Impact of Malnutrition Using Geriatric Nutritional Risk Index in Heart Failure With Preserved Ejection Fraction

JACC: HEART FAILURE

VOL.

-, NO. -, 2019

ª 2019 THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION. PUBLISHED BY ELSEVIER. ALL RIGHTS RESERVED.

Impact of Malnutrition Using Geriatric Nutritional Risk Index in Heart Failure With Preserved Ejection Fraction Masatoshi Minamisawa, MD, PHD,a,b,* Sara B. Seidelmann, MD, PHD,a,* Brian Claggett, PHD,a Sheila M. Hegde, MD, MPH,a Amil M. Shah, MD, MPH,a Akshay S. Desai, MD, MPH,a Eldrin F. Lewis, MD, MPH,a Sanjiv J. Shah, MD,c Nancy K. Sweitzer, MD,d James C. Fang, MD,e Inder S. Anand, MD,f Eileen O’Meara, MD,g Jean-Lucien Rouleau, MD,g Bertram Pitt, MD,h Scott D. Solomon, MDa

ABSTRACT OBJECTIVES This study sought to investigate the relationship between malnutrition and adverse cardiovascular (CV) events in heart failure with preserved ejection fraction (HFpEF). BACKGROUND Malnutrition is associated with poor prognosis in a wide range of illnesses, however, the prognostic impact of malnutrition in HFpEF patients is not well known. METHODS Baseline malnutrition risk was determined in 1,677 patients with HFpEF enrolled in the Americas regions of the TOPCAT (Aldosterone Antagonist Therapy for Adults With Heart Failure and Preserved Systolic Function) trial, according to 3 categories of the geriatric nutritional risk index (GNRI) as previously validated: moderate to severe, GNRI of <92; low, GNRI of 92 to <98; and absence of risk, GNRI of $98. The relationships between malnutrition risk and the primary composite outcome of CV events (CV death, heart failure hospitalization, or resuscitated sudden death) and all-cause death were examined. RESULTS Approximately one-third of patients were at risk for malnutrition (moderate to severe: 11%; low: 25%; and absence of risk: 64%). Over a median of 2.9-years’ follow-up, compared to those with absent risk for malnutrition, moderate to severe risk was associated with significantly increased risk for the primary outcome, CV death and all-cause death (hazard ratio [HR]: 1.34; 95% confidence interval [CI]: 1.02 to 1.76; HR: 2.06; 95% CI: 1.40 to 3.03; and HR: 1.79; 95% CI: 1.33 to 2.42, respectively) after multivariate adjustment for age, sex, history of CV diseases, and laboratory biomarkers. CONCLUSIONS Patients with HFpEF are at an elevated risk for malnutrition, which was associated with an increased risk for CV events in this population. (J Am Coll Cardiol HF 2019;-:-–-) © 2019 the American College of Cardiology Foundation. Published by Elsevier. All rights reserved.

From the aCardiovascular Division, Brigham and Women’s Hospital, Boston, Massachusetts; bDepartment of Cardiovascular Medicine, Shinshu University Hospital, Matsumoto, Nagano, Japan; cDivision of Cardiology, Northwestern University Feinberg School of Medicine, Chicago, Illinois; dDivision of Cardiovascular Medicine, University of Arizona, Tuscon, Arizona; eDivision of Cardiovascular Medicine, University of Utah School of Medicine, Salt Lake City, Utah; fMinneapolis Veterans’ Affairs Hospital, Minneapolis, Minnesota; gMontreal Heart Institute, Montreal, Quebec, Canada; and the hDepartment of Internal Medicine, University of Michigan School of Medicine, Ann Arbor, Michigan. *Dr. Minamisawa and Seidelmann are joint first authors. Supported by U.S. National Institutes of Health/National Heart, Lung, and Blood Institute (NIH/NHLBI) contract HH-SN268200425207C. The content of this article does not necessarily represent the views of the National Heart, Lund, and Blood Institute or of the Department of Health and Human Services. Dr. Minamisawa received support from the Japanese Circulation Society, the Japanese Society of Echocardiography, and the Uehara Memorial Foundation Overseas Research Fellowship. Dr. Shah has personal fees from Philips Ultrasound and Bellerophon Therapeutics and research support from Novartis. Dr. Desai is a consultant for Novartis, Abbott, AstraZeneca, Boston Scientific, Boehringer Ingelheim, Regeneron, Relypsa, Corvidia, and Zogenix. Dr. Pitt is a consultant for Bayer and AstraZeneca. Dr. Solomon has received research grants from Alnylam, Amgen, AstraZeneca, Bellerophon, Bayer, Bristol-Myers Squibb, Celladon, Cytokinetics, Eidos, Gilead, GlaxoSmithKline, Ionis, Lone Star Heart, Mesoblast, MyoKardia, NIH/ NHLBI, Novartis, Sanofi, Pasteur, and Theracos; and is a consultant for Akros, Alnylam, Amgen, AstraZeneca, Bayer, Bristol-Myers Squibb, Cardior, Corvia, Cytokinetics, Gilead, GlaxoSmithKline, Ironwood, Merck, Myokardia, Novartis, Roche, Takeda, Theracos, Quantum Genetics, Cardurion, AoBiome, Janssen, Cardiac Dimensions, and Tenaya. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Manuscript received March 30, 2019; revised manuscript received April 24, 2019, accepted April 25, 2019.

ISSN 2213-1779/$36.00

https://doi.org/10.1016/j.jchf.2019.04.020

2

Minamisawa et al.

JACC: HEART FAILURE VOL.

-, NO. -, 2019 - 2019:-–-

Impact of Malnutrition Assessment in HFpEF

H

ABBREVIATIONS AND ACRONYMS BNP = B-type natriuretic peptide

CV = cardiovascular

eart failure with preserved ejection

hospitalizations, a B-type natriuretic peptide (BNP)

fraction (HFpEF) is a growing

concentration in the previous 60 days of $100 pg/ml

health care problem in the aging

or N-terminal pro–B-type natriuretic peptide (NT-

population and is associated with significant

proBNP) concentration of $360 pg/ml. Details of the

risk for recurrent cardiovascular (CV) events,

study design and baseline characteristics have been

including death and rehospitalization for

published previously (13,14). In the overall popula-

worsening HF. Comorbidities are common

tion, randomization to spironolactone did not reduce

in patients with HFpEF and may contribute

the composite endpoint of CV death, HF hospitali-

to the increased risk (1–4). Malnutrition is

zation, or resuscitated sudden death but was asso-

preserved ejection fraction

highly prevalent in patients with several

ciated with a lower incidence of HF hospitalization

LVEF = left ventricular ejection

chronic illnesses such as cancer and end-

(15). However, in the Americas region, use of spi-

fraction

stage renal disease and is associated with

ronolactone

NT-proBNP = N-terminal pro–

decreased

impaired

reduction in the primary outcome, CV death, hospi-

B-type natriuretic peptide

wound healing, and worsened prognosis (5–

talization for HF (12). A total of 1,677 patients with

GNRI = geriatric nutritional risk index

HF = heart failure HFpEF = heart failure with

immune

function,

was

associated

with

a

significant

7). Although malnutrition also has been reported to

adequate GNRI data were enrolled (Figure 1). The

be associated with a high rate of mortality in HF pa-

primary outcome was the composite of CV events,

tients, there are few studies evaluating the relation-

including CV death, heart failure hospitalization, or

ship between malnutrition and various types of CV

resuscitated sudden death.

