University of Groningen

Chronic heart failure diagnostics and application of neuropeptides in residential elderlyBarents, Maaike

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2015

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Barents, M. (2015). Chronic heart failure diagnostics and application of neuropeptides in residential elderly.[Groningen]: University of Groningen.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 15-09-2020

Chronic heart failure diagnostics and application of neuropeptides

in residential elderly

Maaike Barents

geboren 27 december 1950

te Utrecht

2

Paranimfen:

Els Barents

Christien Enzing

Barents, Maaike

Chronic heart failure diagnostics and application of neuropeptides in residential elderly

Dissertation University of Groningen, The Netherlands – with references – with summary in

English and Dutch.

book: ISBN: 978-90-367-8323-1

Ebook : ISBN: 978-90-367-8322-4

PDF without DRM

All rights reserved. No part of this publication may be reproduced, stored in retrieval system, or

transmitted, in any form or by any means, electronically, mechanically, by photocopying, or oth-

erwise, without the prior written permission of the author.

Cover design: Simone Koster, simone@anderkaliber.nl

Druk & lay-out: Gildeprint Drukkerijen B.V., Enschede www.gildeprint.nl

3

Chronic heart failure diagnostics and applica-tion of neuropeptides in residential elderly

Proefschrift

ter verkrijging van de graad van doctor aan de

Rijksuniversiteit Groningen

op gezag van de

rector magnificus prof. dr. E. Sterken

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

woensdag 23 december 2015 om 16:15 uur

door

Maaike Barents geboren op 27 december 1950

te Utrecht

4

Promotores

Prof. F.A.J. Muskiet

Prof. J.L. Hillege

Copromotor

Dr. M.J.L. de Jongste

Beoordelingscommissie

Prof. A.W. Hoes

Prof. W.P.A. Achterberg

Prof. A.A. Voors

Financial support by the Dutch Heart Foundation for the publication of this study was

gratefully acknowledged.

5

We recognize thankfully the support for this thesis given by Corparis Welzijn met Zorg, www.corparis.nl.

6

7

CONTENTS List of abbreviations 10

Chapter 1 15

General introduction to heart failure 1.2 Rationale 15

1.3 Mortality 16

1.4 Morbidity 17

1.5 Defining heart failure and its causes 18

1.6 Prevalence of heart failure (HF) 20

1.7 HF disease burden (DALY) 20

1.8 Quality of life 21

1.9 Diagnostics and algorithm 22

1.10 Natriuretic peptides 23

1.11 Therapies for heart failure 24

1.12 Residential elderly 27

1.13 Heart failure in residential elderly 27

1.14 Aim and outline of the thesis 29

Chapter 2 Diagnostics and prevalence of heart failure in residential elderly. Chapter 2.1 41

Prevalence and misdiagnosis of chronic heart failure in nursing home

residents: the role of B-type natriuretic peptides. Barents M, van der Horst IC, Voors AA, Hillege JL, Muskiet FA, de Jongste MJ.

Neth Heart J 2008;16(4):123-8.

8

Chapter 2.2 61 Underestimated prevalence of chronic heart failure among the nursing

home residents in Aruba. Stoutjesdijk E, Brouns RM, Barents M, de Jongste MJ, Besselink HJ, Cheng JD,

Wever R, Muskiet FA.

West Indian Med J 2014;63(6):610-5. Chapter 2.3 79 Validation of neuropeptides in stable heart failure patients. Patients with stable chronic heart failure exhibit also high intraindividual

Biological variation of amino-terminal pro-brain natriuretic peptide (NT-proBNP)

in urine when compared to the variation in plasma.

Schimmel AM, Barents M, de Jongste MJ, Römer JW, Steward HN, Muskiet FA.

Accepted for publication in part in Clinical Chemistry.

