Dietary Habits are Related to Outcomes in Patients With Advanced Heart Failure Awaiting Heart Transplantation

Dietary habits are related to outcomes in patients with advanced

heart failure awaiting heart transplantation

Short title: Dietary habits and waiting list outcomes

Heike Spaderna, PhD, Department of Psychology, Johannes Gutenberg

University, Mainz, Germany

Daniela Zahn, PhD, Department of Psychology, Johannes Gutenberg

University, Mainz, Germany

Johanna Pretsch, MS, DFG Graduiertenkolleg, University Koblenz-

Landau, Landau, Germany

Sonja L. Connor, MS, RD, Lipid Disorders and Clinical Nutrition,

Oregon Health & Science University, Portland, OR, USA

Armin Zittermann, MD, Clinic for Thoracic and Cardiovascular

Surgery, Heart Center North Rhine-Westphalia, Bad Oeynhausen,

Germany

Stefanie Schulze Schleithoff, PhD, Clinic for Thoracic and

Cardiovascular Surgery, Heart Center North Rhine-Westphalia, Bad

Oeynhausen, Germany

Katrina A. Bramstedt, PhD, Bond University School of Medicine,

Queensland, Australia

Jacqueline M. A. Smits, Eurotransplant International Foundation,

Leiden, The Netherlands

Gerdi Weidner, PhD, Department of Biology, San Francisco State

University, San Francisco, CA

1

Sources of support: This work was supported by grants from the

Alexander-von-Humboldt Foundation (GW); Eurotransplant International

Foundation; German Academic Exchange Service (GW); German Research

Foundation (SP 945/1-1, SP 945/1-3, SP 945/1-4 to HS, and MA 155/75-

1 to GW); and the Johannes Gutenberg-University, Mainz (HS).

Corresponding author and address for reprint requests: Dr. Heike

Spaderna, Psychologisches Institut, Johannes Gutenberg-Universität,

Binger Str. 14-16, 55099 Mainz, Germany. Telephone: 0049-6131-39-

39166. Fax: 0049-6131-39-39154, [email protected]; or Gerdi

Weidner, PhD, Department of Biology, San Francisco State University,

3150 Paradise Dr., Tiburon, CA 94920, USA. Telephone: 001-415-331-

8058. Fax: 001-415-435-7121, [email protected]

Abstract

Background: Empirical evidence supporting the benefits of

dietary recommendations for patients with advanced heart

failure is scarce. We prospectively evaluated the relation of

dietary habits to pre-transplant clinical outcomes in the

multi-site observational Waiting for a New Heart Study.

Methods: 318 heart transplant candidates (82% male, 53±11

years) completed a Food Frequency Questionnaire [foods high in

2

salt, saturated fats, poly-/monounsaturated fats (PUFA+MUFA),

fruit/vegetables/legumes, and fluid intake] at time of

waitlisting. Cox proportional hazard models controlling for

heart failure severity (e.g., Heart Failure Survival Score,

creatinine) estimated cause-specific hazard ratios (HR)

associated with each dietary habit individually, and with all

dietary habits entered simultaneously.

Results: During follow-up (median=338 days; range 13 to 1394),

54 patients died, 151 were transplanted (110 in high-urgency

status, 41 electively), and 45 became delisted (15

deteriorated, 30 improved). Two robust findings emerged:

frequent intake of salty foods, which correlated positively

with saturated fat and fluid intake, was associated with

transplantation in high-urgency status [HR=2.90, 95%

confidence interval (CI): 1.55-5.42]. Frequent intake of foods

rich in PUFA+MUFA reduced the risk for death/deterioration

(HR=0.49, 95% CI: 0.26-0.92).

Conclusion: These results support the importance of dietary

habits for the prognosis of patients listed for heart

transplantation, independent of heart failure severity.

3

Keywords: Epidemiology; prognosis; waiting list; heart

transplantation

4

Introduction

The prevalence of heart failure is estimated to rise by 25%

between 2010 and 2030, threatening not only the health of

individuals, but also contributing to medical costs.1 Among

patients with advanced heart failure awaiting heart

transplantation the numbers of patients listed and

transplanted in “high-urgency” status is increasing

substantially,2 challenging organ allocation and post-

transplant survival.3

To slow disease progression, guidelines for the

management of chronic heart failure and transplantation

candidates consistently include nutritional recommendations

regarding dietary salt and fluid restriction,4-7 especially for

patients with hyponatremia or fluid overload. Nonetheless, the

clinical benefits of restricting salt and fluid intake remain

unclear,8 especially for the prognosis of patients with

advanced heart failure awaiting transplantation. Furthermore,

evidence suggests that the beneficial effects of ω-3

polyunsaturated fatty acids (ω-3PUFA) on cardiovascular health

may extent to patients with heart failure.9 In addition, fruit

and vegetable intake has been associated with reduced

5

incidence of heart failure.10 However, empirical evidence

supporting the value of these dietary factors for the

prognosis of patients with advanced heart failure is scarce.

We evaluated the role of dietary habits assessed at time

of waitlisting on outcomes in the multisite Waiting for a New

Heart Study, a prospective study of patients with advanced

heart failure newly listed for cardiac transplantation.

Employing a competing risks approach and time-to-event

methods, we examined associations between consumption

frequencies of salty foods, foods high in poly- and

monounsaturated fatty acids (PUFA+MUFA), foods high in

saturated fatty acids, fruits/vegetables/legumes, and the

outcomes death on the waiting list, delisting due to clinical

deterioration, high-urgency transplantation, delisting due

clinical improvement, and elective transplantation.

Methods

Procedure and participants

The Waiting for a New Heart Study is an ongoing prospective

multi-site observational study of patients newly listed for

cardiac transplantation in 17 hospitals (16 in Germany, 1 in

6

Austria). Its primary objective is the identification of

psychosocial and behavioral predictors of pre-transplant

outcomes. For the present report patients were followed-up

until January 2009. The study procedures have been described

previously. Briefly, recruitment took place between April 1,

2005 and December 31, 2006. Written informed consent was

obtained from patients, who were newly registered on the

waiting list. Exclusion criteria were aged <18, being listed

for combined heart-lung transplantation, re-transplantation,

not being fluent in German, and too severely ill to

participate, as rated by the local physician. Of 479 newly

listed patients 380 met inclusion criteria.12 Questionnaires

were mailed to 340 patients who consented, and completed by

318 patients within a median of 15 days since listing

(interquartile range 15.25). Comparisons of non-participants

with participating patients have been reported previously.12

The study was approved by local ethic committees and conforms

to the principles outlined in the Declaration of Helsinki.

