2019 年 1 巻 2 号 p. 87-93
Background: The purpose of the study was to evaluate the impact of nutritional status on 1-year mortality in hospitalized patients with acute decompensated heart failure (ADHF).
Methods and Results: We enrolled 457 hospitalized ADHF patients. Previously established objective nutritional indexes (controlling nutritional status [CONUT], prognostic nutritional index [PNI], geriatric nutritional risk index [GNRI], and subjective global assessment [SGA]) were evaluated at hospital admission. Malnutrition was defined as CONUT score ≥5, PNI score <38, GNRI score <92, and SGA scores B and C. The frequencies of malnutrition based on CONUT, PNI, GNRI, and SGA were 31.5%, 21.4%, 44.9%, and 27.8%, respectively. All indexes were related to the occurrence of 1-year mortality on univariate Cox regression analysis (P<0.05). We constructed a reference model using age, body mass index, systolic blood pressure, sodium concentration, and renal function on multivariable Cox regression analysis. Adding SGA to the reference model significantly improved both net reclassification improvement (NRI) and integrated discrimination improvement (0.344, P=0.002; 0.012, P=0.049; respectively). Other indexes (CONUT, PNI, and GNRI scores) significantly improved NRI (0.254, P=0.019; 0.273, P=0.013; 0.306, P=0.006; respectively).
Conclusions: Nutritional screening assessed at hospital admission was appropriate for the prediction of 1-year mortality in hospitalized patients with ADHF.
The prevalence of heart failure increases with age.1 In Japan, the number of elderly people is increasing rapidly, and the prevalence of acute decompensated heart failure (ADHF) is steadily increasing.2 Malnutrition is an important factor in the exacerbation of chronic heart failure.3,4 Regarding the relationship between heart failure and malnutrition, intestinal ischemia affected by low output syndrome causes absorption disorder and decreases peristalsis. Simultaneously, intestinal edema affected by gastrointestinal congestion causes abdominal fullness and loss of appetite, which decreases oral intake despite hypermetabolic conditions due to chronic inflammation in patients with heart failure.3 Consequently, chronic heart failure results in malnutrition.5 Taken together, nutrition status has been extensively examined in patients with chronic heart failure. There are several reports on CHF patients, while few studies have evaluated the relationship between clinical prognosis and nutritional status in ADHF. In addition, appropriate nutrition measurement in ADHF has not been well evaluated. Therefore, the purposes of the present study were to evaluate the impact of nutritional status on 1-year mortality and to assess appropriate nutrition measurements in hospitalized patients with ADHF.
We enrolled 457 consecutive patients with ADHF who were admitted to Kasugai Municipal Hospital, Aichi, Japan, between November 2009 and August 2015. For patients with multiple admissions, the first eligible hospitalization for ADHF was evaluated. We excluded patients with ADHF caused by acute coronary syndrome or malignancy.
Medical records were reviewed to assess clinical characteristics (age, gender, previous medical history, etiology, comorbidities, and laboratory data) and 1-year mortality rates collected from outpatient visits. All patients were followed up for 1 year and we examined the association between 1-year mortality and nutrition status in patients with ADHF retrospectively. We compared all data according to survival status.
The study protocol complied with the Declaration of Helsinki and was approved by the Committees on the Ethics of Human Research of Kasugai Municipal Hospital. Written informed consent was obtained from each patient.