events in large, well-characterized cohorts of patients

GERIATRIC NUTRITIONAL RISK INDEX. The GNRI

with HFpEF (8–11). The geriatric nutritional risk index

was determined by using the following formula:

(GNRI) has been validated as a screening tool for

GNRI ¼ (14.89  serum albumin (g/dl)) þ (41.7 

malnutrition in elderly patients by using 3 objective

weight (kg)/ideal body weight (kg)). Ideal body

parameters that are routinely measured in patients

weight was derived by using the following equa-

with HFpEF: body height and weight and serum albu-

tions of Lorentz. Ideal body weight for men (kg) was

min (6–11). Several single-center studies found that

calculated by height [cm] - 100 - [height - 150/4].

malnutrition, as assessed by GNRI values, was associ-

For women, ideal body weight (kg) was calculated

ated with worse prognosis in patients with HF (8–10).

by height [cm] - 100 - [height - 150/2.5]. The weight-

The present authors hypothesized that GNRI may pre-

to- ideal body weight ratio was set to 1 if the pa-

dict adverse events in a large multicenter, interna-

tient’s body weight exceeded the ideal body weight

tional clinical trial that enrolled patients with

(6,16). These variables were used at the baseline

well-characterized HFpEF.

visit in TOPCAT enrollment. Categorization of the

METHODS

patients was performed according to the following cutoffs:

moderate

to

severe

malnourishment

STUDY POPULATION. A total of 1,767 subjects were

risk: <92; low risk: 92 to <98; absence of risk: $98.

enrolled in the Americas regions of the TOPCAT

In contrast to the 4 classes proposed by Bouillance

(Aldosterone Antagonist Therapy for Adults With

et al. (6), we combined the moderate (GNRI: 82

Heart Failure and Preserved Systolic Function) trial

to <92) and severe risk (GNRI: <82) groups because

(United

the number of subjects with severe risk was very

States, Canada,

Brazil,

and

Argentina).

TOPCAT was a multicenter, international, random-

small (n ¼ 28) in this study, which could lead to

ized,

statistical error.

double-blind,

placebo-controlled

trial

that

tested the effects of aldosterone antagonist spi-

STATISTICAL ANALYSIS. Continuous variables were

ronolactone on CV morbidity and mortality. The

summarized as mean  SD if normally distributed or

present study examined only patients enrolled in the

as median and interquartile range otherwise. Cate-

Americas region because of significant differences in

gorical variables were expressed as numbers and

population characteristics and outcomes by region

percentages. Clinical characteristics by GNRI cate-

(12). Eligible subjects were at least 50 years of age

gories were presented with p values for trend using

and had signs and symptoms of HF and a left ven-

linear regression for continuous normally distributed

tricular ejection fraction (LVEF) $45% according to

variables, the Cuzick nonparametric trend test for

local site readings. Randomization was stratified by

continuous non-normally distributed variables, and

the presence of 1 of the following inclusion criteria:

chi-square trend test.

at

least

one

hospitalization

in

the

previous

Kaplan-Meier curves were calculated from the date

12 months for which HF was a major component of

of enrollment to the incidence of CV events and were

the hospitalization or, if there were no qualifying

compared using the log-rank test. Poisson models

JACC: HEART FAILURE VOL.

-, NO. -, 2019

Minamisawa et al.

- 2019:-–-

Impact of Malnutrition Assessment in HFpEF

were used to estimate the incidence rates. A Cox proportional hazards regression analysis was per-

F I G U R E 1 Study Design

formed to identify for predicting CV events, using variables that included clinical characteristics and risk factors. Multivariate analysis was performed to adjust for the effects of baseline risk factors on the incidence of CV events. The GNRI value was adjusted for potential confounders without strong correlation with other variables. Model 1 was adjusted for age, sex, and race (white). Model 2 included the same variables as model 1, including lifestyle factors (smoking, alcohol drinks in the past week, metabolic equivalent-hours per week [activity level], cooking salt score, home meals, and live alone). Model 3 additionally adjusted for CV comorbidities and laboratory data (New York Heart Association [NYHA] functional class, hypertension, diabetes mellitus, prior HF hospitalization, prior myocardial infarction, prior stroke, atrial fibrillation, any cancer, use of

GNRI ¼ geriatric nutritional risk index; TOPCAT ¼ Aldosterone Antagonist Therapy for

angiotensin-converting

Adults With Heart Failure and Preserved Systolic Function.

enzyme

inhibitors/angio-

tensin receptor blockers, usage of beta-blockers, hemoglobin levels at baseline, serum sodium levels at baseline, log total bilirubin levels at baseline, log estimated glomerular filtration rate [eGFR] at baseline, and randomized treatment assignment [e.g., spironolactone vs. placebo]). To assess whether the accuracy of predicting CV events would improve after the addition of GNRI to a baseline

model

with

established

risk

factors,

including age, sex, smoking status, hypertension, diabetes mellitus, prior myocardial infarction, atrial fibrillation, and NYHA functional class, Harrell Cstatistics, net reclassification improvement (NRI), and integrated discrimination improvement (IDI) were calculated. The change in C-statistic values was compared between a baseline model plus body mass index (BMI) or serum albumin alone and the model plus GNRI to evaluate incremental prognostic information by adding an assessment of GNRI values over BMI or serum albumin alone. A p value < 0.05 was considered statistically significant. All analyses were performed using STATA version 14.1 software (Stata Corp., College Station, Texas) and R version 3.3.2 software (Vienna, Austria).

RESULTS

64.1 to 79.5) years old, and approximately 51% were female. There were no significant differences in age or sex among the 3 groups. Self-reported race of “white” was less common in patients in the moderate to severe risk group. Compared to the low or absentrisk group, those in the moderate to severe risk group were more likely to have worse NYHA functional classes, diabetes mellitus, greater insulin use, lower hemoglobin, serum sodium, and eGFR; higher total bilirubin and natriuretic peptide concentrations; and higher tricuspid regurgitation jet velocity. Histories of hypertension, dyslipidemia, HF hospitalization,

myocardial

infarction,

stroke,

and

atrial

fibrillation were similar among the 3 groups. During the median 2.9-year follow-up period (interquartile range: 1.9, 4.2 years), the primary outcomes were observed in 494 patients (incidence rate [per 100 patient-years]: 11.4%). There was a statistically significant, linear association between higher GNRI and lower incidence of the primary outcome (p < 0.0001 for overall trend) (Figure 2). In the Kaplan-Meier analysis, patients with moderate to severe risk for malnutrition showed worse prognoses than the other groups regarding the primary outcome

Median (25th, 75th percentile) GNRI values and mean

(incidence rate [per 100 patient-years]: moderate to

 SD were 99.8 (95.3, 104.2) and 99.6  6.9, respec-

severe risk: 17.8; low risk: 13.3; and absence of risk:

tively. Patients were divided into 3 malnutrition risk

9.8, respectively; log-rank p < 0.0001) (Central

groups; moderate to severe (n ¼ 188), low (n ¼ 424),

Illustration). In the univariate Cox proportional haz-

and absence of risk (n ¼ 1,065) (Figure 1). The pa-

ards analysis, lower GNRI values were associated with

tients’ baseline clinical characteristics are listed in

higher incidence of the primary outcome, CV death,

Table 1. The median age in this study was 72.4 (range

hospitalization for HF, all-cause death, non-CV or

3

4

Minamisawa et al.

