Neutrophil-to-lymphocyte ratio (NLR) has been associated with poor outcomes in patients with acute coronary syndromes. However, its role for risk stratification in acute decompensated heart failure (ADHF) has not been well described. In this study, 1,212 consecutive patients admitted with ADHF who had total white blood cell and differential counts measured at admission were analyzed. The patients were divided into tertiles according to NLR. The association between NLR and white blood cell types with all-cause mortality was assessed using Cox regression analysis. During a median follow-up period of 26 months, a total of 284 patients (23.4%) had died, and a positive trend between death and NLR was observed; 32.8%, 23.2%, and 14.2% of deaths occurred in the higher, middle, and lower tertiles, respectively (p <0.001). After adjusting for confounding factors, multivariate analysis demonstrated that patients in the higher NLR tertile had the highest mortality (adjusted hazard ratio 2.23, 95% confidence interval (CI) 1.63 to 3.02, p <0.001), followed by those in the middle tertile (adjusted hazard ratio 1.62, 95% CI 1.16 to 2.23, p = 0.001). Furthermore, tertiles of NLR were superior in predicting long-term mortality compared with white blood cell, neutrophil, and relative lymphocyte counts. Patients in the higher NLR tertile (adjusted odds ratio 3.46, 95% CI 2.11 to 5.68, p <0.001) had a significantly higher 30-day readmission rate. In conclusion, higher NLR, an emerging marker of inflammation, is associated with an increased risk for long-term mortality in patients admitted with ADHF. NLR is a readily available inexpensive marker to aid in the risk stratification of patients with ADHF.
White blood cell (WBC) count and its subtypes are classic markers of inflammation in cardiovascular disease. Earlier studies have shown that increased WBC counts are associated with increased incidence of heart failure hospitalization and mortality. In addition, neutrophilia has been associated with increased incidence of acute decompensated heart failure (ADHF) in patients admitted with acute myocardial infarctions, while relative lymphocytopenia has shown to be an independent predictor of mortality in heart failure. Recently, neutrophil-to-lymphocyte ratio (NLR) has emerged as potent composite inflammatory marker, with higher levels associated with increased mortality in patients with acute coronary syndromes. The aims of this study were to evaluate the association between NLR and long-term outcomes in patients admitted with ADHF, independent of standard risk factors, and to compare its discriminative prognostic efficacy with total WBC count and neutrophil and relative lymphocyte counts.
Methods
The study population (n = 1,212) included consecutive patients admitted (from January 2006 to December 2008) with diagnoses of ADHF to the New England Heart Institute at Catholic Medical Center, a large urban hospital in Manchester, New Hampshire. The diagnosis of ADHF was based on standard guidelines. Patients with medical conditions known to affect the total and differential WBC counts, such as disorders of the hematopoietic system, history of cancer and/or previous treatment with chemotherapy, infection, and chronic inflammatory conditions; glucocorticoid therapy and/or histories of glucocorticoid use 3 months before the admission; and acute myocardial infarction or coronary revascularization within the past 6 months were excluded from the study. Also, patients who did not have differential WBC counts or who were lost to follow-up were excluded. Data were collected by the study personnel from review of the hospital’s medical records and included demographic characteristics, clinical signs and symptoms, laboratory data, medications, and echocardiographic parameters. Complete blood counts with automated differential counts, which included total WBCs, neutrophils, and lymphocytes, were obtained at the time of admission, and data were entered prospectively. NLR was calculated as the ratio of the neutrophil and lymphocytes, both obtained from the same automated blood sample at the admission of the study population. Death from any cause during a median follow-up period of 26 months (interquartile range 15 to 36) was the primary end point. Data pertaining to the death were obtained from the hospital’s medical records, the Social Security Death Index, patients’ family members, and patients’ physicians. The secondary end point included 30-day readmission related to ADHF. Additional analysis included the evaluation of the association between NLR tertile and long-term mortality in the patients groups with preserved (≥50%) and reduced (<50%) left ventricular ejection fractions (LVEFs) during their follow-up. The hospital’s institutional review board approved the study.