Chapter 3 111 BNP and NT-proBNP, predictors of 1-year mortality in nursing home residents. Barents M, Hillege HH, van der Horst IC, de Boer RA, Koster J, Muskiet FA, de Jongste MJ. J Am Med Dir Assoc 2008;9(8):580-5. Chapter 4 129 Chronic heart failure, related to help with activities of daily living in nursing

home residents. Barents M, Hillege HH, Muskiet FA, de Jongste MJ.

Act Adapt Aging 2011;35(2):98-110.

9

Chapter 5 151 Physical exercise performed by residential elderly of high age is safe

and does not affect chronic heart failure symptoms, NT-proBNP

and cardiovascular-metabolic indices: a pilot, single-blinded RCT. Barents M, Weening E, Muskiet FA, Hillege HH, de Jongste MJ. Submitted.

Chapter 6 173 Chapter 6.1 Summary 174 Chapter 6.2 Discussion 185 Chapter 6.3 Recommendations and future perspectives 191 Chapter 7 199 Chapter 7.1 Samenvatting 200

Chapter 7.2 Discussie 211

Chapter 7.3 Aanbevelingen en toekomst perspectieven 217

Dankwoord en CV 221

10

ABBREVIATIONS AND DEFINITIONS ACE-i angiotensin-converting enzyme inhibitor

ADL activities of daily living

Advance care planning aims to help patients, establish decisions about future care that

take effect when they lose capacity

AF atrial fibrillation

ARB angiotensin-2 receptor blocker

AWBZ Algemene Wet Bijzondere Ziektekosten, funding source of elderly care

medicine except geriatric rehabilitation (Ziektekosten Wet)

BNP B-type natriuretic peptide or brain natriuretic peptide

CHF chronic heart failure. Actually, “non-acute” heart failure is a better term

because a physician does not wait to diagnose heart failure till it has

become chronic” (three months). But, we have used “chronic” heart

failure in all publications and decided therefore not to replace the term

comorbidity diseases of which a person suffers, near the disease which is focussed at.

Comorbidity is standardised by the Charlson comorbidity index. This com-

bined age and comorbidity is developed to estimate the relative risk of

death from prognostic clinical variables. In many studies, the index is ap-

plied to correct for existing comorbidity (1)

COPD chronic obstructive pulmonary diseases

CRT-P/CRT-D chronic resynchronization therapy-pacemaker/defibrillator

CVD cardiovascular diseases

CVA cerebrovascular accident

CVi intraindividual biological variation

CVRM guideline cardiovascular risk management guideline (2)

11

DALY disability adjusted life years is a measure of overall disease burden, ex-

pressed as the number of years lost due to ill-health, disability or early

death

ECG electrocardiogram

echo echocardiogram

eGFR estimated glomerular filtration rate

elderly persons of 65 years and older; also older persons elderly care physi-

cian the elderly care physician is a physician who specializes in long-

term care for frail elderly people and chronic patients with complex

health problems (3)

ESC European Society of Cardiology

GP general practitioner

HbA1c HbA1c ≤8,5% = GTA (Gron. Transmurale Afspraak ) T2DM

HDL-C high-density lipoprotein cholesterol; HDL-C ≥ 1 mmol/L (39 mg/dL) ma-

les, HDL-C ≥1,2 mmol/L (46 mg/dL) females

HF heart failure

HFpEF heart failure with preserved ejection fraction

HFrEF heart failure with reduced ejection fraction

Hg mercury

ICD implantable cardioverter defibrillator

IR insulin resistance

Lab Noord Laboratorium Noord (Laboratory North) in The Netherlands

LDL-C low-density lipoprotein cholesterol; LDL-C < 2,5mmol/L (100 mg/dL)

LTCFs long-term care facilities comprising nursing homes and care homes for the

elderly

LVEF left ventricular ejection fraction

12

MDS minimum data set, items of the Resident Assistant Instrument (RAI).

MDS-items consist of systematic observations of dependency and de

livered aid through which dependence and aid can be quantified.

These observations focus on self-reliance of the individual

residential elderly during predefined activities of daily living.