JACC: HEART FAILURE VOL.

-, NO. -, 2019 - 2019:-–-

Impact of Malnutrition Assessment in HFpEF

T A B L E 1 Baseline Characteristics of Patients According to GNRI

Malnourishment Status

GNRI Body mass index, kg/m2 Albumin, g/dl Age $65 yrs $75 yrs

Absence of Risk (GNRI >98) (n ¼ 1,065)

Low Risk (GNRI 92 to <98) (n ¼ 424)

Moderate/Severe Risk (GNRI <92) (n ¼ 188)

p Value (For Trend)

102.7 (99.8, 105.7)

95.3 (93.8, 96.8)

87.9 (85.3, 90.8)

-

32.4 (27.8, 37.8)

33.6 (28.9, 39.6)

33.6 (27.7, 40.5)

0.019 <0.001

4.1 (3.9, 4.3)

3.6 (3.5, 3.7)

3.2 (3.0, 3.3)

72.3 (64.1, 79.3)

73.4 (65.4, 80.0)

70.6 (61.9, 78.6)

0.90

777 (73.0)

322 (75.9)

121 (64.4)

0.15

427 (40.1)

189 (44.6)

74 (39.4)

0.57

Females

516 (48.5)

235 (55.4)

98 (52.1)

0.07

Whites

867 (81.4)

321 (75.7)

124 (66.0)

<0.001

NYHA functional class (III or IV)

342 (32.1)

157 (37.1)

91 (48.7)

<0.001

Hypertension

967 (90.9)

376 (88.7)

174 (92.6)

0.99

Dyslipidemia

763 (71.7)

304 (71.7)

137 (72.9)

0.79

Diabetes mellitus

440 (41.4)

204 (48.1)

98 (52.1)

0.001

Insulin users

183 (17.2)

112 (26.4)

61 (32.4)

<0.001

Prior HF hospitalization

614 (57.7)

241 (56.8)

121 (64.4)

0.22

Prior myocardial infarction

216 (20.3)

84 (19.8)

39 (20.7)

0.99

Comorbidities

Prior stroke

96 (9.0)

34 (8.0)

21 (11.2)

0.63

450 (42.3)

195 (46.0)

70 (37.2)

0.65

161 (15.1)

67 (15.9)

27 (14.5)

0.99

Aspirin

628 (59.0)

242 (57.1)

112 (59.6)

0.84

Warfarin

355 (33.4)

159 (37.5)

55 (29.3)

0.84

ACE-inhibitors/ARBs

839 (78.9)

337 (79.5)

145 (77.1)

0.76

Beta-blockers

807 (75.8)

360 (84.9)

144 (76.6)

0.06

Statins

697 (65.5)

276 (65.1)

121 (64.4)

0.75

60 (5.6)

29 (6.9)

17 (9.1)

Atrial fibrillation Bone fracture Medications

Lifestyle factors Current smoker

0.63

Alcohol drinks in the past week

0.86

0

787 (74.0)

307 (72.6)

140 (75.3)

1-4

187 (17.6)

84 (19.9)

34 (18.3)

>5

90 (8.5)

32 (7.5)

12 (6.4)

Activity level, MET h/week

1.0 (0.0, 4.5)

1.2 (0.0, 4.0)

0.7 (0.0, 3.0)

0.15

Cooking salt score*

1.0 (0.0, 4.0)

1.0 (0.0, 4.0)

0.0 (0.0, 4.0)

0.014

Almost none

79 (7.4)

18 (4.3)

13 (7.1)

25%

66 (6.2)

22 (5.3)

6 (3.3)

50%

97 (9.1)

43 (10.3)

19 (10.4)

75%

169 (15.9)

71 (17.0)

21 (11.4)

Almost all

651 (61.3)

264 (63.2)

125 (67.9)

303 (28.5)

115 (27.2)

61 (32.6)

Home meals

0.08

Live alone

0.49

Continued on the next page

unknown death, hospitalization for any reason, hos-

hospitalization for any reason, and hospitalization for

pitalization for CV reason, and hospitalization for

non-CV reasons. In the univariate Cox proportional

non-CV reasons (Table 2). After multivariate adjust-

hazards analysis, the moderate to severe versus the

ments were made for age, sex, and race (white)

group with no malnutrition was associated with

(model 1), lower GNRI values predicted a higher

incidence of adverse events. After multivariate

incidence of CV events. In the model 2 adjusted for

adjustment for model 3, moderate to severe risk

lifestyle factors, lower GNRI values remained an in-

remained associated with significantly increased risk

dependent predictor for CV events. Similar results

for the primary outcome, CV death, all-cause death,

were observed after further adjustments for CV

hospitalization for any reason, and hospitalization for

comorbidities and laboratory data regarding the

non-CV reasons (HR: 1.34; 95% CI: 1.02 to 1.76; HR:

primary outcome, CV death, hospitalization for HF,

2.06; 95% CI: 1.40 to 3.03; HR: 1.79; 95% CI: 1.33 to

all-cause

2.42; HR: 1.28; 95% CI: 1.05 to 1.57; and HR: 1.53;

death,

non-CV

or

unknown

death,

-, NO. -, 2019

JACC: HEART FAILURE VOL.

Minamisawa et al.

- 2019:-–-

Impact of Malnutrition Assessment in HFpEF

T A B L E 1 Continued

Absence of Risk (GNRI >98) (n ¼ 1,065)

Low Risk (GNRI 92 to <98) (n ¼ 424)

Moderate/Severe Risk (GNRI <92) (n ¼ 188)

p Value (For Trend)

Hemoglobin, g/dl

13.1 (12.1, 14.2)

12.4 (11.5, 13.6)

11.9 (10.7, 13.2)

<0.001

Sodium, mEq/l

140 (138, 142)

140 (138, 142)

139 (138, 141)

<0.001

0.58 (0.40, 0.80)

0.60 (0.40, 0.82)

0.60 (0.50, 0.90)

23 (18, 30)

22 (18, 27)

22 (18, 32)

0.79 0.29

Malnourishment Status

Laboratory and echocardiography data

Total bilirubin, mg/dl Aspartate transaminase, U/l

0.002

Alanine transaminase, U/l

21 (15, 31)

24 (16, 32)

22 (15, 31)

Alkaline phosphatase, U/l

83 (66, 110)

82 (66, 108)

89 (70, 119)

0.12

Creatinine, mg/dl

1.1 (0.9, 1.4)

1.1 (0.9, 1.4)

1.2 (0.9, 1.4)

0.05

62.1 (50.1, 77.2)

59.0 (48.1, 74.5)

58.4 (47.8, 75.5)

0.049

237 (143, 435) (n ¼ 394)

267 (156, 448) (n ¼ 187)

288 (192, 465) (n ¼ 89)

0.023

865 (501, 1,688) (n ¼ 224)

1148 (655, 2,035) (n ¼ 88)

1916 (1,050, 3,131) (n ¼ 34)

eGFR, ml/min per 1.73 m2 surface area BNP, pg/ml NT-proBNP, pg/ml LVEF, %

<0.001

58 (51, 64)

60 (55, 65)

58 (54, 62)

0.58

LVEDVi, ml/m2

44.8 (37.9, 54.1) (n ¼ 320)