Patients were categorized on the basis of NLR into lower (tertile 1), middle (tertile 2), and higher (tertile 3) tertiles. Continuous variables are expressed as mean ± SD if normally distributed and as medians with interquartile ranges for those with skewed distributions. Categorical variables are expressed as frequencies and percentages. Continuous variables were compared using analysis of variance and Kruskal-Wallis tests for those with normal and skewed distributions, respectively. Chi-square tests were used to compare categorical variables. Univariate and multivariate Cox regression models were used to evaluate the independent association of the NLR tertiles with long-term mortality; unadjusted and adjusted hazard ratios (HRs) are reported with their respective 95% confidence intervals (95% CIs). Variables included in the model were age, gender, cardiac risk factors, the LVEF, medications, hemoglobin, blood urea nitrogen, creatinine, B-type natriuretic peptide, and atrial fibrillation. The discriminative abilities of NLR, neutrophil count, WBC count, and relative lymphocyte count in predicting long-term mortality were compared using area under the receiving-operating characteristic curve, and the final discriminative ability of the multivariate model was analyzed using Harrell’s C-statistics. To analyze the association of NLR tertile with long-term mortality on LVEF, the patient population was divided into 2 groups: those with preserved (≥50%) and reduced (<50%) LVEFs. Data were analyzed for independent association of NLR tertile on long-term mortality by univariate and multivariate Cox regression model. A logistic regression model was used to analyze the independent association of NLR tertile on short-term ADHF readmission risk; unadjusted and adjusted odds ratios are reported with their respective 95% CIs. Kaplan-Meier estimates of the cumulative hazard for long-term mortality according to NLR tertile were plotted for the entire study group and for the patients belonging to LVEF groups. A p value <0.05 was considered statistically significant. Statistical analyses were performed using SPSS (PASW) version 17.0 for Windows (SPSS, Inc., Chicago, Illinois).
Results
Of the total of 1,296 consecutive patients admitted to our institution with ADHF during the study duration, 84 patients were excluded because they did not meet the inclusion criteria, and thus 1,212 patients became eligible for the study. The baseline characteristics of the study population according to NLR tertile are listed in Table 1 . Membership in the higher NLR tertile was associated with older age, female gender, systemic hypertension, diabetes mellitus, history of coronary artery disease, and history of atrial fibrillation. Patients in the higher NLR tertile presented clinically with higher heart rates and increased incidence of pedal edema, rales on lung examination, jugular venous distension, lower LVEFs, and New York Heart Association class III or IV. Analysis of laboratory parameters demonstrated that membership in the higher NLR tertile was associated with increased serum B-type natriuretic peptide, blood urea nitrogen, serum creatinine, and hemoglobin levels. Patients in the higher NLR tertile were more likely to receive aspirin, statins, nitrates, digoxin, diuretics, and angiotensin-converting enzyme inhibitors and/or angiotensin receptor blockers, with no significant differences in β-blocker therapy. Examination of chest x-ray finding showed that membership in the higher NLR tertile was associated with a higher incidence of cardiomegaly, pleural effusion, and interstitial edema.
| Clinical Characteristic | NLR Tertile | |||
|---|---|---|---|---|
| Low (n = 408) | Medium (n = 396) | High (n = 408) | p Value | |
| Age (years) | 71 ± 15 | 75 ± 12 | 76 ± 13 | <0.001 |
| Women | 168 (41.2%) | 216 (53.0%) | 222 (56.1%) | <0.001 |
| Diabetes mellitus | 162 (40%) | 174 (43%) | 192 (48%) | 0.03 |
| Hypertension | 348 (85.0%) | 348 (87.8%) | 360 (88.2%) | 0.04 |