Assessment of dietary habits

7

A food frequency questionnaire adapted from the Fragebogen zur

Erfassung des Gesundheitsverhaltens (FEG); Questionnaire for the

Assessment of Health Behavior 14 was administered to assess

consumption frequencies of 33 food items and 5 alcoholic

beverages (beer, red wine, white wine/champagne, spirits,

other alcoholic drinks). Participants were asked to specify

how often they consumed the listed foods and drinks (4 =

“daily”, 3 = “several times a week”, 2 = “occasionally”, 1 =

“never”). Food items were grouped a priori and independently

by two dieticians according to their content of salt,

saturated or polyunsaturated and monounsaturated fatty acids,15

or fresh fruits, vegetables, and legumes. Based on these

ratings, four scores were calculated to measure frequency of

intake of salty foods, foods high in saturated fats, foods high in

polyunsaturated and monounsaturated fats (PUFA+MUFA), and

fruits/vegetables/legumes by adding frequency ratings of each food

item (Figure 1). Three foods that are high in salt and

saturated fats were used in both the salt and the fat scores.

Each score was divided by the number of items included.

Frequency of alcohol consumption was computed in the same

manner and was used as a covariate. Psychometric properties of

8

the FEG food frequency questionnaire including its retest

reliability over a period of 4 to 6 months have been

reported.14

Fluid consumption was based on “average daily fluid intake”

(1 = “less than 1 liter”, 2 = “1 to 1.5 liters”, 3 = “1.5 to 2

liters”, 4 = “more than 2 liters”. In addition to considering

fluid consumption per se, we also created a variable that

adjusts fluid intake for cardiac performance by dividing fluid

consumption by cardiac index. The reason for this is that

restrictions in fluid intake are often based on the patient’s

heart failure severity, i.e, more severe restrictions for

patients with highly advanced heart failure.16 The cardiac

index is a well-accepted and objectively measured indicator of

heart failure severity that relates cardiac output to body

surface area, thus adjusting heart performance to the size of

the individual. Therefore, values >1 denote a high fluid

intake relative to a low cardiac index and values <1 denote a

low fluid intake in the presence of better cardiac function

(i.e., high cardiac index).

Assessment of medical, demographic, and other covariates

9

Eurotransplant provided medical information at time of listing

(Table 1), including anthropometric variables, medications,

and seven medical parameters to calculate the Heart Failure

Survival Score (HFSS; Table 1). The seven parameters consist

of resting heart rate, ejection fraction, mean blood pressure,

peakVO2, serum sodium, intraventricular conduction delay,

ischemic diagnosis and are included in Table 1. The HFSS has

been developed in ambulatory patients undergoing

transplantation,17 and has acceptable prognostic performance

even in the era of beta-blocker use. Data on serum total

cholesterol and LDL-cholesterol at time of listing could be

obtained from 11 hospitals.

Demographic variables (including in/outpatient status) were

assessed via questionnaire. Other lifestyle variables, such as smoking

history, alcohol consumption (based on the five items included

in the food frequency questionnaire), and physical activity

were considered as covariates.

Endpoints

Waiting list outcomes were based on type and date of waiting

list status change until January 2009 since date of wait

10

listing, provided by Eurotransplant. Outcomes were death on

the waiting list, high-urgency transplantation (i.e.,

transplantation after having received high- urgency status

because of health decline), elective transplantation

(transplantation while not in high-urgency status), delisting

due to severe clinical deterioration, or delisting due to

clinical improvement. As only 15 patients were delisted

because of clinical deterioration, this outcome and death were

combined into one endpoint. Transplantation in high-urgency

status is indicated if patients show signs of clinical

deterioration, such as receiving intensive care with a)

Cardiac index <2.2 l/m²/min or mixed venous oxygen saturation

<55%, while on inotropic therapy for at least 48 h and

beginning secondary organ failure, and b) life threatening

assist device complications. High-urgency status has to be

approved by a Eurotransplant audit group and requires weekly

re-evaluation.

Statistical analysis

Missing data in food items comprised less than 1%. To deal

with missing data in medical baseline parameters ranging from

11

0.6% (heart rate) to 24.8% (peakVO2), a semi-parametric

multiple imputation procedure was employed. Analyses were

conducted across the 10 imputed data sets and results were

pooled using R 2.12.0 and the packages mitools 2.0.1, cmprsk

2.2-1, survival 2.36-2, and Zelig version 3.4-8.

Absolute numbers, percentages, means and standard

deviations were computed for all variables. For descriptive

purposes, sample characteristics are reported for original

data. Associations between continuous variables were assessed

using Pearson correlation coefficients. Associations between

dietary habits and categorical variables were assessed via

point biserial correlations. To compare frequency data between

groups (e.g., patients with and without hyponatremia), Chi-

square tests were used.

To examine whether food groups assessed at time of

listing were related to waiting list outcomes, we employed a

competing risks approach, considering the mutual exclusive

outcomes death/deterioration, high-urgency transplantation,

elective transplantation, and delisting due to clinical

improvement as competing events. This implies that, if a

patient experiences one event (e.g., high-urgency

12

transplantation), the probability to experience the other

events of interest (e.g., death) is altered.23 Thus, we plotted

cumulative incidence functions for all outcome types, i.e.,

the proportion of patients having experienced an outcome over

the course of time and report subdistribution estimates for

each outcome.

To evaluate if dietary habits were associated with time

until outcomes, Cox proportional hazard regression was

employed, considering the impact of each of the 5 continuously

measured dietary variables on each of the competing events.24

First, each of the five dietary habits was tested in

univariate analyses, for which we report cause-specific, i.e.,

outcome-specific, hazard ratios. Second, multivariate analyses

were conducted, adjusted for the standard covariates: age,

sex, disease duration, BMI, and heart failure severity (serum

creatinine, HFSS, inpatient status), and other health

behaviors. Significant effects were further evaluated by

considering other medical variables associated with the

particular food group or fluid intake (correlations with P <

0.05) as additional covariates. Finally, an adjusted model

using all dietary habits and fluid intake adjusted for cardiac

13

index together in one step was built for each of the competing

outcomes to evaluate the robustness of the findings. To

illustrate the significant effects of dietary habits on

competing outcomes, cumulative incidence functions were

plotted for groups of “rare” (below the median) and “often”

consumption (above the median), as determined by median split

of frequency scores. Patients lost to follow-up were censored

at their time of delisting. The proportional hazard assumption

of included variables was evaluated using scaled Schoenfeld

residuals. As recommended,21 analyses were repeated with the

unimputed data, i.e., the reduced sample with complete data.

Results were considered statistically significant if two-sided

P-values were less than 0.05.

Results

Baseline characteristics

Sample characteristics of 318 newly listed transplant

candidates (18 to 75 years of age) are presented in Table 1.

Consumption frequencies are displayed in Figure 1. Mean fluid

intake adjusted for cardiac index was 1.47 (SD = 0.61),

ranging from 0.34 to 5.50. Thirteen percent of the patients

14

reported to drink more than 2 L/d (Figure 1). Patients with

hyponatremia (serum sodium <130 mmol/L; n = 11, 3.5% of the

entire sample) were more likely to be among those who drank >2

L (9.5%) than to be among those who drank <2 L (2.6%; χ2(1) =

5.26, P = 0.044), and 9 out of the 11 had fluid intake/cardiac

index ratios >1 (range 1.1 to 4.0; data not shown).