45.4 (37.5, 56.2) (n ¼ 171)

44.0 (36.5, 53.0) (n ¼ 69)

0.77

LVESVi, ml/m2

18.0 (13.4, 22.3) (n ¼ 320)

18.1 (14.4, 23.2) (n ¼ 171)

15.6 (13.4, 21.1) (n ¼ 69)

0.66

SVi, ml/m2

27.2 (22.7, 33.0) (n ¼ 320)

27.5 (23.1, 31.8) (n ¼ 171)

26.9 (22.9, 32.7) (n ¼ 69)

0.94

LV mass index, g/m2

106.1 (86.4, 123.3) (n ¼ 346)

108.7 (88.0, 130.8) (n ¼ 180)

105.4 (89.2, 122.1) (n ¼ 77)

0.55

Lateral TDI E’, cm/s

8.0 (6.0, 10.5) (n ¼ 191)

8.3 (6.5, 10.0) (n ¼ 106)

6.9 (5.9, 9.9) (n ¼ 44)

0.58

Septal TDI E’, cm/s

6.0 (4.7, 7.8) (n ¼ 190)

5.9 (4.8, 7.8) (n ¼ 119)

5.6 (4.7, 7.3) (n ¼ 45)

0.48

Average E/E’ ratio

13.3 (9.6, 17.0) (n ¼ 137)

14.3 (10.3, 18.1) (n ¼ 87)

13.9 (11.8, 18.7) (n ¼ 32)

0.11

28.5 (21.9, 36.2) (n ¼ 317)

30.4 (23.8, 39.5) (n ¼ 164)

26.7 (19.6, 36.2) (n ¼ 69)

0.99

2.7 (2.4, 3.1) (n ¼ 208)

2.8 (2.5, 3.1) (n ¼ 114)

2.9 (2.6, 3.2) (n ¼ 46)

0.026

48.7 (42.7, 53.8) (n ¼ 264)

48.9 (43.2, 54.7) (n ¼ 139)

49.7 (44.6, 53.4) (n ¼ 58)

LAVi, ml/m2 TR jet velocity, m/s RV FAC (%)

0.50

Values are median (25th, 75th percentiles), n (%), or mean  SD. *Cooking salt score ¼ sum of salt added to staple foods, soup, meat, and vegetables during cooking. Range is 0–12: none ¼ 0; 1/8 tsp ¼ 1; 1/4 tsp ¼ 2; $1/2 tsp ¼ 3 for each food category. ACE ¼ angiotensin-converting enzyme; ARB ¼ angiotensin receptor blocker; BNP ¼ B-type natriuretic peptide; E’ ¼ peak early diastolic mitral annular tissue velocity; eGFR ¼ estimated glomerular filtration rate; GNRI ¼ geriatric nutritional risk index; HF ¼ heart failure; LAVi ¼ left atrial volume indexed to BSA; LVEDVi ¼ LV end-diastolic volume indexed to BSA; LVEF ¼ left ventricular ejection fraction; LVESVi ¼ LV end-systolic volume indexed to BSA; NT-proBNP ¼ N-terminal pro–B-type natriuretic peptide; MET ¼ metabolic equivalent; NYHA ¼ New York Heart Association; RV FAC ¼ right ventricular fractional area change; TDI ¼ tissue Doppler imaging; TR ¼ tricuspid regurgitation.

95% CI: 1.22 to 1.92, respectively). However, the as-

model plus BMI did not show an incremental value

sociation between moderate to severe risk and higher

compared to the baseline model alone. Furthermore,

risk for hospitalization for HF, non-CV or unknown

the change in the C-statistic values between the

death, hospitalization for CV reasons was made

model plus BMI and the model plus GNRI was

nonsignificant

significantly different (p < 0.019). There was a

after

multivariate

adjustment for

model 3 (Table 2).

marginally significant difference of change in the

There was no significant interaction of spi-

C-statistic values between the model plus serum

ronolactone treatment with GNRI values for the

albumin and the model plus GNRI (p ¼ 0.034)

primary outcome, CV death, hospitalization for HF,

(Table 3, Online Figure 1). The NRI and IDI for the

and all-cause death (p value for interaction ¼ 1.00,

primary outcome also significantly increased after

0.93, 0.64, and 0.81, respectively). Regarding model

adding serum albumin (0.093 and 0.009, respec-

discrimination, C-statistics for the primary outcome

tively) and GNRI (0.096 and 0.010, respectively) to

were greater in the baseline model plus serum al-

the baseline model. Furthermore, the incremental

bumin (0.645; p < 0.001) compared with the base-

values of GNRI over BMI or serum albumin alone for

line model alone (0.628) and were greater in the

predicting the incidence of other outcomes were

model plus GNRI (0.647; p < 0.001), whereas the

marginal (Table 3).

5

6

Minamisawa et al.

JACC: HEART FAILURE VOL.

F I G U R E 2 Association between GNRI and Incidence Rate of the Primary Outcome

-, NO. -, 2019 - 2019:-–-

Impact of Malnutrition Assessment in HFpEF

severe malnutrition risk and incidence of CV events remained significant after adjustment for lifestyle and clinical risk factors. GNRI was first reported by Bouillanne et al. (6) in 2005 for predicting the risks of malnutrition-related complications (bedsores and infections) and mortality in hospitalized elderly patients ($65 years of age). GNRI has been recently reported to be significantly correlated with biochemical and anthropometric markers of malnutritional status in those younger than 65 years of age (16). Although previous studies showed that the GNRI was associated with mortality among HF patients in a single-center study, the impact of malnutrition risk in a large cohort of patients with HFpEF was not well known (9). The present study demonstrated the utility of GNRI for predicting future adverse CV events in patients $50 years of age in a multicenter, international trial. There are several potential explanations for the relationship between low GNRI and CV events in patients with HFpEF. First, malnutrition, which can be caused

The incidence rate of the primary outcome (events per 100 person-years) is shown after adjustment for age, sex, and race (white) (left y-axis) (log) and GNRI (x-axis). The solid black curve shows the incidence, with 95% confidence intervals, of the estimates. Primary outcome

by metabolic derangements, has been shown to be a prognostic marker in other patient populations such

was the composite of cardiovascular death, hospitalization for heart failure, or resuscitated

as cancer, renal failure, and acute HF (5–8). Chronic

sudden death. Poisson models were used to estimate the incidence rates. Histograms show

diseases, such as HF, are associated with increased

the population distribution of GNRI. GNRI ¼ geriatric nutritional risk index.

production of catabolic cytokines, muscle catabolism, and appetite suppression and, thereby, lower albumin levels. Aging also decreases one’s metabolic

To evaluate the impact of GNRI on the primary outcome within subgroups, patients were divided into various subpopulations. With adjustment for age, sex, and race (model 1), lower GNRI, suggesting a greater malnutrition risk, was consistently associated with the higher incidence of primary outcomes (Online Figure 2). Additionally, in a sensitivity analysis by $65 or <65 years of age, testing for interaction was negative with regard to other CV events, such as CV death, hospitalization for HF, or all-cause death after adjusting for model 1 (interaction p ¼ 0.73, p ¼ 0.76, or p ¼ 0.20, respectively) (Online Table 1, Online Figures 3 and 4).