Body mass index (BMI) was calculated using the following formula: BMI=mass (kg)/height2 (m2). We defined anemia as serum hemoglobin <13 mg/dL in men and <12 mg/dL in women.6
Previously established objective nutritional indexes (controlling nutritional status [CONUT], prognostic nutritional index [PNI], geriatric nutritional risk index [GNRI], and subjective global assessment [SGA]) were evaluated at hospital admission. The control criteria are shown in Figure 1. CONUT score was determined using serum albumin level, total cholesterol level, and total lymphocyte count.7 The formula was as follows: CONUT score=albumin score (≥3.5 g/dL, 0 points; 3.0–3.4 g/dL, 2 points; 2.5–2.9 g/dL, 4 points; <2.5 g/dL, 6 points)+total lymphocyte score (≥1,600/mL, 0 points; 1,200–1,599/mL, 1 point; 800–1,199/mL, 2 points; <800/mL, 3 points)+total cholesterol score (≥180 mg/dL, 0 points; 140–179 mg/dL, 1 point; 100–139 mg/dL, 2 points; <100 mg/dL, 3 points). PNI score was calculated using serum albumin level and total lymphocyte count.8 The formula was as follows: PNI=10×serum albumin concentration (g/dL)+0.005×total lymphocyte count (/μL). GNRI score was calculated using serum albumin level, and body weight and height.9 The formula was as follows: GNRI=14.89×serum albumin (g/dL)+41.7×(actual body weight/ideal body weight). Actual body weight/ideal body weight was set to 1 when the patient’s body weight exceeded the ideal body weight. The ideal body weight in the present study was calculated using BMI 22 kg/m2. SGA score was based on history and physical examination: patients were subjectively rated as well nourished (A), moderately malnourished (B), or severely malnourished (C).10 The examiner asked every patient about any involuntary weight loss and/or changes in dietary intake in the past 6 months and 2 weeks, the presence of any gastrointestinal symptoms for >2 weeks (nausea, vomiting, diarrhea, and anorexia), and his/her performance status. Physical examination assessed the severity of subcutaneous fat loss, muscle wasting, and ankle and sacral edema, as well as ascites on a 4-point scale (0–3).
Criteria algorithms for nutrition indexes. Malnutrition was defined as CONUT score ≥5, PNI score <38, GNRI score <92, and SGA scores B,C. Malnutrition was shown in reds. CONUT, controlling nutritional status; GNRI, geriatric nutritional risk index; H, height; PNI, prognostic nutritional index; SGA, subjective global assessment.
Malnutrition was defined as CONUT score ≥5, PNI score <38, GNRI score <92, and SGA scores B and C.
Statistical AnalysisThe distribution of continuous variables was examined using the Shapiro-Wilk test. Continuous variables are expressed as median (IQR), and categorical variables as n (%). Comparison of continuous variables was performed using the Student’s t-test or Mann-Whitney U-test. Categorical variables were analyzed using the chi-squared test or Fisher’s exact probability test. All baseline variables with P<0.05 on univariate Cox regression analysis were entered into multivariate Cox regression analysis to determine independent predictors of 1-year mortality.
We then defined the baseline model, which included factors significantly and independently associated with 1-year mortality on multivariate Cox regression analysis. Finally, we calculated the C-index, net reclassification improvement (NRI), and integrated discrimination improvement (IDI). Differences were considered statistically significant at P<0.05. Statistical analysis was performed using IBM SPSS Statistics 18.0 (IBM, Somers, NY, USA), R version 3.2.1 (R Project for Statistical Computing), and JMP version 5.1 (SAS Institute, Cary, NC, USA).
Baseline patient characteristics are listed in Table 1. Eighty-one patients died (17.7%) within 1 year. The rate of cardiovascular and non-cardiovascular death was 70.4% and 29.6%, respectively. Seventy-nine patients (17.3%) were re-admitted due to heart failure within 1 year. The average age was 79 years (IQR, 70–86 years), and 53.4% of the patients were male. Patients who died were significantly older (P<0.001), whereas BMI, systolic blood pressure (SBP), diastolic blood pressure, heart rate, and the frequency of hypertensive heart disease were significantly higher in patients who survived (P<0.05). With regard to laboratory data, blood urea nitrogen, serum potassium, serum creatinine, and brain-type natriuretic peptide were significantly higher in patients who died (P<0.05), whereas serum sodium, estimated glomerular filtration rate (eGFR), serum albumin, serum total cholesterol, and hemoglobin, and total lymphocyte count were significantly lower in patients who died (P<0.05). Left ventricular ejection fraction and the frequency of systolic dysfunction were not different between the 2 groups. The prevalence of atrial fibrillation or atrial flutter was 25.9% in patients who died, 26.3% in patients who survived, and there was no significant difference between the 2 groups. The use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, β-blocker, aldosterone blocker, and diuretics at hospital admission was similar between the 2 groups (Supplementary Table). The frequency of malnutrition based on each nutrition index is shown in Figure 2. Malnutrition was observed in 31.5% based on CONUT score, 21.4% based on PNI score, 44.9% based on GNRI score, and 27.8% based on SGA score.