Associations among dietary habits and with demographic and

medical characteristics are displayed in Table 2. There were

no significant correlations of food groups with medications,

including diuretics (data not shown). In order to obtain an

estimate of the accuracy of self-reported food intake, we

correlated consumption frequency of foods high in saturated

fats with plasma cholesterol levels of patients who were not

taking lipid-lowering medications (n = 77). Frequent

consumption of foods high in saturated fats was significantly

associated with LDL cholesterol (r = 0.25, P < 0.05), thus

providing validation for this food group.

Association of dietary habits with endpoints

Participants were observed for a mean follow-up of 462.8 days

(SD = 396.2, median = 338, Min = 13, Max = 1394). Cumulative

15

incidences of outcomes were 36.6% for high-urgency

transplantation (n = 110), 25.5% for death/deterioration (n =

54 deaths, n = 15 delistings due to clinical deterioration),

13.3 for elective transplantation (n = 41), 10.6% for delisting

due to improvement (n = 30). Six patients (2.0%) were lost to

follow up; one woman and three men withdrew their consent for

transplantation, one male patient was delisted due to non-

compliance, for one male patient the reason for delisting was

not documented. Sixty-two patients (12.4%) were still on the

waiting list by end of follow-up.

In univariate analyses a more frequent salty food intake was

associated with shortened time to transplantation in high

urgency status, reflecting clinical deterioration (Table 3,

Figure 2). Multivariate adjustment controlling for standard

covariates did not alter this finding: a 1-unit increase in

consumption frequency of salty food intake (for example, an

increase from “occasional” to “several times a week”) was

associated with an almost 3-fold hazard for this outcome

(Table 3). Additional adjustment for diuretic use reduced the

risk associated with salty foods only minimally (HR = 2.88,

95% CI: 1.54, 5.37, P < 0.001). Because diabetes and previous

16

heart surgery were correlated with reduced salt intake (Table

2), we further adjusted for these two variables. Results did

not change substantially (HR = 2.84, 95% CI: 1.51, 5.34, P =

0.0012).

A similar pattern of findings for this outcome emerged

for foods high in saturated fats, which is not surprising given

that both scores were highly correlated. After multivariate

adjustment, the effect of foods high in saturated fats on

high-urgency transplantation was reduced (P = 0.088; Table 3).

The same was true for elective transplantation.

Consumption of foods rich in PUFA+MUFA was positively

associated with a reduced hazard ratio for

death/deterioration, which was maintained after controlling

for standard covariates (Table 3, Figure 3). Thus, a 1-unit

increase in consumption frequency of foods rich in PUFA+MUFA

(e.g., from “occasionally” to “several times a week”) was

associated with a 50% risk reduction for this outcome.

Additional adjustment for diabetes and cardioverter

defibrillator (positively correlated with PUFA+MUFA) did not

alter this result (HR = 0.49, 95% CI: 0.26, 0.92, P = 0.028).

This effect was maintained when each of the items “vegetable

17

oil” and “fish, seafood” was entered separately into the model

(both P-values < 0.05).

A trend emerged pointing at an increased chance for

delisting due to improvement with more frequent consumption of

fruits/vegetables/legumes (Table 3). Additional adjustment for

cardiac index strengthened this result (HR = 3.89, 95% CI:

1.14, 13.29, P = 0.030). There was also a trend for

fruits/vegetables/legumes to be associated with an increased hazard

ratio for high-urgency transplantation (Table 3). However,

this effect was statistically insignificant when cardiac index

was added to the set of standard covariates (HR = 1.58, 95%

CI: 0.93, 2.68, P = 0.091).

Fluid intake per se was not associated with any of the

outcomes (data not shown; all P-values > 0.33). However, when

fluid intake was adjusted for cardiac performance (cardiac

index), a trend emerged for higher fluid intake/cardiac index to

increase the risk for high-urgency transplantation in

unadjusted and adjusted models (Table 3). Among the 9 patients

with hyponatremia who also had unfavorable fluid intake (fluid

intake/cardiac index >1) 5 experienced deterioration of health

(2 died, one was delisted due to deterioration, 2 received

18

high-urgency transplantation); one was delisted due to other

reasons, and 3 electively transplanted.

In the above univariate and multivariate analyses (Table

3) each of the dietary habits were considered individually. In

our final analyses, we entered all dietary habits and

covariates together into one Cox model for each outcome (Table

4), in order to evaluate the contribution of a specific

dietary habit in the presence of the other dietary habits.

This procedure confirmed the main findings obtained in the

previous models: Frequency of salty food consumption remained

independently associated with an increased risk for high-

urgency transplantation (HR = 2.91, 95% CI: 1.29, 6.60, P =

0.011), regardless of other dietary habits and disease

severity. Frequent intake of foods rich in PUFA+MUFA

significantly reduced the risk of death/deterioration (HR =

0.48, 95% CI: 0.24, 0.95, P = 0.034), independently of other

dietary habits and heart failure severity. The remaining three

dietary habits did not contribute significantly to any of the

outcomes.

Repeating analyses with the unimputed data revealed

results comparable to the ones described above, despite

19

reduced sample size and lower event numbers. The association

of fluid intake/cardiac index with high-urgency

transplantation became significant in univariate and

multivariate models including only this dietary habit (P <

0.01 and P < 0.02).

Discussion

This multi-site prospective study indicates that dietary

habits are related to the prognosis of patients with advanced

heart failure awaiting transplantation. Specifically,

transplant candidates who reported frequent consumption of

salty foods at time of waitlisting had an increased risk for

deterioration of health status as indicated by high-urgency

transplantation. In contrast, more frequent consumption of

foods rich in PUFA+MUFA was independently associated with

decreased risk for death/deterioration. These two findings

were robust across various models and the effects were

independent of heart failure severity, inpatient status, age,

sex, disease duration, BMI, and other dietary habits and

health behaviors.

20

Transplantation in high-urgency status was the most

prevalent outcome observed in our study cohort and occurred

rather early during waiting time. It is conceivable that

frequent consumption of salty foods is associated with an

increased sodium intake and thus contributes to deterioration

of health status. The latter was shown in recent studies of

stable ambulatory heart failure patients.25-27 For example, in

the study by Arcand, patients with a high sodium intake of

>2.7 g/d had an increased risk for acute decompensation

compared to patients with lower salt intake.25 Also, 24-hour

urinary sodium excretion indicating sodium intake >3 g/d was

associated with shorter event-free survival among patients

with NYHA class III/IV.26

It has been suggested that high salt intake might affect

the organism via activation of the sympathetic nervous system

and the renin-angiotensin-aldosterone system, thereby

contributing to aldosterone-mediated vascular damage such as

development of extracellular matrix and fibrosis, oxidative

stress, inflammation, and endothelial dysfunction.28 However,

the impact of sodium intake on these systems remains

controversial in patients with compensated heart failure who

21

receive ACE-inhibitors and beta-blockers.29-31 Yet, a high sodium

intake increases extracellular fluid volume,31 thus

exacerbating heart failure symptoms. Additional analyses in

our study support this finding. Patients who ate salty foods

“often” were more likely to be symptomatic as indicated by

NYHA class IV (30%) than patients who ate salty foods “rarely”

(18.6%; P = 0.050). Such clinical deterioration may enforce the

need for high-urgency listing, and, consequently high-urgency

transplantation, as patients in this status have priority in

organ allocation. Patients thus rather rapidly transplanted

avoid death on the waiting list (only 6 patients died while

upgraded to high-urgency status) and other competing outcomes.

