces gene expression of inflammatory mediators in adipose tissue independent of energy balance. Front Endocrinol (Lausanne). 2021;12:682726. [DOI] [PMC free article] [PubMed] [Google Scholar]