| Parameter | Total | Died | Survived | P-value |
|---|---|---|---|---|
| No. subjects (n) | 457 | 81 | 376 | |
| Age (years) | 79.0 (70.0–86.0) | 84.0 (78.5–90.0) | 78.0 (69.0–84.0) | <0.001* |
| Male gender | 53.4 | 48.2 | 54.5 | 0.298 |
| BMI (kg/m2) | 22.4 (19.9–25.1) | 20.6 (18.8–22.8) | 22.7 (20.4–25.6) | <0.001* |
| Current or former smoking (%) | 43.5 | 38.3 | 44.7 | 0.289 |
| SBP (mmHg) | 156 (129–188) | 137 (120–175) | 161 (133–189) | <0.001* |
| DBP (mmHg) | 88 (72–107) | 78 (60–94) | 91 (74–110) | <0.001* |
| Heart rate (beats/min) | 102 (86–125) | 98 (81–114) | 105 (87–125) | 0.045* |
| Etiology (%) | ||||
| Ischemic heart disease | 31.5 | 32.1 | 31.4 | 0.900 |
| Valvular heart disease | 12.4 | 16.1 | 11.7 | 0.297 |
| Cardiomyopathy | 13.1 | 13.6 | 13.0 | 0.895 |
| Hypertension | 21.4 | 12.4 | 23.4 | 0.021* |
| Arrhythmia | 12.0 | 7.4 | 13.0 | 0.137 |
| Serum TC (mg/dL) | 157 (132–821) | 147 (121–176) | 159 (135–184) | 0.009* |
| Serum TG (mg/dL) | 74 (55–102) | 79 (55–105) | 74 (55–99) | 0.797 |
| Serum HDL-C (mg/dL) | 41 (33–51) | 40 (30–46) | 42 (34–52) | 0.017* |
| Serum LDL-C (mg/dL) | 94 (71–116) | 86 (58–112.3) | 95 (73–117) | 0.017* |
| FPG (mg/dL) | 134 (108–188) | 129 (110–190) | 135 (108–188) | 0.998 |
| Blood HbA1c (NGSP, %) | 5.9 (5.4–6.7) | 5.7 (5.2–6.7) | 5.9 (5.5–6.7) | 0.051* |
| BUN (mg/dL) | 24.9 (17.4–35.3) | 33.3 (24.1–51.3) | 23.4 (16.7–33.0) | <0.001* |
| Serum sodium (mEq/L) | 141 (138–143) | 139 (137–142) | 141 (139–143) | <0.001* |
| Serum potassium (mEq/L) | 4.3 (3.8–4.7) | 4.5 (4.1–5.0) | 4.2 (3.8–4.6) | <0.001* |
| Serum creatinine (mg/dL) | 1.1(0.8–1.7) | 1.4 (1.0–2.5) | 1.1 (0.8–1.6) | <0.001* |
| eGFR (mL/min/1.73 m2) | 45.3 (25.9–63.5) | 30.9 (15.5–54.4) | 47.0 (27.6–64.5) | <0.001* |
| Serum uric acid (mg/dL) | 7.2 (5.5–8.8) | 7.4 (5.5–9.9) | 7.2 (5.5–8.6) | 0.221 |
| Serum albumin (mg/dL) | 3.6 (3.2–3.9) | 3.5 (3.1–3.7) | 3.6 (3.2–3.9) | 0.009* |
| Serum CRP (mg/L) | 7.5 (2.2–24.7) | 7.4 (1.8–47.5) | 7.5 (2.2–21.8) | 0.397 |
| BNP (pg/mL) | 756 (410–1,405) | 971 (507–1,623) | 701 (389–1,355) | 0.038* |
| WBC (103/μL) | 7.9 (6.1–10.6) | 7.8 (6.5–10.8) | 8.0 (6.1–10.5) | 0.766 |
| Total lymphocytes (/μL) | 1,332 (868–1,996) | 1,071 (630–1,656) | 1,416 (910–2,036) | 0.001* |
| Hb (mg/dL) | 11.7 (10.1–13.5) | 10.4 (9.1–12.0) | 11.9 (12.0–13.8) | <0.001* |
| Anemia | 62.9 | 84.0 | 58.4 | <0.001* |
| LVEF (%) | 50.0 (39.0–62.8) | 54.5 (43.0–60.8) | 49.8 (38.0–63.0) | 0.319 |
| Comorbidity (%) | ||||
| Type 2 DM | 39.6 | 40.7 | 39.4 | 0.818 |
| Dyslipidemia | 23.4 | 21.0 | 23.9 | 0.566 |
| Hypertension | 63.2 | 59.3 | 64.1 | 0.415 |
| AF or AFl | 26.3 | 25.9 | 26.3 | 0.940 |