Therefore, it is not surprising that salty food consumption

was not associated with the remaining outcomes.

The relevance of restricting fluid intake in patients with

advanced heart failure remains unclear. Of two studies

concluding that fluid restriction might not be advisable, one

did not include patients in NYHA class IV.32 The second study

reported relatively small amounts of fluid (1466 mL/d±607

mL/d) in 34 patients with free fluid intake.33 In our sample,

13.2% of participants reported to drink >2 L/d; patients with

22

hyponatremia were particularly likely to report excessive

fluid intake. While fluid intake per se was not related to

outcomes, there was some suggestion that fluid intake adjusted

for cardiac index (indicating greater burden on the heart)

might be associated with accelerated high-urgency

transplantation. It is of note that consumption of salty

foods, the most robust dietary habit to predict

transplantation in high-urgency status, was also positively

correlated with fluid intake as well as intake of foods high

in saturated fats. Thus, dietary recommendations targeting all

three behaviors, that is, reduce salt and saturated fat intake

together with fluid intake, may be most beneficial for this

patient population.

Our other robust finding indicated that in patients who

were not transplanted urgently frequent intake of foods rich

in PUFA+MUFA was associated with significantly reduced risk of

death on the waiting list. It is conceivable that this effect

was due to higher amounts of ω-3PUFA contained in fish oils

and vegetable oils such as flaxseed, canola, walnut, or

soybean oil.9 ω-3PUFAs have been linked to anti-arrhythmic

action, lowering of triglyceride levels, anti-inflammatory

23

effects, improved endothelial function, and improved

mitochondrial function leading to increased efficiency of

oxygen use by the heart and skeletal muscles. These effects

have mostly been ascribed to ω-3PUFA of marine origin, but

also to plant-based ω-3PUFA such as α-linolenic acid.

Interestingly, in the GISSI-HF trial, ω-3PUFA supplementation

resulted in small but significant reductions in mortality

compared to a placebo group.36 Furthermore, animal studies

suggest that plant based ω-6PUFA (α-linoleic acid), which is

also contained in ω-3rich soybean, walnut, and canola oil, but

additionally in safflower, grape seed, or sunflower oil, may

have beneficial effects on heart failure via remodeling of

cardiac cardiolipin, an important mitochondrial phospholipid.37

In addition, MUFA and phenolic compounds of olive oil might

contribute to beneficial health effects via reduction of

oxidative stress and inflammation.15 Thus, our findings are in

line with those who suggest that consuming fish and vegetable

oils that are high in PUFA and MUFA might be beneficial in

heart failure.

Interestingly, the beneficial effect of foods rich in

PUFA+MUFA was observed about one year after waitlisting (cf.

24

Figure 3). At this time, the majority of transplantations in

high-urgency status had already taken place. Thus, the

relatively soon occurring of high-urgency transplantation

might have precluded the protective mechanisms of PUFA+MUFA to

develop in these patients. However, our study design does not

allow identifying potential mechanisms underlying these

associations. Hence, it remains unclear if, for example the

detrimental effect of salty/fatty foods and concomitant fluid

intake was stronger than the beneficial effect of foods rich

in PUFA+MUFA. Similarly, different mechanisms might be

involved in the associations of PUFA+MUFA with reduced

death/deterioration on one hand and improvement of health

status on the other hand, particularly as less severe heart

failure (as indicated by HFSS) was predictive of this latter

outcome. Regardless of the above issues, our study clearly

points to dietary habits associated with unhealthy eating

(frequent intake of salty foods, rare intake of foods rich in

PUFA+MUFA) are associated with outcomes reflecting clinical

deterioration, i.e., transplantation in high-urgency status,

death on the waiting list or delisting due to clinical

deterioration.

25

Aside from the robust findings regarding the intake of

salty foods and foods rich in PUFA+MUFA, there was also some

indication for the prognostic relevance of

fruits/vegetables/legumes: frequent intake increased the

chance for clinical improvement and subsequent delisting but

only with additional adjustment for cardiac index. This result

is in line with studies indicating that frequent fruit and

vegetable consumption provides protection from fatal ischemic

heart disease and other non-communicable chronic diseases.

Associations of frequent consumption of

fruits/vegetables/legumes with high-urgency transplantation

were not observed consistently and disappeared when fluid

intake/cardiac index and the other dietary habits were also

considered. It is possible that consumed fruits may have

differed in their water content. Because the assessment of

fruit in this questionnaire did not permit to disentangle the

potentially adverse effects of juicy fruits, which might

contribute to fluid intake, a more detailed assessment of

fruits and vegetables would be advisable in future studies.

Our study has several limitations. First, a food frequency

questionnaire was employed to assess the frequency of foods

26

that were consumed, which does not allow for any conclusions

about actual nutrient intakes or total amount of calories

consumed. However, relationships between eating habits and

measures of health have been shown without assessment of

nutrient intake. For example, improvements in eating habits

(lower cholesterol and saturated fat foods, more fruits,

vegetables, grains, legumes) were associated with significant

plasma cholesterol lowering.40 Another limitation, and a

problem that all measures based on self-reports share, is that

people’s responses may be prone to memory and reporting

biases. However, reasonable correlations among food groups

support the validity of our assessments. For example, frequent

intake of foods high in salt was correlated with frequent

intake of foods high in saturated fats and increased fluid

intake. Also, the positive correlation between frequent intake

of foods high in saturated fats and plasma LDL-cholesterol

levels validates patients' self-reports by a biomarker. The

somewhat unexpected finding that higher BMI scores were

related to less frequent intake of foods high in saturated

fats, but with a more frequent intake of foods rich in PUFAs

and MUFAs, could indicate efforts of overweight and obese

27

patients to improve their eating habits, in an attempt to

reduce their BMI. A high BMI has been related to adverse

outcomes after heart transplantation and, therefore, obese

patients are advised to loose weight.41 This reasoning receives

support from post-hoc analyses of additional data that had

been included in our study. Significantly more overweight and

obese patients reported that they were adhering to a diet

recommended by their physician compared to patients with

normal weight (data not shown, overweight, 61%; obese, 66%;

normal weight, 48%, respectively), and both overweight and

obese patients reported more frequent consumption of low fat,

low caloric foods than patients with normal weight

(overweight, 53%; obese, 71%; normal weight 43%,

respectively). A more detailed dietary assessment including

calories consumed could shed more light on this issue in

future studies.

