Malnutrition in Heart Failure with Preserved Ejection Fraction

Abstract

Background

Malnutrition may be an important geriatric condition in adults with heart failure with preserved ejection fraction (HFpEF), but studies on its prevalence and associated clinical outcomes are limited. The aim of this study was to determine if malnutrition is associated with short-term morbidity and mortality in ambulatory patients with HFpEF.

Methods

We examined 231 patients with a clinical diagnosis of HFpEF seen at two dedicated academic HFpEF programs (Weill Cornell Medicine and Michigan Medicine) from June 2018 - April 2022. Malnutrition was defined by Mini-Nutritional Assessment Short Form (MNA-SF) scores ≤ 11. The primary endpoint was a 6-month composite of all-cause mortality and all-cause hospitalization. A Cox proportional-hazard models was used to examine the association between malnutrition and the primary endpoint, adjusting for race, prior hospitalization history, and the validated Meta-Analysis Global Group in Chronic (MAGGIC) heart failure prognostic risk score.

Results

The median age of the cohort was 73 years (interquartile range 64–81). The most common comorbid conditions included hypertension (prevalence 81%), atrial fibrillation (43%), and obesity (63%). The prevalence of malnutrition was 42% (n=97), and MNA-SF scores did not significantly correlate with body mass index (R = - 0.02, p = 0.71). At the 6-month follow-up, 62 patients (26.8%) were hospitalized and 4 patients died (1.7%). In a fully-adjusted analysis, malnutrition was independently associated with the composite outcome of all-cause mortality and all-cause hospitalization (HR 1.94 [ 95% CI: 1.17 – 3.20], p = 0.01).

Conclusion

Despite a high prevalence of obesity, 2 out of 5 ambulatory adults with HFpEF are malnourished. Malnutrition was independently associated with adverse outcomes at 6 months. Future work is necessary to develop interventions that can address malnutrition.

Introduction

Heart failure with preserved ejection fraction (HFpEF) is the predominant subtype of heart failure (HF) in the United States, affecting over 3 million people.¹ HFpEF has been accepted as a true geriatric syndrome.^2–4 The mean age of patients with HFpEF is 76 years,⁵ and its prevalence rises exponentially with advancing age.⁶ This is important because advanced age is associated with the concurrence of multiple co-morbid conditions, as well as deficits across multiple domains including cognition, physical function, and social environment.² While recent studies have demonstrated the prognostic importance of comorbid conditions,^7,8 cognitive impairment,⁹ and frailty^10,11 in HFpEF, malnutrition may be an important geriatric condition^12,13 that is not well-studied.

Malnutrition refers to an imbalance of anabolic-catabolic metabolism with a protein-energy deficit.¹⁴ Its etiology is often multifactorial and can include decreased food intake, increased nutrient loss, increased metabolic rate, decreased absorption via intestinal edema, and cytokine dysregulation involving cortisol, renin, angiotensin, aldosterone, epinephrine, and tumor necrosis factor alpha.¹⁵ As a result, malnutrition leads to a loss of muscle, adipose tissue, and bone mass; and is associated with myriad adverse clinical outcomes including increased mortality and prolonged hospitalizations.¹⁶ The consequences of malnutrition have been described in the general population, as well as subpopulations with HF. However, malnutrition studies conducted in HF have predominantly included adults with heart failure with reduced ejection fraction (HFrEF).^15,17 Given well-known differences in the pathophysiology of HFpEF compared to HFrEF, there is a need to examine the influence of malnutrition on outcomes in HFpEF separately from HFrEF. Understanding the prevalence and effect of malnutrition in a condition often driven by obesity, such as HFpEF, is particularly important since the presence of obesity is often misconstrued as a marker of sufficient nutrition. Thus, we leveraged prospective data from two dedicated HFpEF programs in two different geographic regions of the United States to examine whether malnutrition is associated with 6-month adverse outcomes in ambulatory patients with HFpEF.