MET Metabolic Equivalent Task (MET), an unit for metabolism in rest

METS metabolic syndrome

MeSH medical subject headings; controlled key words of MEDLINE database

MI myocardial infarction

MMSE minimum mental state estimation

multi medication use a daily intake of at least five different drugs (4)

multimorbidity multiple co-occurring, chronic or long-term diseases or conditions, none

considered as index disease (5)

MSD musculoskeletal disorders

NICE guideline National Institute for Health and Care Excellence

NH nursing home

NLM database National Library of Medicine database (Maryland, USA)

NND neurodegenerative disorders

norm physical activity a minimum of physical activity at ≥4 METs for individuals of 20-55

years during 30 min at 5 days a week and ≥3 METs for 55+ years (6)

NP natriuretic peptide

NT-proBNP amino-terminal proB-type natriuretic peptide

NYHA class New York Heart Association classification of heart failure

oldest old persons persons aged > 85 years (medicaldictionary.thefreedictionary.com)

PMC database Pub Med Central database

polypharmacy polypharmacy is defined as the use of five or more drugs, including

prescribed, over-the-counter, and complementary medicines (4)

13

primary care physicians physicians, general practitioners and elderly care physicians

Quol quality of life

RAAS renin-angiotensin-aldosterone system

RAI Resident’s Assessment Instruments, software program for

providing problem lists as base for the individual care-treatment

plan made by the direct care-giver and for the State Governors (7)

RCT randomized controlled trial

RIVM Rijks Instituut voor Volksgezondheid en Milieu

RD renal dysfunction

residential elderly elderly residing in nursing homes or care homes; also

institutionalised frail elderly

TIA transient ischemic attack

TMWD ten meter walk distance

TUG timed up and go test

TG triglycerides; fasting TG ≤8mmol/L

T2DM type 2 diabetes mellitus

Verenso vereniging van specialisten ouderengeneeskunde en sociaal geriaters (So-

ciety of elderly care physicians and social geriatricians)

VZH verzorgingshuis, care home for the elderly. One of the institutes of

LTCFs

14

15

Chapter 1 GENERAL INTRODUCTION.

This chapter provides a comprehensive review of heart failure (HF), which is the central theme

of this thesis. Since HF is a syndrome predominantly affecting the elderly, it imposes an impor-

tant disease burden in elderly medicine. Elderly medicine is a relatively young specialism. On a

daily basis, elderly care physicians deal with the question when and how to apply general me-

dicinal guidelines and in which cases the resident’s high age has to be taken into account. Is-

sues as diagnostics and exercise are relevant for the frailest elderly with HF. Due to the com-

plexity of frail elderly, optimal diagnosis and treatment, remain a major challenge for the practic-

ing physician in order to improve outcome. At the end of this chapter we focus on HF in the

frailest elderly, i.e. those who have to rely on long-term care facilities (LTCFs), since heart fail-

ure in the frailest elderly is the subject of this thesis.

1.2 RATIONALE.

In The Netherlands, the average life expectancy has increased from 70 years for men and 73

years for women in 1950 to 79 and 83 years in 2013, respectively (www.cbs.nl and

www.nationaalkompas.nl). Main determinants of the increased life expectancy are the imple-

mentation of better hygienic and preventive measures, easy availability of sufficient and healthy

food in conjunction with a shrewd and balanced lifestyle and the development of an armamen-

tarium of medical treatments, such as pharmacotherapy and surgical techniques. One of the

most important achievements in reducing overall mortality originates from the enormous pro-

gress made in the management of cardiovascular disease (CVD) during the last 60 years. As a

consequence the many survivors of CVD grow older and acquire other non-CV diseases, single

or multiple, often accompanied by disabilities. When multimorbidity is accompanied by the loss

of psychosocial resources, we speak of frailty (8). Frail elderly are at risk to deteriorate even