reserve of albumin, and therefore, the nutritional status of elderly and chronically ill subjects can be affected by relatively small and/or acute stresses (16). Second, patients with low GNRI were more likely to have characteristics of frailty, consistent with previous reports (8,9). Frailty represents a state of increased vulnerability to stressors resulting from multisystem dysregulation that accompanies aging and is associated with a higher risk of impaired physical functioning and mortality among older adults (17). Nutrition status indicated by MiniNutritional Assessment is associated with the degree of frailty (18). Third, the change in HF patients’ body composition is important because the reduction in lean mass and muscle wasting, as defined by using

DISCUSSION

the criteria of sarcopenia, is associated with worse

In this post hoc analysis of the TOPCAT database, the

trition risk, as assessed by GNRI values, is associated

prognostic impact of malnutrition is reported in pa-

with muscle dysfunction in the elderly (20). Given the

tients with HFpEF. It was found that patients with

fact that multiple comorbidities often coexist and

moderate to severe risk of malnutrition assessed by

overlap

the GNRI values had a higher risk of the primary

values may reflect an advanced phase of systemic

composite CV outcome, CV death, all-cause death,

illness, contributing to the progression of HFpEF.

exercise capacity (19,20). Indeed, increased malnu-

in

patients

with

HFpEF,

lower

GNRI

hospitalization for any reason, and hospitalization for

In the present study, the median GNRI value was

non-CV reasons than those without risk for malnu-

99.8 (moderate to severe risk: 11.2%), which was

trition. Second, the association between moderate to

higher than in the previous studies: 96.5 (for elderly

JACC: HEART FAILURE VOL.

-, NO. -, 2019

7

Minamisawa et al.

- 2019:-–-

Impact of Malnutrition Assessment in HFpEF

C ENTR AL I LL U STRA T I O N Malnutrition Risk and Cardiovascular Outcomes

A

Primary Outcome

Cardiovascular Death

50%

50%

40%

40%

30%

30%

20%

20%

10%

10% Log rank P < 0.0001

0% 0 Number at risk Moderate/Severe 188 Low Risk 424 Absent Risk 1065

C

B

1

2 Years

3

4

132 330 898

90 236 639

61 145 398

36 88 244

Log rank P < 0.0001

0% 0 Number at risk Moderate/Severe 188 Low Risk 424 Absent Risk 1065

Hospitalization for Heart Failure

D

1

2 Years

3

4

157 386 979

114 290 728

84 190 477

52 114 302

All-Cause Death

50%

50%

40%

40%

30%

30%

20%

20% 10%

10% Log rank P = 0.0001

0% 0 Number at risk Moderate/Severe 188 Low Risk 424 Absent Risk 1065

1

2 Years

3

4

132 331 899

90 238 640

61 147 399

36 89 245

Moderate/Severe

Log rank P < 0.0001

0% 0 Number at risk Moderate/Severe 188 Low Risk 424 Absent Risk 1065 Low Risk

1

2 Years

3

4

161 396 1001

118 301 759

86 204 505

54 129 325

Absent Risk

Minamisawa, M. et al. J Am Coll Cardiol HF. 2019;-(-):-–-.

Kaplan-Meier curves for (A) primary outcome, (B) cardiovascular death, (C) hospitalization for heart failure, and (D) all-cause death according to GNRI. The primary outcome was the composite of cardiovascular death or hospitalization for heart failure. GNRI ¼ geriatric nutritional risk index.

subjects in long-term care (mean age 75  8), mod-

have contributed to the observed higher median GNRI

erate/severe risk: 18.3%), 98.2 (for elderly patients

value in the present study. Furthermore, BMI values

with acute decompensated HF (mean age 79  7),

were marginally larger in the moderate to severe risk

moderate/severe risk: 33.1%) [8, 20]. The median age

group. This finding suggests that malnutrition risk

in the present study was 3 to 7 years younger than

even in subjects with higher BMI can still be high,

previous studies, which may, in part, account for this

although it is possible that patients in the moderate to

difference. Additionally, the median BMI value in the

severe malnutrition group may be overestimated.

TOPCAT trial was larger than that in other chronically

Prior analyses of the TOPCAT participants suggested

ill patient cohorts that studied GNRI, which might

that abdominal obesity and greater frailty were

8

Minamisawa et al.

JACC: HEART FAILURE VOL.

-, NO. -, 2019 - 2019:-–-

Impact of Malnutrition Assessment in HFpEF

T A B L E 2 Cox Proportional Hazards Analyses of Clinical Outcomes

Malnourishment Status

Primary outcomes (Number of events) Event rate (per 100 patient-yrs)

Overall (n ¼ 1,677) GNRI, 1 per SD Decrease* HR (95% CI) p Value

Absent Risk (n ¼ 1,065)

Low Risk (n ¼ 424) HR (95% CI) p Value (vs. Absent)

Moderate/ Severe Risk (n ¼ 188) HR (95% CI) p Value (vs. Absent)

(n ¼ 494)

(n ¼ 278)

(n ¼ 140)

(n ¼ 76)

Continuous GNRI* Treatment p Trend C-Statistic Interaction

11.4 (10.5–12.5)

9.8 (8.7–11.0)

13.3 (11.3–15.7)

17.8 (14.2–22.3)

Unadjusted (n ¼ 1,677)

1.29 (1.18–1.41) p < 0.0001

Reference

1.36 (1.11–1.67) p ¼ 0.003

1.80 (1.40–2.32) p < 0.001

<0.001

0.56

Model 1 (n ¼ 1,677)

1.28 (1.18–1.40) p < 0.001

Reference

1.36 (1.11–1.67) p ¼ 0.003

1.75 (1.35–2.26) p < 0.001

<0.001

0.59

Model 2 (n ¼ 1,656)

1.28 (1.17–1.40) p < 0.001

Reference

1.40 (1.14–1.72) p ¼ 0.002

1.71 (1.32–2.22) p < 0.001

<0.001

0.63

Model 3 (n ¼ 1,632)

1.20 (1.09–1.32) p < 0.001

Reference

1.24 (1.00–1.53) p ¼ 0.049

1.34 (1.02–1.76) p ¼ 0.038

0.015

0.68 p ¼ 1.00

Treatment effect, unadjusted HR (95% CI) (Spironolactone vs. placebo) Cardiovascular death (number of events) Event rate (per 100 patient-yrs)

0.84 (0.66–1.06)

0.67 (0.48–0.94)

0.94 (0.60–1.48)

(n ¼ 212)

(n ¼ 114)

(n ¼ 55)

(n ¼ 43)

4.3 (3.7–4.9)

3.6 (3.0–4.3)

4.4 (3.4–5.7)

8.2 (6.1–11.1)

Unadjusted (n ¼ 1,677)

1.30 (1.14–1.47) p < 0.0001

Reference

1.24 (0.90–1.71) p ¼ 0.19

2.34 (1.65–3.32) p < 0.001

<0.001

0.57

Model 1 (n ¼ 1,677)

1.31 (1.15–1.50) p < 0.001

Reference

1.22 (0.88–1.68) p ¼ 0.24

2.40 (1.68–3.43) p < 0.001

<0.001

0.62

Model 2 (n ¼ 1,656)

1.29 (1.12–1.48) p < 0.001

Reference

1.26 (0.91–1.74) p ¼ 0.17

2.31 (1.60–3.33) p < 0.001

<0.001

0.64

Model 3 (n ¼ 1632)