Methods

Study Population

We examined patients seen at two dedicated outpatient HFpEF Programs at two large academic institutions (Weill Cornell Medicine and Michigan Medicine) in June 2018 through April 2022. Of note, patients with a prior diagnosis of HFpEF and also those seeking evaluation for a potential new diagnosis of HFpEF are seen at these programs. As part of routine care in these programs, patients undergo geriatric assessments^11,18 guided by a recent expert recommendation.¹⁹ Geriatric assessments incorporate assessments of multiple domains using validated tools. Among others, this includes routine assessments of nutrition using the mini-nutritional assessment short form (MNA-SF). These data are inputted into a registry, which is approved by a local Ethics Committee. There were no dedicated interventions in those who had a positive MNA screen, as this study was intended to define the prevalence and prognostic implications of malnutrition in HFpEF. Establishing effective, reimbursable strategies to manage malnutrition is an important aim of national policy and future research.²⁰

For the current study, we examined patients with a clinical diagnosis of HFpEF. Consistent with guidelines, HFpEF was defined as: (i) presence of HF symptoms based on Framingham criteria;²¹ (ii) preserved LVEF ≥ 50%;²² and (iii) absence of alternative HF etiologies, including severe valvular disease, hypertrophic cardiomyopathy, pericardial disease, pulmonary arterial hypertension, cardiac amyloidosis, or other infiltrative diseases. To ensure data was complete, we excluded 10 patients from our cohort due to a lack of clear follow-up in the electronic medical record.

Exposure: Malnutrition

Malnutrition was determined based on the validated MNA-Short Form (MNA-SF),^23,24 and defined as an MNA-SF score ≤ 11. This included those with malnutrition (MNA-SF score < 8, or at risk for malnutrition (MNA-SF score [8–11])²⁵ these classifications were combined because prior work has shown that malnutrition and being at-risk for malnutrition (per MNA-SF) have similar short-term predictive value in older adults.²⁶ The MNA-SF assesses 6 categories: food intake, weight loss, mobility, physical stress or acute illness, cognitive status, and BMI. The MNA-SF takes < 5 minutes to conduct, does not require extensive training, and has excellent diagnostic performance for detecting malnutrition.²⁵

Primary outcome

The primary outcome was a composite of all-cause mortality and all-cause non-elective hospitalization over a 6-month follow-up period after baseline administration of the MNA-SF. Outcomes were ascertained based on review of the electronic medical record. Of note, the electronic medical record used at both institutions incorporates nearly all encounters within the health system, and provides extensive access to healthcare encounters that occur outside of the institution. Moreover, since the majority of patients included in this study were seen multiple times in their respective HFpEF Programs, there were ample encounters within the electronic medical record to ascertain outcomes.

Statistical Analysis

Continuous variables were expressed as median and inter-quartile range (IQR): the differences were compared using the Wilcoxon rank-sum test. Categorical variables were expressed as count and percentage, and assessed using either a Pearson’s Chi-squared test or Fisher’s Exact test.

A scatter plot with 95% confidence interval was created to visualize the relationship between MNA-SF scores and body mass index (BMI). The Pearson’s correlation coefficient was calculated to determine if any correlation exists.

A Kaplan–Meier curve with a log-rank test was used to examine crude differences in outcomes among those with and without malnutrition. To evaluate the association of malnutrition with the composite outcome, we conducted a Cox proportional hazards model analysis adjusting for the following covariates: race (white vs. non-white), severity of illness using the validated Meta-Analysis Global Group in Chronic (MAGGIC) heart failure score, and prior all-cause hospitalization history within 6 months of the baseline encounter.²⁷ Proportional hazards assumption was checked using Schoenfeld residuals against time. The MAGGIC prognostic risk score has been validated in HFpEF as a predictor of morbidity and mortality,²⁸ and includes the following variables: Age, Gender, Diabetes Mellitus, Chronic Obstructive Pulmonary Disease, Smoking Status, BMI, Systolic Blood Pressure, Creatinine, Left Ventricular Ejection Fraction, first HF diagnosis within prior 18 months of baseline visit, New York Heart Association class, and current Beta Blocker, Angiotensin Converting Enzyme Inhibitor, or Angiotensin Receptor Blocker usage. Data for the MAGGIC score and hospitalization history were obtained based on chart review conducted at the time of the baseline assessment. A P-value of < 0.05 was considered statistically significant for all analyses. Statistical analyses were performed with R 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Among 231 patients with HFpEF, 42% had malnutrition (MNA-SF score 8–11: 35.9%; score < 8: 6.1%; Table 1 shows the baseline clinical characteristics of patients with HFpEF stratified by their nutritional status. The median age of patients in this cohort was 73 years (IQR [64–81]), and the majority were female (65.8%) and white (73.6%). The median LVEF was 62 (IQR [60–67]) and median BMI was 33 (IQR [27–40]). The most common comorbid conditions were hypertension (80.5%), atrial fibrillation (42.9%), and obesity (63.2%). In this cohort, 42.0% (n = 97) of patients had malnutrition. Those without malnutrition had slightly higher systolic blood pressure and ACE-I/ARB usage. Those with malnutrition were more likely to be White and be diagnosed with HF within 18 months prior to the initial encounter. As shown in Figure 1, MNA-SF scores and BMI were not significantly correlated (R = - 0.024, p = 0.71). Nearly half of the cohort had an all-cause hospitalization in the prior 6 months—the prevalence of at least 1 all-cause hospitalization in the prior 6 months was numerically higher in those with malnutrition (53.6%) compared to those without (45.5%).