16

faster and geriatric interventions are required. The frailest elderly live in LTCFs, like nursing

homes or care homes, and are referred to as ’residential elderly’. Although this group consumes

a growing percentage of the Dutch health care budget they are excluded from research studies

because of high age, multimorbidity and cognitive disorders. Until recently, these elderly were

diagnosed and treated in the same way as adults with a single disease. At the start of this study

there was no HF guideline available for the group of residential elderly. This was because the

HF guideline of Verenso from 1999 was out of date (Verenso, Vereniging van specialisten

ouderengeneeskunde en sociaal geriaters). In 2003 and in 2008, the Dutch multidisciplinary

HF guideline, based on the European Society of Cardiology (ESC) HF guideline, recognised the

lack of research data for females and the 75+ age-group (both genders) (9;10). Meanwhile,

natriuretic peptides (NP) proved their utility for HF diagnostics, prognosis and therapy-

optimization. In addition the importance of non-pharmacological treatment such as physical ex-

ercise was well recognized. These niches of knowledge on CHF diagnostics and non-

pharmacological therapy formed the rationale for studying this excluded group of the residential

elderly.

1.3 MORTALITY FROM CARDIOVASCULAR DISEASES.

Since its peak in the late 1950s (for women) and the early 1970s (for men) cardiovascular

(CV) mortality has gradually declined. This tremendous decrease in CV mortality is remarkable

when we take into account the increasing number of elderly people and the more advanced age

of contemporary Dutch citizens. Also, after correction for age (standardisation), the drop in CV

mortality remains considerable, with a reduction of about 70% (Figure 1).

17

Figure 1. Standardised Cardiovascular Mortality per 100,000 inhabitants in The Netherlands

(8).

1.4 MORBIDITY.

As spin-off of the impressive 68% decrease in CVD mortality, a substantial increase in CVD

morbidity is nowadays reported, with an estimated prevalence of 700,000 subjects in The

Netherlands currently suffering from CVD. Out of these 700,000, around 125,000 have HF

(11). As for CV mortality to date more patients die from HF than from myocardial infarction (MI), since an increasing number of people survive their MI, but subsequently develop HF (Figure 2).

18

Figure 2. Heart failure mortality has surpassed myocardial infarction mortality in The Neth-

erlands since 2011. Number of deaths attributed to myocardial infarction (MI) or heart failure

(HF) as the primary cause of death in The Netherlands from 1980 through 2012 (12).

1.5 DEFINING HEART FAILURE AND ITS CAUSES.

Defining HF is not easy. The most straightforward approach is to define HF as: a condition dur-

ing which the heart is unable to meet the metabolic needs of the body. Unfortunately in daily

practice HF is often more difficult to define. Also, this is caused by the multi-interpretability of

vague key symptoms of HF, notably, fatigue and dyspnea. The significance of symptoms and

signs is even more difficult to interpret in elderly due to the presence of co-morbidities. At pre-

sent three definitions are used to label HF. These definitions describe the diagnosis of HF on a

clinical (symptomatic), a pathophysiological or a therapeutic basis (9;13;14). Given the difficul-ties to define HF, discovery of (the onset of) HF is often late.

Clinical (symptomatic) HF definition: HF is a syndrome in which the patients should have

the following symptoms typical for HF: shortness of breath at rest or during exertion,

0

5000

10000

15000

20000

25000

1980 1985 1990 1995 2000 2005 2010

Myocardial infarction Heart failure

19

and/or fatigue; signs of fluid retention such as pulmonary congestion or ankle swelling;

and objective evidence of an abnormality of the structure or function of the heart at rest

(9). Pathophysiological HF definition: HF is defined as a pathophysiological state in which an

abnormality of cardiac function is responsible for failure of the heart to pump blood at a

rate commensurate with metabolic requirements or to do so only from an elevated filling

pressure (13). Therapeutic definition: cardiomyopathies are a heterogeneous group of diseases of the

myocardium associated with mechanical and/or electrical dysfunction that usually (but

not invariably) exhibit inappropriate ventricular hypertrophy or dilatation and are due to a

variety of causes. Cardiomyopathies either are confined to the heart or are part of gen-

eralized systemic disorders, often leading to cardiovascular death or progressive HF–

related disability (14). In addition to the three definitions, HF is classified by course of disease and functionality.