Data given as % or median (IQR). *P<0.05. AF, atrial fibrillation; AFl, atrial flutter; BMI, body mass index; BNP, brain-type natriuretic peptide; BUN, blood urea nitrogen; CRP, C-reactive protein; DBP, diastolic blood pressure; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; FPG, fasting plasma glucose; Hb, hemoglobin; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; LVEF, left ventricular ejection fraction; NGSP, National Glycohemoglobin Standardization Program; SBP, systolic blood pressure; TC, total cholesterol; TG, triglycerides; WBC, white blood cells.
Distribution of malnutrition according to nutrition index. Abbreviations as in Figure 1.
We compared malnutrition between the 2 groups (Figure 3). In patients who died, malnutrition was observed in 42.0% based on CONUT score, 32.1% based on PNI score, 60.5% based on GNRI score, and 45.7% based on SGA score. In patients who survived, malnutrition was observed in 29.3% based on CONUT score, 19.2% based on PNI score, 41.5% based on GNRI score, and 23.9% based on SGA. In all indexes, malnutrition was significantly more prevalent in patients who died than in those who survived (P<0.05; Figure 4).
Prevalence of malnutrition according to nutrition index and survival status. Abbreviations as in Figure 1.
Kaplan-Meier curves for all-cause mortality-free survival up to 1 year according to malnutrition status using (A) CONUT score; (B) PNI score; (C) GNRI score; and (D) SGA. Abbreviations as in Figure 1.
We examined predictors of 1-year mortality on univariate and multivariate Cox regression analyses (Table 2). Variables with P<0.05 on univariate Cox regression analysis were included in the multivariate Cox regression analysis. Age, BMI, SBP, eGFR, and serum sodium were significantly associated with 1-year mortality in this study. Therefore, we defined the baseline model using age, BMI, SBP, eGFR, and serum sodium.
| Parameter | Univariate analysis | Multivariate analysis | ||
|---|---|---|---|---|
| P-value | HR (95% CI) | P-value | HR (95% CI) | |
| Age (years) | <0.001* | 1.06 (1.04–1.09) | 0.001* | 1.04 (1.02–1.07) |
| Male gender (%) | 0.249 | 0.88 (0.71–1.09) | ||
| BMI (kg/m2) | <0.001* | 0.86 (0.80–0.91) | <0.001* | 0.89 (0.83–0.96) |
| SBP (mmHg) | <0.001* | 0.99 (0.98–1.00) | 0.005* | 0.99 (0.99–1.00) |
| DBP (mmHg) | <0.001* | 0.98 (0.97–0.99) | ||
| Heart rate (beats/min) | 0.063 | 0.99 (0.98–1.00) | ||
| Serum TC (mg/dL) | 0.002* | 0.99 (0.98–1.00) | ||
| Serum TG (mg/dL) | 0.617 | 1.00 (0.99–1.00) | ||
| Serum HDL-C (mg/dL) | 0.005* | 0.97 (0.95–0.99) | ||
| Serum LDL-C (mg/dL) | 0.004* | 0.99 (0.98–1.00) | ||
| FPG (mg/dL) | 0.866 | 1.00 (1.00–1.00) | ||
| Blood HbA1c (NGSP, %) | 0.062 | 0.78 (0.57–1.01) | ||
| BUN (mg/dL) | <0.001* | 1.04 (1.03–1.05) | ||
| Serum sodium (mEq/L) | 0.025* | 0.95 (0.92–0.99) | 0.044* | 0.96 (0.92–1.00) |
| Serum potassium (mEq/L) | <0.001* | 1.62 (1.23–2.10) | ||
| Serum creatinine (mg/dL) | <0.001* | 1.23 (1.11–1.35) | ||
| eGFR (mL/min/1.73 m2) | <0.001* | 0.98 (0.97–0.99) | 0.007* | 0.99 (0.97–1.00) |
| Serum uric acid (mg/dL) | 0.091 | 1.08 (0.99–1.18) | ||
| Serum albumin (mg/dL) | 0.005* | 0.55 (0.37–0.83) | ||
| Serum CRP (mg/L) | 0.092 | 1.05 (0.99–1.10) | ||