1.24 (1.07–1.43) p ¼ 0.004

Reference

1.15 (0.82–1.61) p ¼ 0.41

2.06 (1.40–3.03) p < 0.001

0.001

0.68 p ¼ 0.93

Treatment effect, unadjusted HR (95% CI) (Spironolactone vs. Placebo) Hospitalization for HF (number of events) Event rate (per 100 patient-yrs)

0.70 (0.48–1.01)

0.57 (0.33–0.98)

1.01 (0.55–1.83)

(n ¼ 379)

(n ¼ 212)

(n ¼ 111)

(n ¼ 56)

8.7 (7.9–9.7)

7.4 (6.5–8.5)

10.5 (8.7–12.7)

13.1 (8.7–17.1)

Unadjusted (n ¼ 1,677)

1.28 (1.16–1.42) p < 0.0001

Reference

1.40 (1.11–1.76) p ¼ 0.004

1.73 (1.29–2.33) p < 0.001

<0.001

0.56

Model 1 (n ¼ 1,677)

1.28 (1.16–1.41) p < 0.001

Reference

1.40 (1.11–1.76) p ¼ 0.005

1.67 (1.24–2.25) p ¼ 0.001

<0.001

0.60

Model 2 (n ¼ 1,656)

1.26 (1.14–1.40) p < 0.001

Reference

1.44 (1.14–1.82) p ¼ 0.002

1.61 (1.19–2.18) p ¼ 0.002

<0.001

0.64

Model 3 (n ¼ 1,632)

1.15 (1.03–1.28) p ¼ 0.013

Reference

1.20 (0.95–1.54) p ¼ 0.13

1.17 (0.85–1.61) p ¼ 0.34

0.18

0.70 p ¼ 0.64

Treatment effect, unadjusted HR (95% CI) (Spironolactone vs. placebo)

0.89 (0.68–1.17)

0.67 (0.46–0.97)

0.79 (0.46–1.34)

(n ¼ 366) 7.1 (6.4–7.9)

(n ¼ 198) 6.0 (5.2–6.9)

(n ¼ 99) 7.6 (6.2–9.2)

(n ¼ 69) 12.9 (10.1–16.3)

Unadjusted (n ¼ 1,677)

1.31 (1.19–1.45) p < 0.0001

Reference

1.27 (1.00–1.62) p ¼ 0.052

2.17 (1.65–2.86) p < 0.001

<0.001

0.56

Model 1 (n ¼ 1,677)

1.35 (1.21–1.49) p < 0.001

Reference

1.25 (0.98–1.59) p ¼ 0.075

2.31 (1.75–3.06) p < 0.001

<0.001

0.63

Model 2 (n ¼ 1,656)

1.33 (1.19–1.47) p < 0.001

Reference

1.30 (1.02–1.66) p ¼ 0.036

2.23 (1.68–2.97) p < 0.001

<0.001

0.65

Model 3 (n ¼ 1,632)

1.22 (1.09–1.36) p < 0.001

Reference

1.19 (0.92–1.53) p ¼ 0.18

1.79 (1.33–2.42) p < 0.001

<0.001

0.69

All-cause death (number of events) Event rate (per 100 patient-yrs)

Treatment effect, unadjusted HR (95% CI) (Spironolactone vs. Placebo)

p ¼ 0.81 0.83 (0.62–1.09)

0.66 (0.45–0.99)

0.96 (0.60–1.54) Continued on the next page

associated with the higher incidence of all-cause

with high BMIs have often been considered at low-

mortality (21,22). Interestingly, BMI values were

risk for malnutrition, the present findings stress the

greatest in the frailest class, in accordance with the

importance for evaluating nutritional status across all

present study (22). Although geriatric populations

BMI categories.

JACC: HEART FAILURE VOL.

-, NO. -, 2019

Minamisawa et al.

- 2019:-–-

Impact of Malnutrition Assessment in HFpEF

T A B L E 2 Continued

Malnourishment Status

Non-cardiovascular or Unknown death (number of events) Event rate (per 100 patient-yrs)

Overall (n ¼ 1,677) GNRI, 1 per SD Decrease* HR (95% CI) p Value

Absent Risk (n ¼ 1,065)

Low Risk (n ¼ 424) HR (95% CI) p Value (vs. Absent)

Moderate/ Severe Risk (n ¼ 188) HR (95% CI) p Value (vs. Absent)

(n ¼ 154)

(n ¼ 84)

(n ¼ 44)

(n ¼ 26)

Continuous GNRI* Treatment p Trend C-Statistic Interaction

3.0 (2.5–3.5)

2.5 (2.0–3.1)

3.4 (2.5–4.5)

4.8 (3.3–7.1)

Unadjusted (n ¼ 1,677)

1.34 (1.15–1.56) p < 0.0001

Reference

1.33 (0.92–1.91) p ¼ 0.13

1.91 (1.23–2.97) p ¼ 0.004

0.003

0.55

Model 1 (n ¼ 1,677)

1.39 (1.18–1.63) p < 0.001

Reference

1.29 (0.89–1.86) p ¼ 0.18

2.11 (1.35–3.30) p < 0.001

0.001

0.66

Model 2 (n ¼ 1,656)

1.36 (1.16–1.60) p < 0.001

Reference

1.35 (0.93–1.95) p ¼ 0.11

2.01 (1.27–3.17) p ¼ 0.003

0.002

0.70

Model 3 (n ¼ 1,632)

1.20 (1.02–1.43) p ¼ 0.032

Reference

1.24 (0.84–1.82) p ¼ 0.27

1.42 (0.87–2.29) p ¼ 0.16

0.12

0.76 p ¼ 0.98

Treatment effect, unadjusted HR (95% CI) (Spironolactone vs. placebo) Hospitalization for any reason (number of events) Event rate (per 100 patient-yrs)

1.05 (0.68–1.61)

0.81 (0.45–1.47)

0.90 (0.42–1.94)

(n ¼ 1009)

(n ¼ 606)

(n ¼ 269)

(n ¼ 134)

34.5 (32.4–36.7)

30.9 (28.5–33.4)

38.4 (34.1–43.3)

50.3 (42.5–59.6)

Unadjusted (n ¼ 1,677)

1.16 (1.09–1.23) p < 0.0001

Reference

1.21 (1.05–1.40) p ¼ 0.010

1.54 (1.27–1.85) p < 0.001

<0.001

0.54

Model 1 (n ¼ 1,677)

1.15 (1.08–1.22) p < 0.001

Reference

1.19 (1.03–1.37) p ¼ 0.019

1.49 (1.23–1.80) p < 0.001

<0.001

0.56

Model 2 (n ¼ 1,656)

1.15 (1.08–1.23) p < 0.001

Reference

1.22 (1.05–1.41) p ¼ 0.008

1.49 (1.23–1.80) p < 0.001

<0.001

0.58

Model 3 (n ¼ 1,632)

1.09 (1.02–1.17) p ¼ 0.008

Reference

1.13 (0.97–1.32) p ¼ 0.11

1.28 (1.05–1.57) p ¼ 0.014

0.008

0.62

Treatment effect, unadjusted HR (95% CI) (Spironolactone vs. placebo)

p ¼ 0.65 0.95 (0.81–1.12)

0.82 (0.65–1.04)

1.00 (0.71–1.40)

Primary outcome was the composite of cardiovascular death, hospitalization for worsening heart failure, or resuscitated sudden death. *GNRI 1 SD ¼ 6.9. Model 1 is adjusted for randomized treatment assignment, age, sex, and race (white). Model 2 includes the same variables as Model 1 þ lifestyle factors (smoking, alcohol drinks in the past week, MET-h/week, cooking salt score, home meals, and lives alone). Model 3 includes the same variables as Model 2 þ cardiovascular comorbidities and laboratory data (NYHA functional class, hypertension, diabetes mellitus, prior HF hospitalization, prior myocardial infarction, prior stroke, atrial fibrillation, any cancer, the use of ACE-inhibitors/ARBs, use of beta-blockers, hemoglobin levels at baseline, serum sodium levels at baseline, log total bilirubin levels at baseline, and log eGFR levels at baseline). CI ¼ confidence interval; HR ¼ hazard ratio; other abbreviations as in Table 1.