Another limitation pertains to the low number of women in

our sample (representing the typical proportion of women in

this population) and, as a consequence, the low number of

events in women (18 deaths/delistings due to deterioration, 17

HU-transplantations). This did not allow for an evaluation of

28

interaction effects of dietary habits with gender. Also, some

of the Cox models had fewer than 10 events per predictor.

However, this does not necessarily lead to bias in

multivariate Cox regression.42 Nevertheless, stability of

significant results was confirmed by comparing these results

with those from models excluding irrelevant (i.e., P > 0.50)

predictors,42 and evaluating each dietary habit’s association

with outcomes in the presence of other dietary variables and

control variables. Thus, we consider findings consistently

emerging in all models as our most robust findings, namely the

associations including salty foods and foods rich in

PUFA+MUFA. In addition, the prospective observational design of

our study limits drawing conclusions about cause and effect

relationships. It is possible that other variables not

measured by us (e.g., cachexia) could have affected clinical

outcomes. In spite of the above limitations, it is important

that our findings were obtained by treating outcomes as

competing events rather than combined endpoints. This approach

provides a more differentiated picture of factors that

influence mutually exclusive events. For example, the impact

of salty foods on deterioration of health status leading to

29

subsequent high-urgency transplantation would have been missed

in analyses of transplant-free survival. Thus, our findings

obtained under stringent statistical control of potentially

confounding variables, might stimulate the development of

clinical trials testing the effects of dietary interventions

in heart failure and of experiments elucidating potential

biological mechanisms by which diet can influence disease

progression.

To conclude, dietary habits are of relevance for the

prognosis of patients with severe heart failure awaiting

cardiac transplantation. Specifically, our findings confirm

the importance of limiting intake of foods high in salt and

saturated fats, and suggest that concomitantly monitoring a

patient’s fluid intake may be advisable. In addition, our

results point to the importance of protective dietary factors

(frequent intake of foods high in “good fats”, PUFA+MUFA) for

reducing the risk of death while waiting for a new heart. In

sum, dietary interventions to improve eating habits and

adherence to dietary recommendations may be of great benefit

to this patient population.

30

Disclosures

None declared.

Acknowledgements

We are indebted to Katharina Schury for her assistance in data

collection, to Theresa Rebelein Vina Bunyamin, and Larissa

Urban for their assistance in preparation of data and

manuscript. We also thank the hospitals and the patients for

their participation.

31

References

1. Heidenreich PA, Trogdon JG, Khavjou OA, Butler J, Dracup

K, Ezekowitz MD, et al. Forecasting the future of

cardiovascular disease in the United States: a policy

statement from the American Heart Association. Circulation

2011;123:933-44.

2. Smits JM, Rahmel A, Laufer G, Meiser B. Considerations for

patients awaiting heart transplantation. Thoracic and

Cardiovascular Surgeon 2010;58 Suppl 2:S179-84.

3. Haneya A, Haake N, Diez C, Puehler T, Cremer J, Schmid C,

et al. Impact of the eurotransplant high-urgency heart

allocation system on the outcome of transplant candidates

in Germany. Thorac Cardiovasc Surg 2011;59:93-8.

4. Dickstein K, Cohen-Solal A, Filippatos G, McMurray JJ,

Ponikowski P, Poole-Wilson PA, et al. ESC guidelines for

the diagnosis and treatment of acute and chronic heart

failure 2008: the Task Force for the diagnosis and

treatment of acute and chronic heart failure 2008 of the

European Society of Cardiology. Developed in collaboration

with the Heart Failure Association of the ESC (HFA) and

32

endorsed by the European Society of Intensive Care

Medicine (ESICM). Eur J Heart Fail 2008;10:933-89.

5. Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis GS,

Ganiats TG, et al. 2009 focused update incorporated into

the ACC/AHA 2005 Guidelines for the Diagnosis and

Management of Heart Failure in Adults: a report of the

American College of Cardiology Foundation/American Heart

Association Task Force on Practice Guidelines: developed

in collaboration with the International Society for Heart

and Lung Transplantation. Circulation 2009;119:e391-479.

6. Jessup M, Banner N, Brozena S, Campana C, Costard-Jackle

A, Dengler T, et al. Optimal pharmacologic and non-

pharmacologic management of cardiac transplant candidates:

Approaches to be considered prior to transplant

evaluation: International Society for Heart and Lung

Transplantation guidelines for the care of cardiac

transplant candidates--2006. J Heart Lung Transplant

2006;25:1003-23.

7. Lindenfeld J, Albert NM, Boehmer JP, Collins SP, Ezekowitz

JA, Givertz MM, et al. HFSA 2010 Comprehensive Heart

33

Failure Practice Guideline. Journal of Cardiac Failure

2010;16:e1-194.

8. Lainscak M, Blue L, Clark AL, Dahlstrom U, Dickstein K,

Ekman I, et al. Self-care management of heart failure:

practical recommendations from the Patient Care Committee

of the Heart Failure Association of the European Society

of Cardiology. Eur J Heart Fail 2011;13:115-26.

9. Duda MK, O'Shea KM, Stanley WC. omega-3 polyunsaturated

fatty acid supplementation for the treatment of heart

failure: mechanisms and clinical potential. Cardiovasc Res

2009;84:33-41.

10. Djousse L, Driver JA, Gaziano JM. Relation between

modifiable lifestyle factors and lifetime risk of heart

failure. JAMA 2009;302:394-400.

11. Payne-Emerson H, Lennie TA. Nutritional considerations in

heart failure. Nurs Clin North Am 2008;43:117-32.

12. Spaderna H, Weidner G, Zahn D, Smits JMA. Psychological

characteristics and social integration of patients with

ischemic and non-ischemic heart failure newly listed for

heart transplantation: The Waiting for a New Heart Study.

Appl Psychol Health Wellbeing 2009;1:188-210.

34

13. Spaderna H, Mendell NR, Zahn D, Wang Y, Kahn J, Smits JM,

et al. Social isolation and depression predict 12-month

outcomes in the "waiting for a new heart study". J Heart

Lung Transplant 2010;29:247-54.

14. Dlugosch GE, Krieger W. Fragebogen zur Erfassung des

Gesundheitsverhaltens (FEG) [Questionnaire for the

assessment of health behavior]. Frankfurt: Swets Test

Services; 1995.

15. Covas MI, Konstantinidou V, Fito M. Olive oil and

cardiovascular health. J Cardiovasc Pharmacol 2009;54:477-

82.

16. Arzneimittelkommission der deutschen Ärzteschaft.

Chronische Herzinsuffizienz [Chronic Heart Failure]. 3.

ed. Berlin 2007.

17. Lund LH, Aaronson KD, Mancini DM. Predicting survival in

ambulatory patients with severe heart failure on beta-

blocker therapy. Am J Cardiol 2003;92:1350-4.

18. Zahn D, Weidner G, Beyersmann J, Smits JMA, Deng MC,

Kaczmarek I, et al. Combined risk scores and depression as

predictors for competing waiting-list outcomes in the

35

Waiting for a New Heart Study. Transpl Int 2010;23:1223-

32.