| BNP (pg/mL) | 0.036* | 1.00 (1.00–1.00) | ||
| WBC (103/μL) | 0.216 | 1.00 (1.00–1.00) | ||
| Total lymphocytes (/μL) | 0.024* | 1.00 (1.00–1.00) | ||
| Hb (mg/dL) | <0.001* | 0.79 (0.72–0.87) | ||
| Anemia (%) | <0.001* | 1.85 (1.40–2.55) | 0.079 | 1.30 (0.97–1.82) |
| LVEF (%) | 0.217 | 1.01 (0.99–1.03) | ||
*P<0.05. Abbreviations as in Table 1.
Finally, we calculated the C-index, NRI, and IDI (Table 3). Adding each nutrition index to the baseline model improved the prediction of 1-year mortality compared with the baseline model alone (P<0.05). The addition of CONUT, PNI, or GNRI score significantly improved NRI (P<0.05). The addition of SGA significantly improved both NRI and IDI (P<0.05).
| Parameter | C-index (95% CI) | P-value | NRI | P-value | IDI | P-value |
|---|---|---|---|---|---|---|
| Baseline model | 0.769 (0.716–0.823) | Ref. | Ref. | Ref. | ||
| Baseline+CONUT score | 0.771 (0.717–0.825) | 0.743 | 0.254 | 0.019* | 0.003 | 0.192 |
| Baseline+PNI | 0.772 (0.718–0.826) | 0.604 | 0.273 | 0.013* | 0.005 | 0.130 |
| Baseline+GNRI | 0.770 (0.716–0.824) | 0.874 | 0.306 | 0.006* | 0.002 | 0.358 |
| Baseline+SGA | 0.779 (0.726–0.832) | 0.259 | 0.344 | 0.012* | 0.002 | 0.049* |
*P<0.05. Baseline model included age, BMI, SBP, eGFR, and serum sodium. CONUT, controlling nutritional status; GNRI, geriatric nutritional index; IDI, integrated discrimination improvement; NRI, net reclassification improvement; PNI, prognostic nutrition index; SGA, subjective global assessment. Other abbreviations as in Table 1.
The principal findings of this study were as follows: (1) malnutrition was a predictor of 1-year mortality even in patients with ADHF (similar to the results obtained in a previous study involving patients with chronic heart failure); and (2) adding each nutrition index improved the prediction of 1-year outcome in hospitalized patients with ADHF.
Low BMI or sarcopenia is related to poor prognosis in patients with heart failure.11–14 In contrast, some reports suggest that the severity of chronic disease contributes to frailty status.15 The frailty cycle centered on sarcopenia is attracting attention with regard to chronic diseases that may affect elderly individuals. Regarding the relationship between heart failure and nutrition, malnutrition was observed in 25% of hospitalized patients with chronic heart failure.16 Although the present study also noted variation, the percentage of malnutrition was >20% for each nutrition index.
In the present study patients with malnutrition had a poorer prognosis than the subjects with adequate nutritional status. In short, malnutrition increased the risks of readmission and mortality due to heart failure.17,18
Heart failure is a progressive disease that worsens with repeated exacerbation and readmission, regardless of treatment. In this study, nutrition and heart failure were involved even in the acute phase of heart failure.