Some studies have shown that nutritional in-

of multidisciplinary care management including

terventions with and without nutritional supple-

nutritional care to reduce the risk of mortality in HF

mentation may be beneficial for patients with chronic

(4). Our results provide evidence to support this

HF, showing improvement in the functional class or

statement.

quality of life (23,24). Nutritional intervention in 120

The present study demonstrated that lower GNRI

malnourished, hospitalized HF patients reduced the

values are associated with a higher incidence of CV

risk of all-cause death and the risk of readmission for

events in patients with HFpEF. Particularly as no

worsening of HF (25). Hummel et al. (26) recently

pharmacological strategy has been established for

showed the potentially beneficial effects of dietary

the treatment of HFpEF patients, identifying modi-

intervention on HF-related readmissions in patients

fiable risk factors for adverse events is critical (3).

discharged from HF hospitalization. Adherence to the

GNRI serves as a screening tool for malnutrition risk

Mediterranean diet have also been suggested to be

and would require additional assessment for a

beneficial in people with HF, including those with

definitive diagnosis of malnutrition. However, GNRI

HFpEF (27,28). Further research is required to

provides

determine if nutritional interventions aimed at

measured biomarkers for first-pass nutritional risk

improving GNRI provide a survival benefit or slow

screening in HFpEF patients. Although the present

the progression of symptoms in malnourished HFpEF

study did not evaluate the actual prevalence of

patients. Current guidelines emphasize the addition

malnutrition,

a

practical

GNRI

tool

could

by

using

clinicians

to

routinely

identify

9

10

Minamisawa et al.

JACC: HEART FAILURE VOL.

-, NO. -, 2019 - 2019:-–-

Impact of Malnutrition Assessment in HFpEF

T A B L E 3 Discrimination of Each Predictive Model for Cardiovascular Outcomes Using C-statistics, NRI, and IDI

Predictive Models

C-statistics

p Value

NRI

0.628

<0.0001

Reference

0.628

0.905

p Value

IDI

p Value

Primary outcomes Established risk factors (Model 1) þ BMI alone

Reference

þ Albumin alone

0.645

<0.0001

0.093

0.020

0.009

<0.001

þ GNRI alone

0.647

<0.0001

0.096

0.020

0.010

<0.001

þ GNRI alone vs. þ BMI alone

0.019*

<0.0001

þ GNRI alone vs. þ albumin alone

0.00218*

0.034

Cardiovascular death 0.589

<0.0001

Reference

þ BMI alone

0.602

0.027

0.092

0.13

0.004

þ Albumin alone

0.609

0.002

0.064

0.44

0.003

0.14

þ GNRI alone

0.615

<0.0001

0.073

0.26

0.006

0.047

þ GNRI alone vs. þ BMI alone

0.013*

0.40

þ GNRI alone vs. þ albumin alone

0.0060*

0.028

Established risk factors (Model 1)

Reference 0.08

Hospitalization for HF 0.647

<0.0001

þ BMI alone

0.647

0.501

þ Albumin alone

0.662

<0.0001

0.102

0.007

0.008

0.007

þ GNRI alone

0.663

<0.0001

0.102

0.007

0.009

<0.001

þ GNRI alone vs. þ BMI alone

0.016*

0.004

þ GNRI alone vs. þ albumin alone

0.0008*

0.286

Established risk factors (Model 1)

Reference

Reference

All-cause death 0.622

<0.0001

þ BMI alone

0.623

0.285

þ Albumin alone

0.634

<0.0001

0.055

0.23

0.007

0.027

þ GNRI alone

0.637

<0.0001

0.055

0.13

0.010

0.013

þ GNRI alone vs. þ BMI alone

0.0136*

0.140

þ GNRI alone vs. þ albumin alone

0.00315*

0.034

Established risk factors (Model 1)

Reference

Reference

Established risk factors included age, sex, smoking status, hypertension, diabetes mellitus, prior myocardial infarction, atrial fibrillation, and NYHA functional class. *Differences between 2 models. IDI ¼ integrated discrimination improvement; NRI ¼ net reclassification improvement; other abbreviations as in Table 1.

HFpEF patients at elevated risk for future CV events

BMI, total cholesterol, total lymphocyte count, the

and who may benefit from nutritional support.

Subjective

Global

Assessment

(SGA),

Mini-

Nutritional Assessment short-form (MNA-SF), and STUDY LIMITATIONS. First, this is a post hoc anal-

controlling nutritional status (CONUT) score (10).

ysis and unmeasured confounding factor might

The SGA and MNA-SF require subjective data eval-

have affected the outcomes regardless of adjust-

uated by medical staff. The CONUT score is a sum

ments. Second, the number of events in moderate

of 3 parameters: the serum albumin level, the total

to severe groups was relatively small. Therefore,

cholesterol level, and the total lymphocyte level.

result should be interpreted cautiously, despite

The authors were unable to evaluate and compare

showing that the 1 SD GNRI decrease was related

another nutritional assessment tools, such as the

to a higher risk of CV events. Third, the GNRI

SGA, MNA-SF, or CONUT scores for risk stratifica-

values were only evaluated at baseline, and the

tion in HFpEF patients. GNRI was marginally supe-

authors were unable to elucidate its changes during

rior to serum albumin alone in the C-statistics, NRI,

the follow-up period. Fourth, as systemic inflam-

and IDI. Therefore, whether GNRI is more useful

mation markers such as C-reactive protein were not

among malnutrition risk tools for predicting CV

measured,

proin-

events could not be determined. Despite these

flammatory state on malnutrition risk categories

limitations, our findings provide new insight into

and clinical outcomes were not evaluated in the

the

present study. Fifth, several nutritional indexes

screening and CV prognosis in HFpEF patients.

have been developed, including serum albumin,

Further studies are needed to determine whether

the

influence

of

a

greater

association

between

malnutritional

risk

JACC: HEART FAILURE VOL.

-, NO. -, 2019

Minamisawa et al.

- 2019:-–-

Impact of Malnutrition Assessment in HFpEF

the nutritional-related risk assessment contributes to predicting CV events and clarify the utility for risk stratification in this patient population.