19. Spaderna H, Zahn D, Schulze Schleithoff S, Stadlbauer T,

Rupprecht L, Smits JMA, et al. Depression and disease

severity as correlates of everyday physical activity in

heart transplant candidates. Transpl Int 2010;23:813-22.

20. Stewart AL, Mills KM, King AC, Haskell WL, Gillies D,

Ritter PL. CHAMPS Physical Activity Questionnaire for

Older Adults: Outcomes for interventions. Med Sci Sports

Exerc 2001;33:1126-41.

21. Sterne JA, White IR, Carlin JB, Spratt M, Royston P,

Kenward MG, et al. Multiple imputation for missing data in

epidemiological and clinical research: potential and

pitfalls. BMJ 2009;338:b2393.

22. Van Buuren S. Multiple imputation of discrete and

continuous data by fully conditional specification. Stat

Methods Med Res 2007;16:219-42.

23. Bailey RC, Jia-Yeong Lin M, Krakauer H. Time-to-event

modeling of competing risks with intervening states in

transplantation. Am J Transplant 2003;3:192-202.

36

24. Beyersmann J, Dettenkofer M, Bertz H, Schumacher M. A

competing risks analysis of bloodstream infection after

stem-cell transplantation using subdistribution hazards

and cause-specific hazards. Stat Med 2007;26:5360-9.

25. Arcand J, Ivanov J, Sasson A, Floras V, Al-Hesayen A,

Azevedo ER, et al. A high-sodium diet is associated with

acute decompensated heart failure in ambulatory heart

failure patients: a prospective follow-up study. Am J Clin

Nutr 2011;93:332-7.

26. Lennie TA, Song EK, Wu JR, Chung ML, Dunbar SB, Pressler

SJ, et al. Three gram sodium intake is associated with

longer event-free survival only in patients with advanced

heart failure. Journal of Cardiac Failure 2011;17:325-30.

27. Paterna S, Parrinello G, Cannizzaro S, Fasullo S, Torres

D, Sarullo FM, et al. Medium term effects of different

dosage of diuretic, sodium, and fluid administration on

neurohormonal and clinical outcome in patients with

recently compensated heart failure. Am J Cardiol

2009;103:93-102.

28. Lastra G, Dhuper S, Johnson MS, Sowers JR. Salt,

aldosterone, and insulin resistance: impact on the

37

cardiovascular system. Nat Rev Cardiol 2010;7:577-84.

29. Beich KR, Yancy C. The heart failure and sodium

restriction controversy: challenging conventional

practice. Nutr Clin Pract 2008;23:477-86.

30. Alvelos M, Ferreira A, Bettencourt P, Serrao P, Pestana M,

Cerqueira-Gomes M, et al. The effect of dietary sodium

restriction on neurohumoral activity and renal

dopaminergic response in patients with heart failure. Eur

J Heart Fail 2004;6:593-9.

31. Damgaard M, Norsk P, Gustafsson F, Kanters JK, Christensen

NJ, Bie P, et al. Hemodynamic and neuroendocrine responses

to changes in sodium intake in compensated heart failure.

Am J Physiol Regul Integr Comp Physiol 2006;290:R1294-301.

32. Holst M, Stromberg A, Lindholm M, Willenheimer R. Liberal

versus restricted fluid prescription in stabilised

patients with chronic heart failure: result of a

randomised cross-over study of the effects on health-

related quality of life, physical capacity, thirst and

morbidity. Scand Cardiovasc J 2008;42:316-22.

33. Travers B, O'Loughlin C, Murphy NF, Ryder M, Conlon C,

Ledwidge M, et al. Fluid restriction in the management of

38

decompensated heart failure: no impact on time to clinical

stability. J Cardiac Fail 2007;13:128-32.

34. Mehra MR, Lavie CJ, Ventura HO, Milani RV. Fish oils

produce anti-inflammatory effects and improve body weight

in severe heart failure. J Heart Lung Transplant

2006;25:834-8.

35. Kromhout D, Giltay EJ, Geleijnse JM, Group* ftAOT. n–3

Fatty acids and cardiovascular events after myocardial

infarction. N Engl J Med 2010;363:2015-26.

36. Tavazzi L, Maggioni AP, Marchioli R, Barlera S, Franzosi

MG, Latini R, et al. Effect of n-3 polyunsaturated fatty

acids in patients with chronic heart failure (the GISSI-HF

trial): a randomised, double-blind, placebo-controlled

trial. Lancet 2008;372:1223-30.

37. Sparagna GC, Lesnefsky EJ. Cardiolipin remodeling in the

heart. J Cardiovasc Pharmacol 2009;53:290-301.

38. Crowe FL, Roddam AW, Key TJ, Appleby PN, Overvad K,

Jakobsen MU, et al. Fruit and vegetable intake and

mortality from ischaemic heart disease: results from the

European Prospective Investigation into Cancer and

Nutrition (EPIC)-Heart study. Eur Heart J 2011;32:1235-43.

39

39. Weidner G. Sustainability in medicine - a case for the

prevention of chronic non-communicable diseases.

Environmentalist 2012;32:353-359.

40. Connor SL, Gustafson JR, Sexton G, Becker N, Artaud-Wild

S, Connor WE. The Diet Habit Survey: A new method of

dietary assessment that relates to plasma cholesterol

changes. J Am Diet Assoc 1992;92:41-7.

41. Mehra MR, Kobashigawa J, Starling R, Russell S, Uber PA,

Parameshwar J, et al. Listing Criteria for Heart

Transplantation: International Society for Heart and Lung

Transplantation Guidelines for the Care of Cardiac

Transplant Candidates—2006; J Heart Lung Transplant

2006;25:1024-42.

42. Vittinghoff E, McCulloch CE. Relaxing the rule of ten

events per variable in logistic and Cox regression. Am J

Epidemiol 2007;165:710-8.

40

Appendix

The Waiting for a New Heart Study Group consists of the

following sites and investigators: Eurotransplant

International Foundation: Dr. Smits, Dr. Rahmel, Prof. Dr.

Meiser; Med. Klinik I/Kardiologie, Pneumologie, Angiologie

Universitätsklinikum Aachen: Prof. Dr. Kelm, Dr. Koch; Herz-

Zentrum Bad Krozingen: Prof. Dr. Neumann, Dr. Zeh; Herz- und

Diabeteszentrum Nordrhein-Westfalen Bad Oeynhausen: Prof. Dr.

Körfer, (now Prof. Dr. Gummert), Prof. Dr.Zittermann;

Herzzentrum Dresden: Prof. Dr. Strasser, Dr. Thoms;

Herzchirurgische Klinik der Universitätsklinik Erlangen: Prof.

Dr. Weyand, Dr. Tandler; Med. Klinik III/Kardiologie Klinikum

der Universität Frankfurt: Prof. Dr. Zeiher, Dr. Seeger;

Klinik für Thorax-, Herz-, und Gefäßchirurgie des Klinikums

Fulda: Dr. Dörge; Abt. Kardiologie, Universitätsklinikum

Gießen und Marburg, Standort Gießen, Dr. Heidt, Dr.