Plasma albumin concentration is controlled by several factors, including rate of albumin synthesis, catabolic rate, albumin distribution, and exogenous albumin loss.19–21 Hypoalbuminemia is an independent prognostic predictor in patients with chronic heart failure.16 Lymphocyte count reflects inflammation and the stress response.22 Lower lymphocyte count independently predicts clinical events in patients with chronic heart failure.23 Total cholesterol level reflects lipid metabolism and has also been reported as an independent predictor of clinical events in patients with chronic heart failure.24 CONUT score is a nutritional screening tool used in various fields.7 It is a screening tool used to comprehensively evaluate protein metabolism, immunity, and lipid metabolism. PNI is a nutritional screening tool that has been proposed in the field of surgery to evaluate nutritional status and immunity.8 GNRI is often used for nutritional screening in elderly people.9 It is also used for nutritional screening in patients undergoing hemodialysis.25 These nutrition indexes include albumin level, total cholesterol level, and lymphocyte count proportions to some degree. Objective nutrition indexes may also provide additional prognostic information in patients with chronic heart failure.26 Therefore, we thought that screening using these indexes at the time of admission would be useful to predict 1-year mortality in this study.
SGA is a nutrition screening tool that consists of 5 interviews and 5 physical examinations including both subjective and objective assessments. SGA has been used in patients of almost all ages and can also evaluate the risk of delayed healing or infection.10 In addition, SGA is a subjective comprehensive evaluation that has a significant correlation with albumin level, total cholesterol level, and total lymphocyte count.27 SGA is also related to sarcopenia.28,29 In addition, sarcopenia has also been reported as an independent predictor in patients with chronic heart failure.30 SGA has also been suggested to be a useful nutritional screening tool for chronic heart failure.31 SGA is an indicator that considers both malnutrition and sarcopenia. This simple score enables all medical staff to evaluate comprehensive nutritional status easily.8 Moreover, this assessment does not change with timing, indicating sufficient reproducibility. Therefore, SGA might have the greatest advantage because it can be evaluated using only subjective assessments.
In chronic heart failure, nutrition occupies an important position. Stratifying nutrition status at hospital admission in patients with ADHF can lead to an early response by a nutrition support team (NST) and a multifaceted approach, such as cardiac rehabilitation from the acute phase of hospitalization. Taken together, this could reduce the poor prognosis of ADHF.
Study LimitationsThis study had several limitations. First, it involved only a small number of enrolled patients and was conducted at a single center. Second, nutritional screening also differs depending on the etiology of heart failure. It was difficult, however, to analyze the data for each etiology of death because of the small number of enrolled patients. As a result, larger sample sizes and multicenter studies are needed. Finally, we evaluated these nutrition indexes only at hospital admission. Additional screening should be done during the hospital stay or at discharge.
Nutrition screening at hospital admission might improve the prediction of 1-year mortality in hospitalized patients with ADHF. Nutrition screening at hospital admission could be useful for the stratification of ADHF patients.
We are grateful to the staff at Kasugai Municipal Hospital for their assistance in collecting the medical data.
H.I. received lecture fees from Astellas Pharma, Bayer Pharmaceutical, Daiichi-Sankyo Pharms, and MSD. T.M. received lecture fees from Bayer Pharmaceutical, Daiichi-Sankyo, Dainippon Sumitomo Pharma, Kowa, MSD, Mitsubishi Tanabe Pharma, Nippon Boehringer Ingelheim, Novartis Pharma, Pfizer Japan, Sanofi-aventis, and Takeda Pharmaceutical. T.M. also received unrestricted research grant for Department of Cardiology, Nagoya University Graduate School of Medicine from Astellas Pharma, Daiichi-Sankyo, Dainippon Sumitomo Pharma, Kowa, MSD, Mitsubishi Tanabe Pharma, Nippon Boehringer Ingelheim, Novartis Pharma, Otsuka Pharma, Pfizer Japan, Sanofi-aventis, Takeda Pharmaceutical, and Teijin Pharma. The other authors declare no conflicts of interest.
Please find supplementary file(s);
http://dx.doi.org/10.1253/circrep.CR-18-0018