With respect to course: after 3 months HF is considered as chronic. Chronic HF (CHF) can be

stable or gradually progressive, or can exacerbate into acute HF. Regarding the functionality,

HF is distinct in diastolic HF, also named HF with preserved left-ventricular ejection fraction

>40-50% (LVEF; HFpEF), and systolic HF, also named HF with reduced LVEF (HFrEF) (9).

For diagnostic purposes, for both HFrEF and HFpEF, typical signs and symptoms of HF are

obligatory. For HFpEF two more requirements are needed: a non dilated left ventricle and

documentation of diastolic dysfunction. The diagnosis of HFpEF is more difficult than that of

HFrEF because it is largely one of exclusion, e.g. of potential non-cardiac causes of the pa-

tient’s symptoms. As for the individual patient’s restrictions, the New York Heart Association

(NYHA) classification is used (15). NYHA places patients in one of four categories based on

how much they are limited during physical activity.

All possible causes for HF may be divided into derailments of the heart muscle as a result of

20

cardiomyopathy, coronary artery disease (CAD), atrial fibrillation or other rhythm or conducting

disturbances, infection or intoxication. In addition, diseases of the heart valves, afflictions of the

pericardium or (cardiac) shunts may cause HF. Finally, circulatory derangements may induce

HF, such as coarctation, thrombo-embolic process, intoxication, high output failure, infection,

hyperthyroidism and fluid overload. In general, HFrEF is most often caused by CAD which is

predominantly generated through hypertension (HT).

1.6 PREVALENCE OF HEART FAILURE.

Due to an increasingly ageing population, already in 1990 a growing number of individuals with

HF was predicted in The Netherlands (12). After 22 years, the estimation of HF became reality,

with a prevalence of about 120,000 – 130,000 subjects. Moreover, the number of people with

HF is expected to increase to reach about 200,000 individuals in 2018 (16). Between 20 to

30% of the general population will ultimately develop HF, when they are over 70 years of age

(17).

1.7 HEART FAILURE DISEASE BURDEN.

Apart from the increasing HF mortality, the morbidity or disease burden due to HF will rise as

well. The “Rijksinstituut voor Volksgezondheid en Milieu” (RIVM) has quantified “disease bur-

den” in a model of disability-adjusted life-years (DALY) (18), expressed as the number of years

lost due to ill-health, disability or early death through HF. The authors have estimated the DALY

for each of four risk factors for HF separately. These risk factors comprise hypertension (HT;

moderate HT is SBP of 140-160 and serious HT ≥160 mm Hg), moderately-high total choles-

terol (200-239 mg/dL; 5.17-6.18 mmol/L) and seriously-high total cholesterol (≥240 mg/dL;

≥6.20 mmol/L), overweight, obesity and inactivity. Inactivity is defined as <4 metabolic equiva-

lents of tasks (METs) for age 20-55 years and <3 METs for age 55+. A DALY was found of

21

1.6 years for seriously-high total cholesterol and up to 2.9 years for serious HT (Figure 3). The

consequences of DALY for the AWBZ will be that an increasing part of the Dutch health care

budget have to be devoted to the HF disease burden.

Figure 3. Disability-adjusted life-years (DALY). A DALY-model estimated for a Dutch popu-

lation classified by different risk factors for heart failure.

Disability-Adjusted Life-Years (DALY), expressed as number of years lost due to ill-health, dis-

ability or early death. The bars are equally long and the upper bar represents the life-

expectancy (LE) of a 20 year old with a lifelong optimal HF risk profile. The middle parts of the

bars reflect the years lived with HF. The right parts of the bars show the years lost due to HF

(18).