Table 1 –

Baseline study population characteristics stratified by malnutrition status

Characteristic, n (%)	Overall, N = 231	Malnutrition Absent (MNA-SF Score > 11), N = 134	Malnutrition Present (MNA-SF Score ≤ 11), N = 97	P-value
Socio-demographics
Age in years, median (IQR)	73 (64, 81)	74 (68, 80)	72 (60, 82)	0.19
Women	152 (65.8%)	87 (64.9%)	65 (67.0%)	0.74
White Race	170 (73.6%)	92 (68.7%)	78 (80.4%)	0.045
Non-White Race	61 (26.4%)	42 (31.3%)	19 (19.6%)
Hispanic Ethnicity	19 (8.2%)	12 (9.0%)	7 (7.2%)	0.26
Comorbid conditions
Coronary Artery Disease	64 (27.7%)	37 (27.6%)	27 (27.8%)	0.97
Atrial Fibrillation	99 (42.9%)	62 (46.3%)	37 (38.1%)	0.22
Diabetes Mellitus	96 (41.6%)	52 (38.8%)	44 (45.4%)	0.32
Hypertension	186 (80.5%)	108 (80.6%)	78 (80.4%)	0.97
COPD	48 (20.8%)	26 (19.4%)	22 (22.7%)	0.54
Current Smoker	5 (2.2%)	2 (1.5%)	3 (3.1%)	0.65
Obesity	146 (63.2%)	86 (64.2%)	60 (61.9%)	0.72
BMI in kg/m2, median (IQR)	33 (27, 40)	32 (28, 38)	33 (27, 42)	0.62
Systolic BP in mmHg, median (IQR)	130 (117, 142)	134 (119, 148)	124 (115, 138)	0.004
eGFR, median (IQR)	62 (44, 80)	63 (45, 82)	59 (42, 79)	0.39
HF Characteristics
Heart failure diagnosed within previous 18 months	63 (27.3%)	27 (20.1%)	36 (37.1%)	0.004
NYHA Class				0.53
1	7 (3.0%)	4 (3.0%)	3 (3.1%)
2	77 (33.3%)	41 (30.6%)	36 (37.1%)
3	138 (59.7%)	85 (63.4%)	53 (54.6%)
4	9 (3.9%)	4 (3.0%)	5 (5.2%)
LVEF, median (IQR)	62 (60, 67)	62 (60, 67)	61 (60, 67)	0.87
MAGGIC Score, median (IQR)	22 (16, 26)	21 (17, 26)	22 (16, 27)	0.73
All-cause hospitalization within prior 6 months	113 (48.9%)	61 (45.5%)	52 (53.6%)	0.22
Baseline medication use
Beta Blocker	135 (58.4%)	74 (55.2%)	61 (62.9%)	0.24
ACE-I/ARB	86 (37.2%)	57 (42.5%)	29 (29.9%)	0.050
Diuretics	186 (80.5%)	104 (77.6%)	82 (84.5%)	0.19
Statins	157 (68.0%)	96 (71.6%)	61 (62.9%)	0.16

BP: Blood Pressure; BMI: Body Mass Index; NYHA: New York Heart Association; COPD: Chronic Obstructive Pulmonary Disease; ACE-I: Angiotensin Converting Enzyme Inhibitor; ARB: Angiotensin Receptor Blocker; MAGGIC: Meta-Analysis Global Group in Chronic (MAGGIC) heart failure; LVEF: Left Ventricle Ejection Fraction; CAD: Coronary Artery Disease; CKD: Chronic Kidney Disease; IQR: Interquartile Range; eGFR: estimated glomerular filtration rate

Figure 1:

Comparison of MNA-SF Scores and BMI.