Stadlbauer; Klinik für Chirurgie der Medizinischen Universität

Graz: Prof. Dr. Tscheliessnigg, Dr. Kahn; Universitätsklinik

und Poliklinik für Herz- und Thoraxchirurgie Halle-Wittenberg:

Prof. Dr. Silber, Dr. Hofmann; Universitäres Herzzentrum

Hamburg GmbH: Prof. Dr. Dr. Reichenspurner, Dr. Meffert, Dr.

41

Wagner; Klinik für Herz-, Thorax- und Gefäßchirurgie des

Universitätsklinikums Jena: Prof. Dr. Gummert, Dr. Malessa,

Dr. Tigges-Limmer; Klinik und Poliklinik für Herz- und

Thoraxchirurgie der Universität zu Köln: Dr. Müller-Ehmsen;

Klinik für Herzchirurgie des Herzzentrums Leipzig GmbH: Prof.

Dr. Mohr, Dr. Doll; II. Medizinische Klinik und Poliklinik

Universitätsmedizin Mainz: Prof. Dr. Münzel, Dr. Hink;

Herzchirurgische Klinik der Universität München: Prof. Dr.

Reichart, Dr. Kaczmarek; Klinik und Poliklinik für Herz-,

Thorax- und herznahe Gefäßchirurgie der Universität

Regensburg: Prof. Dr. Birnbaum, Dr. Rupprecht (now Prof. Dr.

Schmid).

42

Table 1. Patient Characteristics

Age 53.1 ± 11.1

Women (%) 58 (18.2)

Married (%) 212 (66.7)

BMI (kg/m2) 25.9 ± 4.0

Physical activity (kcal/week), median (IQR)1177 (1392 – 2931)

Frequency of alcohol consumption1, median (IQR)1.2 (1.0 – 1.6)

Former/current smoker2 (n = 316; %) 240 (76.0)

Inpatients (%) 87 (27.4)

Ischemic diagnosis3 (%) 122 (38.4)

NYHA class (n = 316; %)

II, II-III, III 125 (39.6)

III-IV 114 (36.1)

IV 77 (24.4)

QRS > 0.12 sec3 (n = 301; %) 161 (53.5)

Peak oxygen consumption3 (mL/min/kg; n = 239) 11.1 ± 3.0

Cardiac index (L/min/m²; n = 289), median (IQR)2.0 (1.7 – 2.3)

Ejection fraction3 (%; n = 312), median (IQR)21.5 (15.3 – 28.0)

Heart rate3 (beats/min; n = 316), median (IQR)75 (65 – 86)

Mean arterial blood pressure3 (mmHg; n = 316), median (IQR) 76.7

(70.0 – 84.0)

Sodium3 (mmol/L; n = 316), median (IQR) 137.5 (135.0 – 140.0)

Creatinine (mg/dl; n = 301), median (IQR) 1.3 (1.1 – 1.6)

43

Serum total cholesterol (mg/dl; n = 218) 172.3 ± 52.1

LDL cholesterol (mg/dl; n = 77) 112.8 ± 36.1

HFSS (n = 224), median (IQR) 7.8 (7.2 – 8.4)

To be continued.

Table 1. Continued.

Diabetes mellitus (n = 288; %) 75 (26.0)

Previous heart surgery (n = 290; %) 95 (32.8)

Implantable cardioverter defibrillator (n = 279; %)173 (62.0)

Diuretics (n = 312; %) 279 (89.4)

Beta-blockers (n = 313; %) 272 (86.9)

ACE Inhibitors (n = 312; %) 237 (76.0)

Anticoagulation drugs (n = 288; %) 226 (78.5)

Aldosterone antagonists (n = 312; %) 208 (66.7)

Digitalis (n = 313; %) 154 (49.2)

Antiarrhythmics (n = 287; %) 81 (28.2)

Catecholamines (n = 309; %) 49 (15.9)

Notes. Values are number and percentage (%), mean ± standard

deviation, or median and IQR, interquartile range; HFSS, Heart

Failure Survival Score. Lower scores denote an increased medical

risk; ACE Inhibitors, angiotensin-converting enzyme inhibitors and

AT1 receptor blockers.

44

1Scores for frequency of alcohol consumption denote the mean

frequency (1 = “never” to 4 = “daily”) of 5 alcoholic beverages

divided by 5; higher values indicate higher consumption frequencies.

2Twelve patients reported to be current smokers.

3Included in the Heart Failure Survival Score.

45

Table 2. Associations Among Dietary Habits and of Dietary Habits

with Demographic and Medical Characteristics.

Salty Saturated PUFA+ Fruit/ Fluid Fluid

foods fats MUFA Vegetables/ intake

intake/car-

Legumes diac index

Salty foods – 0.66**** –0.09 0.02 0.22***

0.17**

Saturated fats – –0.32****–0.01 0.16** 0.17**

PUFA+MUFA – 0.25**** –0.05

–0.06

Fruit/Vegetables/Legumes – –0.07

0.05

Fluid intake – 0.60**

Age –0.20*** –0.18** 0.14* 0.09 –0.19*** –0.11

Sex1 –0.08 –0.12* –0.03 0.14* –0.22***–

0.12*

BMI –0.14* –0.25*** 0.16** –0.06 0.13* 0.10

Inpatient1 0.10 0.17** –0.06 0.12 –0.05 0.04

HFSS –0.06 –0.03 0.05 –0.01 –0.10 –0.13*

Cardiac index –0.08 –0.10 0.03 –0.14* 0.03–

0.65****

Diabetes1 –0.11* –0.22*** 0.12** –0.03 0.06 0.10

46

Previous heart surgery1 –0.13* –0.17* 0.09 0.03 –0.05

–0.16*

ICDa –0.06 –0.02 0.13* 0.12 –0.01 –0.01

LDL cholesterol2 0.19 0.25* –0.06 0.05 0.11 –0.02

Notes. N = 318. PUFA+MUFA, poly-/monounsaturated fats; HFSS, Heart

Failure Survival Score; ICD, intracardiac cardioverter-

defibrillator.

1Men are coded 0, women 1. For all other categorical variables 0 =

no, 1 = yes. Coefficients are point biserial correlations.

2Pearson correlations among patients without lipid lowering

medication (n = 77).

*P < 0.05; **P < 0.01; ***P < 0.001; **** P < 0.0001.

47

Table 3. Outcome-specific Hazard Ratios of Waiting List Outcomes Associated with Dietary Habits in 318 NewlyListed Heart Transplant Candidates.