PERSPECTIVES COMPETENCY IN MEDICAL KNOWLEDGE: Patients with HFpEF categorized as having moderate-to-severe risk for

CONCLUSIONS

malnutrition by geriatric nutritional risk index, a validated tool for assessing malnutrition, are at higher risk for adverse out-

Malnutrition risk is associated with the incidence of

comes, including cardiovascular death, HF hospitalization, and

adverse CV events in HFpEF patients from the

resuscitated sudden death. These data suggest that malnutrition

Americas region of the TOPCAT trial. These findings

is a comorbid condition in HFpEF and may be useful for identi-

suggest the potential importance of malnutrition as a

fying patients with HFpEF who are at an elevated risk for future

comorbid condition in HFpEF and warrant validation

CV events.

in an independent study specifically designed to assess the prognostic utility of nutritional status in

TRANSLATIONAL OUTLOOK: Malnutrition assessment would

HFpEF patients.

allow clinicians to identify HFpEF patients at elevated risk for future CV events and those who may benefit from nutritional

ADDRESS FOR CORRESPONDENCE: Dr. Scott D. Sol-

omon,

Cardiovascular

Division,

Brigham

and

Women’s Hospital, 75 Francis Street, Boston, Massa-

support. These findings support the basis for future prospective randomized studies to evaluate the role of nutritional intervention on outcomes in HFpEF patients.

chusetts. E-mail: [email protected].

REFERENCES 1. Shah SJ, Kitzman DW, Borlaug BA, et al.

9. Kinugasa Y, Kato M, Sugihara S, et al. Geriatric

life and clinical outcome in hemodialysis patients.

Phenotype-specific treatment of heart failure with preserved ejection fraction: a multiorgan roadmap. Circulation 2016;134:73–90.

nutritional risk index predicts functional dependency and mortality in patients with heart failure with preserved ejection fraction. Circ J 2013;77:705–11.

Clin Nutr 2016;35:1522–9.

2. Shah AM, Claggett B, Loehr LR, et al. Heart failure stages among older adults in the community:the atherosclerosis risk in communities study. Circulation 2017;135:224–40. 3. Shimokawa H, Miura M, Nochioka K, Sakata Y. Heart failure as a general pandemic in Asia. Eur J Heart Fail 2015;17:884–92. 4. Yancy CW, Jessup M, Bozkurt B, et al., for the Writing Committee Members. American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. 2013 ACCF/ AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. Circulation 2013;128: 240–327. 5. Pressoir M, Desné S, Berchery D, et al. Prevalence, risk factors and clinical implications of malnutrition in French Comprehensive Cancer Centres. Br J Cancer 2010;102:966–71. 6. Bouillanne O, Morineau G, Dupont C, et al. Geriatric nutritional risk index: a new index for evaluating at-risk elderly medical patients. Am J Clin Nutr 2005;82:777–83. 7. Takahashi H, Ito Y, Ishii H, et al. Geriatric nutritional risk index accurately predicts cardiovascular mortality in incident hemodialysis patients. J Cardiol 2014;64:32–6. 8. Honda Y, Nagai T, Iwakami N, et al., for the NaDEF investigators. Usefulness of geriatric nutritional risk index for assessing nutritional status and its prognostic impact in patients aged $65 years with acute heart failure. Am J Cardiol 2016;118:550–5.

10. Lin H, Zhang H, Lin Z, Li X, Kong X, Sun G. Review of nutritional screening and assessment tools and clinical outcomes in heart failure. Heart Fail Rev 2016;21:549–65. 11. 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. J Am Coll Cardiol HF 2018;6:476–86. 12. Pfeffer MA, Claggett B, Assmann SF, et al. Regional variation in patients and outcomes in the Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist (TOPCAT) trial. Circulation 2015;131:34–42. 13. Desai AS, Lewis EF, Li R, et al. Rationale and design of the treatment of preserved cardiac function heart failure with an aldosterone antagonist trial: a randomized, controlled study of spironolactone in patients with symptomatic heart failure and preserved ejection fraction. Am Heart J 2011;162:966–72. 14. Shah SJ, Heitner JF, Sweitzer NK, et al. Baseline characteristics of patients in the treatment of preserved cardiac function heart failure with an aldosterone antagonist trial. Circ Heart Fail 2013; 6:184–92. 15. Pitt B, Pfeffer MA, Assmann SF, et al., for the TOPCAT Investigators. Spironolactone for heart failure with preserved ejection fraction. N Engl J Med 2014;370:1383–92. 16. Beberashvili I, Azar A, Sinuani I, Shapiro G, Feldman L, Sandbank J, Stav K, Efrati S. Geriatric nutritional risk index, muscle function, quality of

17. Nadruz W Jr., Kitzman D, Windham BG, et al. Cardiovascular dysfunction and frailty among older adults in the community: the ARIC study. J Gerontol A Biol Sci Med Sci 2017;72:958–64. 18. Valentini A, Federici M, Cianfarani MA, Tarantino U, Bertoli A. Frailty and nutritional status in older people: the Mini Nutritional Assessment as a screening tool for the identification of frail subjects. Clin Interv Aging 2018;13:1237–44. 19. Emami A, Saitoh M, Valentova M, et al. Comparison of sarcopenia and cachexia in men with chronic heart failure: results from the Studies Investigating Co-morbidities Aggravating Heart Failure (SICA-HF). Eur J Heart Fail 2018;20:1580–7. 20. Cereda E, Vanotti A. The new Geriatric Nutritional Risk Index is a good predictor of muscle dysfunction in institutionalized older patients. Clin Nutr 2007;26:78–83. 21. Tsujimoto T, Kajio H. Abdominal obesity is associated with an increased risk of all-cause mortality in patients with HFpEF. J Am Coll Cardiol 2017;70:2739–49. 22. Sanders NA, Supiano MA, Lewis EF, et al. The frailty syndrome and outcomes in the TOPCAT trial. Eur J Heart Fail 2018;20:1570–7. 23. Colín Ramírez E, Castillo Martínez L, Orea Tejeda A, et al. Effects of a nutritional intervention on body composition, clinical status, and quality of life in patients with heart failure. Nutrition 2004;20:890–5. 24. Rozentryt P, von Haehling S, Lainscak M, et al. The effects of a high-caloric protein-rich oral nutritional supplement in patients with chronic heart failure and cachexia on quality of life, body composition, and inflammation markers: a randomized, double-blind pilot study. J Cachexia Sarcopenia Muscle 2010;1:35–42.

11

12

Minamisawa et al.

JACC: HEART FAILURE VOL.

-, NO. -, 2019 - 2019:-–-

Impact of Malnutrition Assessment in HFpEF

25. Bonilla-Palomas JL, Gámez-López AL, Castillo-Domínguez JC, et al. Nutritional intervention in malnourished hospitalized patients with heart failure. Arch Med Res 2016;

27. Miró Ò, Estruch R, Martín-Sánchez FJ, et al., for the ICA-SEMES Research Group. Adherence to Mediterranean diet and all-cause mortality after an episode of acute heart failure: results of the

preserved ejection fraction. J Am Coll Cardiol Basic Transl Sci 2017;2:513–25.

47:535–40.

MEDIT-AHF Study. J Am Coll Cardiol HF 2018;6: 52–62.

KEY WORDS heart failure with preserved ejection fraction, malnutrition, prognosis

26. Hummel SL, Karmally W, Gillespie BW, et al. Home-delivered meals postdischarge from heart failure hospitalization. Circ Heart Fail 2018;11: e004886.

28. Carbone S, Canada JM, Buckley LF, et al. Dietary fat, sugar consumption, and cardiorespiratory fitness in patients with heart failure with

A PPE NDI X For supplemental figures and table, please see the online version of this paper.