Abbreviations: BMI: Body mass index; MNA-SF: Mini nutritional assessment short form

During the 6-month follow-up period, 62 (26.8%) patients were hospitalized and 4 patients (1.7%) died (Table 2). Those with malnutrition experienced 33 (34.0%) hospitalizations and 2 (2.1%) deaths; while those without malnutrition experienced 29 (21.6%) hospitalization and 2 (1.5%) deaths. The majority of hospitalizations were non-cardiovascular in nature, regardless of nutritional status.

Table 2:

6-month clinical outcomes stratified by nutritional status

Outcome, n (%)	Overall, N = 231	Malnutrition Absent (MNA-SF Score > 11), N = 134	Malnutrition Present (MNA-SF Score ≤ 11), N = 97	P-value
Composite outcome	64 (27.7%)	29 (21.6%)	35 (36.1%)	0.016
Mortality	4 (1.7%)	2 (1.5%)	2 (2.1%)	0.99
All-cause hospitalization	62 (26.8%)	29 (21.6%)	33 (34.0%)	0.036
Cause of Hospitalization				0.070
HF-related	17 (7.4%)	7 (5.2%)	10 (10.3%)
Non-HF, CV related	10 (4.3%)	7 (5.2%)	3 (3.1%)
Non-CV related	35 (15.2%)	15 (11.2%)	20 (20.6%)
Not hospitalized	169 (73.2%)	105 (78.4%)	64 (66.0%)

HF: Heart Failure; CV: Cardiovascular; MNA-SF: Mini-Nutritional Assessment Short Form

The Kaplan Meier analysis (Figure 2) showed that patients with HFpEF who had malnutrition experienced the composite outcome of all-cause hospitalization and mortality significantly more frequently than those who did not have malnutrition (HR: 1.84, 95% CI (1.12–3.01), p = 0.014). In a fully-adjusted Cox proportional hazard model, malnutrition remained significantly associated with the composite outcome of all-cause death or hospitalizations (HR: 1.94 [95% CI 1.17 – 3.20], p = 0.01).

Figure 2:

Kaplan-Meier curve for the composite outcome of all-cause mortality or hospitalization for patients with (MNA-SF ≤11) and without (MNA-SF > 11) malnutrition.

Abbreviations: MNA-SF: Mini nutritional assessment short form

Discussion

In this study of ambulatory patients with HFpEF from two dedicated HFpEF programs in distinct geographic regions of the United States, we found that malnutrition was common and associated with short-term adverse outcomes. These findings highlight the importance of nutrition in the evaluation and management of patients with HFpEF.

Malnutrition is common and a known critical determinant of poor clinical outcomes in acute decompensated HF patients.^15,17,29 Whether malnutrition is associated with adverse outcomes in ambulatory patients with HFpEF is less clear given divergent findings to date. In a sub-group analysis of the Aldosterone Antagonist Therapy for Adults With Heart Failure and Preserved Systolic Function (TOPCAT) trial, malnutrition, based on the Geriatric Nutritional Risk Index (GNRI) criteria, was found to be associated with a myriad of adverse outcomes including cardiovascular death, all-cause death, hospitalization for any reason, and hospitalization for non-cardiovascular reasons over a 3-year follow-up.³⁰ Meanwhile, in a single center study from Spain (n = 555), malnutrition, based on the MNA-SF, was not associated with adverse outcomes at 2 years.³¹ In addition to differences in assessment tools (GNRI vs. MNA-SF), it is important to note that patients with malnutrition from the study done in Spain were provided formal dietary counseling, which may have attenuated its adverse effects. Our study helps to reconcile these differences by examining the association between malnutrition and adverse outcomes using the MNA-SF in a cohort of ambulatory patients with HFpEF, showing that MNA-SF scores are inversely associated with adverse outcomes. Our findings were driven largely by non-cardiovascular hospitalizations—this is important since non-cardiovascular events are common in adults with HFpEF and have important consequences on morbidity and mortality in adults with HFpEF.^32,33 These findings should stimulate future studies examining underlying of mechanisms driving the associations observed. For example, investigation of the microbiome and possibly telomere length in relation to malnutrition (and diet, more broadly) could be fruitful areas to better understand the role of this important geriatric syndrome in HFpEF.