| Cigarette smoker | 294 (72.1%) | 264 (66.7%) | 288 (70.6%) | 0.23 |
| Hyperlipidemia ⁎ | 276 (67.6%) | 264 (66.7%) | 282 (69.1%) | 0.75 |
| Coronary artery disease | 234 (57.4%) | 266 (67.2%) | 282 (69.1%) | <0.001 |
| Previous myocardial infarction | 210 (51.5%) | 204 (51.5%) | 216 (52.9%) | 0.89 |
| Previous coronary artery bypass graft | 114 (27.9%) | 120 (30.3%) | 96 (23.5%) | 0.09 |
| Atrial fibrillation | 138 (33.8%) | 162 (40.9%) | 198 (48.5%) | <0.001 |
| LVEF (%) | 45 ± 19 | 43 ± 20 | 37 ± 21 | 0.001 |
| History of orthopnea | 252 (61.8%) | 246 (62.1%) | 234 (57.4%) | 0.30 |
| Paroxysmal nocturnal dyspnea | 330 (80.9%) | 324 (81.8%) | 336 (82.3%) | 0.86 |
| Systolic blood pressure (mm Hg) | 134 ± 33 | 132 ± 29 | 131 ± 22 | 0.19 |
| Diastolic blood pressure (mm Hg) | 74 ± 19 | 73 ± 17 | 72 ± 12 | 0.53 |
| Heart rate (beats/min) | 81 ± 27 | 82 ± 22 | 86 ± 19 | 0.003 |
| Jugular venous distension | 252 (61.8%) | 252 (63.6%) | 294 (72.1%) | 0.004 |
| Rales on lung examination | 360 (88.2%) | 361 (91.1%) | 372 (91.2%) | 0.01 |
| Pedal edema | 312 (76.5%) | 306 (77.3%) | 330 (80.9%) | 0.01 |
| S3 gallop | 198 (48.5%) | 222 (56.1%) | 234 (57.4%) | 0.01 |
| New York Heart Association heart failure class III/IV | 252 (61.8%) | 276 (69.7%) | 327 (80.1%) | <0.001 |
| Hemoglobin (g/dl) | 12.8 ± 2.1 | 12.3 ± 1.9 | 11.8 ± 2.1 | <0.001 |
| WBC count (×1,000/μl) | 7.6 ± 2.7 | 8.9 ± 3.6 | 11.3 ± 4.1 | <0.001 |
| NLR | 2.8 (2.2–3.8) | 5.1 (4.5–5.8) | 9.6 (7.6–13.1) | <0.001 |
| Blood urea nitrogen (mg/dl) | 30.1 ± 16.2 | 33.4 ± 18.4 | 40.4 ± 27.1 | <0.001 |
| Creatinine (mg/dl) | 1.6 ± 0.9 | 1.6 ± 0.6 | 2.0 ± 1.5 | <0.001 |
| B-type natriuretic peptide (pg/ml) | 713 (405–1,155) | 601 (318–1470) | 1,257 (527–1,340) | <0.001 |
| Albumin (g/dl) | 3.3 ± 0.4 | 3.3 ± 0.5 | 3.2 ± 0.5 | 0.03 |
| Cholesterol (mg/dl) | 144 ± 46 | 149 ± 36 | 142 ± 45 | 0.21 |
| Medications | ||||
| Aspirin | 192 (48.5%) | 216 (54.5%) | 270 (66.1%) | <0.001 |
| Angiotensin-converting enzyme inhibitors/angiotensin receptor blockers | 174 (43.9%) | 198 (50.0%) | 216 (52.9%) | 0.003 |
| Digoxin | 60 (14.7%) | 72 (18.2%) | 150 (36.8%) | <0.001 |
| β blockers | 288 (71.2%) | 306 (77.3%) | 306 (75.0%) | 0.08 |
| Diuretics | 264 (65.3%) | 276 (69.7%) | 300 (73.5%) | 0.002 |
| Nitrates | 42 (10.3%) | 90 (22.7%) | 90 (22.1%) | <0.001 |
| Statins | 204 (50.4%) | 210 (53.0%) | 252 (61.8%) | 0.003 |
| Chest x-ray | ||||
| Cardiomegaly | 354 (87.6%) | 364 (91.9%) | 384 (94.1%) | 0.01 |
| Pleural effusion | 144 (35.6%) | 168 (42.4%) | 192 (47.1%) | 0.001 |
| Cephalization | 348 (86.1%) | 360 (90.9%) | 378 (92.6%) | 0.002 |
| Interstitial edema | 180 (44.5%) | 210 (53.0%) | 258 (63.2%) | <0.001 |
| Cause of heart failure | ||||
| Primary | 195 (47.8%) | 190 (47.9%) | 188 (46.1%) | 0.61 |
| Ischemic | 188 (46.2%) | 168 (42.3%) | 199 (48.8%) | 0.45 |
| Valvular | 16 (4.1%) | 24 (6.1%) | 14 (3.4%) | 0.15 |
| Others | 8 (1.9%) | 14 (3.7%) | 7 (1.7%) | 0.16 |
⁎ Defined as previously diagnosed hyperlipidemia or receiving lipid-lowering medications.
During follow-up (median 26 months), a total of 284 patients (23.4%) died. Patients belonging to NLR tertile 3 had an increased rate of death (32.8% [n = 134]) compared to those in tertile 2 (23.2% [n = 92]) and tertile 1 (14.2% [n = 58]) (p <0.001). In univariate and multivariate analysis with NLR tertile 1 as the reference, patients in NLR tertile 3 (adjusted HR 2.23, 95% CI 1.63 to 3.02, p <0.001) showed a higher mortality rate compared to those in tertile 2 (adjusted HR 1.62, 95% CI 1.16 to 2.23, p = 0.004), as depicted in Figure 1 . The area under the curve for NLR was significantly greater (0.77) than for WBC count (0.66, p <0.001), neutrophil count (0.69, p = 0.001), and lymphocyte minimum count (0.58, p <0.001). A multivariate model demonstrated higher C-statistics including tertile of NLR (0.81) compared to models including tertiles of WBC count (0.78), neutrophil count (0.79), and lymphocyte count (0.65).