Death/ High-urgency Elective Improvement

Deterioration Transplantation Transplantation

HR (95% CI) P HR (95% CI) P HR (95% CI) P HR (95% CI) P

Salty foods

Muniv 1.09 (0.53, 2.26) 0.813 3.34 (1.84, 6.08) <0.0001 1.35 (0.52, 3.48)

0.536 1.33 (0.44, 4.00) 0.617

Madj 1.72 (0.78, 3.80) 0.180 2.90 (1.55, 5.42)1 <0.001 1.59 (0.56, 4.49)

0.382 0.74 (0.20, 2.69) 0.647

Saturated fats

Muniv 1.21 (0.61, 2.41) 0.587 2.82 (1.66, 4.78) <0.001 2.87 (1.21, 6.79)

0.017 0.88 (0.31, 2.50) 0.812

Madj 1.58 (0.72, 3.46) 0.250 1.67 (0.93, 3.01) 0.088 2.29 (0.89, 5.91)

0.086 0.52 (0.14, 1.86) 0.312

PUFA+MUFA

Muniv 0.50 (0.29, 0.85) 0.011 0.89 (0.57, 1.38) 0.590 0.49 (0.24, 0.99)

0.046 1.05 (0.44, 2.52) 0.914

48

Madj 0.49 (0.26, 0.92) 0.0262 1.17 (0.75, 1.85) 0.488 0.50 (0.23, 1.11)

0.090 1.07 (0.42, 2.71) 0.889

Fruits/vegetables/legumes

Muniv 0.81 (0.44, 1.51) 0.511 1.56 (0.93, 2.60) 0.092 1.74 (0.74, 4.10)

0.204 1.92 (0.68, 5.44) 0.221

Madj 0.83 (0.46, 1.52) 0.551 1.77 (1.06, 2.97) 0.0303 1.82 (0.78, 4.23)

0.167 3.44 (1.00, 11.78) 0.0503

Fluid intake/cardiac index

Muniv 1.10 (0.71, 1.69) 0.675 1.39 (0.99, 1.97) 0.064 0.60 (0.31, 1.14)

0.117 0.84 (0.33, 2.10) 0.704

Madj 1.20 (0.78, 1.86) 0.404 1.38 (0.96, 1.99) 0.089 0.73 (0.39, 1.37)

0.327 0.93 (0.37, 2.31) 0.874

49

Notes. HR, hazard ratio. CI, confidence interval. PUFA+MUFA, poly-

and monounsaturated fatty acids. Muniv, unadjusted model. Madj,

adjusted for age, sex, disease duration, body mass index, Heart

Failure Survival Score, creatinine, inpatient status, physical

activity, smoking, frequency of alcohol consumption.

1Additional adjustment for diabetes and previous heart surgery: P <

0.001.

2Additional adjustment for diabetes and implantable cardioverter

defibrillator: P < 0.05.

3Additional adjustment for cardiac index: P > 0.09 for high-urgency

transplantation; P < 0.05 for delisting due to improvement.

50

Table 4. Outcome-specific hazard ratios of waiting list outcomes associated with demographic

characteristics, medical variables and all dietary habits simultaneously.

Death/ High-urgency Elective Delisting due to

Deterioration1 Transplantation1 Transplantation1 Improvement1

HR (95% CI) P HR (95% CI) P HR (95% CI) P HR (95% CI) P

Demographic characteristics and BMI

Age (years) 1.01 (0.99, 1.05) 0.323 0.99 (0.97, 1.01) 0.308 1.01 (0.98, 1.04) 0.626 0.98

(0.94, 1.02) 0.260

Female sex2 2.30 (1.18, 4.47) 0.014 0.96 (0.54, 1.73) 0.903 1.33 (0.56, 3.23) 0.503

1.49 (0.44, 5.04) 0.523

BMI (kg/m2) 0.97 (0.91, 1.04) 0.436 0.97 (0.92, 1.02) 0.247 0.90 (0.80, 0.99) 0.023

1.00 (0.90, 1.10) 0.960

Medical variables indicating heart failure severity

HFSS 0.65 (0.49, 0.86) 0.003 0.80 (0.64, 1.01) 0.059 0.96 (0.69, 1.34) 0.814

1.87 (1.24, 2.81) 0.003

Creatinine (mg/dl) 2.23 (1.42, 3.51) <0.001 1.55 (1.01, 2.38) 0.048 1.24 (0.61, 2.49) 0.554

0.76 (0.25, 2.30) 0.623

51

Inpatient status3 1.31 (0.67, 2.55) 0.435 3.87 (2.52, 5.94) <0.001 3.36 (1.63, 6.93)0.001

1.02 (0.27, 3.86) 0.979

Dietary habits4

Salty foods 1.85 (0.65, 1.66) 0.247 2.91 (1.29, 6.60) 0.011 1.18 (0.29, 4.84) 0.814

1.15 (0.23, 5.64) 0.864

Saturated fats 0.80 (0.27, 2.33) 0.680 0.89 (0.39, 2.01) 0.772 1.89 (0.52, 6.89) 0.336

0.55 (0.11, 2.75) 0.465

PUFA+MUFA 0.48 (0.24, 0.95) 0.034 1.04 (0.63, 1.73) 0.873 0.50 (0.21, 1.21) 0.123

0.77 (0.28, 2.16) 0.624

Fruits/vegetables/legumes 1.15 (0.74, 1.78) 0.741 1.67 (0.98, 2.85) 0.058 2.09 (0.88, 4.98)0.095

3.40 (0.94, 12.31) 0.062

Fluid intake/cardiac index 1.15 (0.74, 1.78) 0.549 1.26 (0.87, 1.82) 0.223 0.70 (0.36, 1.36)

0.295 1.00 (0.40, 2.52) 0.995

Notes. HR, hazard ratio. CI, confidence interval. HFSS, Heart Failure Survival Score. PUFA+MUFA, poly- and

monounsaturated fatty acids.

52

1The models were also adjusted for potentially confounding health

behaviors (smoking, physical activity, frequency of alcohol

consumption).

2Men are coded 0, women 1.

3Outpatients are coded 0, inpatients 1.

4Dietary habits are consumption frequencies except for fluid

intake/cardiac index.

53

Legends

Figure 1.

Dietary habits at time of listing in 318 newly listed heart

transplant candidates. Panel A: Bars denote mean values (M) and

standard deviations (SD) of single food items and dietary habit sum

scores. PUFA+MUFA, poly- and monounsaturated fats. Panel B:

Distribution of patients by daily fluid intake.

Figure 2.

N = 318. Cumulative incidence functions illustrating the probability

for competing waiting list outcomes in the group with rare

consumption of salty foods (below the median of consumption

frequency of salty foods; “rarely”) and the group with frequent

consumption of salty foods (above the median; “often”). Mean

consumption frequency of salty foods with standard deviation: Mrarely =

1.79 0.21 versus Moften = 2.34 0.19, P < 0.001.

Figure 3.

N = 318. Cumulative incidence functions illustrating the probability

for competing waiting list outcomes in the group with rare

consumption of foods rich in poly- and monounsaturated fats

(PUFA+MUFA, below the median of PUFA+MUFA consumption frequency;

“rarely”) and the group with frequent consumption of foods rich in

PUFA+MUFA (above the median of PUFA+MUFA consumption frequency;

54

“often”). Mean consumption frequency of PUFA+MUFA with standard

deviation: Mrarely = 1.98 0.29 versus Moften = 2.68 0.20, P < 0.001.

55