The prevalence of malnutrition and its association with adverse outcomes underscore the importance of considering malnutrition when evaluating adults with HFpEF. Some clinicians may be inclined to use BMI as a proxy for malnutrition; however, our findings here show that MNA-SF scores do not correlate well with BMI. Similar observations have been made when using BMI as a proxy for other geriatric conditions like frailty³⁴ and sarcopenia.³⁵ It is important to clarify that frailty, sarcopenia, and malnutrition are related but distinct clinical entities from each other and also different from BMI; each have important prognostic implications.³⁶ While there have been prior calls to formalize routine evaluation of frailty³⁷ and sarcopenia³⁶ in older adults with various cardiovascular diseases including HF, our findings here additionally support the need to incorporate screening tools for malnutrition as well. To effectively screen for malnutrition, it may be reasonable to use the MNA-SF given its validity and brevity. Optimal approaches to incorporating these evaluations into busy clinical practice have yet to be identified. Future work should examine whether incorporating routine assessments of nutrition and other geriatric conditions² can improve outcomes in HFpEF. In addition, these findings suggest the need for further work to better understand the link amongst geriatric conditions like frailty, sarcopenia, and malnutrition; and to understand underlying pathophysiologic mechanisms driving the associations between geriatric conditions and adverse outcomes.

Prior work suggests that malnutrition may be modifiable. Indeed, several studies suggest that nutritional interventions in malnourished HF patients can improve clinical outcomes.^38–41 For example, Bonilla et. al demonstrated that individualized nutritional regimens can decrease 12-month all-cause mortality (HR 0.37, 95% CI, 0.19–0.72, p = 0.003) and HF readmission rates (10.2 vs. 36.1%, p = 0.001) in hospitalized patients with HF.³⁸ Additionally, Hersberger et. al recently displayed that providing nutritional support to malnourished hospitalized HF patients (19% HFpEF) significantly decreased the 30 & 180 day mortality rate (odds ratio: 0.44; 95% CI: 0.26 to 0.75; p = 0.002), as well as the 30-day risk for major adverse cardiovascular events (odds ratio: 0.50; 95% CI: 0.34 to 0.75; p = 0.001).⁴¹ To our knowledge, these are the only studies to examine the efficacy of nutritional interventions in HF on morbidity and mortality. This indicates a major need for large-scale studies to examine whether addressing malnutrition can improve outcomes in HFpEF, as well as the other HF-subtypes.

There were several limitations in our study. First, we only evaluated nutritional status at baseline; naturally, nutritional status can change over time and differentially impact outcomes. Although we used a validated tool to identify those with malnutrition, we did not have data on key elements of malnutrition like caloric or macronutrient intake. Second, our results are vulnerable to limitations inherent to chart review. This includes the possibility of inaccurate or missing documentation related to clinical characteristics and/or outcomes. Of note, both institutions studied here have robust electronic medical record systems that link to multiple other hospitals within the local region and across the country, which mitigates this concern. Third, although our analysis adjusted for multiple clinical characteristics (via MAGGIC—a validated prognostic tool), there are other conditions that could have impacted outcomes. For example, other geriatric conditions such as frailty, depression, and/or cognition could have impacted our findings, but we did not have the complete data to adjust for these. Future work is needed to better understand the overlap of various geriatric conditions, and their synergistic effect on outcomes.

Conclusion

Our study found that malnutrition occurred in 2 out of 5 ambulatory adults with HFpEF, despite a high prevalence of obesity in this population. We also found that malnutrition was associated with adverse outcomes at 6-months. These results support screening for malnutrition in adults with HFpEF, and highlight the need to develop and study nutritional interventions to improve outcomes in this vulnerable population.

Key Points:

• Malnutrition is prevalent in an obese predominant ambulatory HFpEF patient cohort.

• BMI values are weakly correlated with malnutrition in HFpEF.

• Malnutrition is an independent predictor of short term adverse events in HFpEF.

Why does this matter?

This study supports routine nutritional screenings in ambulatory HFpEF patients and highlights the importance of future research to determine the impact of nutritional interventions.