1.7 QUALITY OF LIFE (QUOL).

The DALY quantifies the time lived with reduced Quol due to HF. Not only the direct loss of

Quol by HF symptoms but also the indirect loss by restrictions, as a result of HF, are important.

These restrictions appear in several Quol domains like mobility and independency. For instance,

healthy HF free

overweight with HF

increased cholesterol loss of LE

strongly increased cholesterol

hypertension

obesity

serious hypertension

inactivity

77 78 79 80 81 82

22

dyspnea limits the ability to execute activities of daily living (ADL) and walking distance (19).

From the perspective of the individual, the severity of HF affects a person’s self-reported func-

tioning (20). It is obvious that HF is related to assistance at ADL, but this relation has not been

investigated by others. Such a relation, if causal, is important for the care-giver who needs to

estimate the amount of required ADL assistance and for the individual who needs that help.

The physician may estimate the prognosis of patients using independent predictors of HF such

as high age, NYHA 3 and 4, loss of appetite (21), N terminal pro B-type natriuretic peptide

(NT-pro)BNP (22) and chronic kidney diseases (CKD) (23). Among others these prognostic

predictors enable a patient with CVD to make personal choices for the last phase of his life. The

choices may permit the patient to retain control over his or her life and thereby prevent de-

crease of autonomy, at least on the domain of decision-making.

1.9 HF DIAGNOSTICS AND ALGORITHM.

In primary care elderly-medicine, HF is diagnosed using CV medical history, for HF predispos-

ing diseases and risk factors (chapter 1.5), clinical examination, electrocardiography (ECG),

natriuretic peptides (chapter 1.10) and echocardiography. These are referred to as diagnostic

tools. The reference standard for the, at least functionally undifferentiated HF diagnosis is based

on all of these tools, occasionally extended with chest X-ray (9;10;24). For further examination

of HF causes or treatment, referral to a cardiologist may be required. However, predictive val-

ues of the separate diagnostic tools are moderate which hamper initial HF diagnostics in pri-

mary care and in geriatric outpatients as well (25;26). This hindrance combined with no easy

access to echocardiography or unwanted referral (resident’s preference), needed a solution to

diagnose HF without outpatient referral. For that purpose HF diagnostic algorithms were devel-

oped based on combinations of diagnostic tools including NPs (21;25-28). However a few is

known whether those HF diagnostic algorithms (9;10;21;24-28) are also suitable for the resi-

dential elderly suspected of HF.

23

1.10 NATRIURETIC PEPTIDES.

At the end of the last century, NPs have emerged as reliable biomarkers for the diagnosis and

prognosis of both systolic and diastolic HF. NP testing is regarded to be the single most useful

test to add to the diagnostic pathway for HF in primary care. Tests of NPs have shown to be

cost effective. However, in HFpEF levels of NPs are increased, but somewhat lower compared

to those in patients with HFrEF. Therefore, the relatively higher NP levels are predictive for

HFrEF, but the moderately increased NP levels do not predict HFpEF very well (29). Especially

the NPs specificity and diagnostic accuracy in acute HF are prominent. Currently, NPs, notably

B-type natriuretic peptide (BNP) and N-terminal-proBNP (NT-proBNP), are commonly applied

diagnostic (30) and prognostic (31) biomarkers. Confounders of NP levels are non-cardiac

factors such as high age, sex and renal dysfunction, since these also increase NP levels. Obe-

sity, diuretics, ACE inhibitors, beta-blockers, ARBs and aldosterone antagonists reduce NP lev-

els (32). All the above mentioned confounders occur widely in residential elderly. Many of

them use CV medications and renal dysfunction is widespread. Therefore, NPs may not predict

HF exclusively (32) and NP cut-off values are not fully validated in this particular group. With

regard to treatment optimization, a meta-analysis shows that NP-guided therapy is superior to

symptom-guided therapy in groups of patients with CHF (33). However a study on individual

CHF patients reveals high intra-individual biological variations of NPs as major limitation of NPs

in the optimization of HF treatments (34).

The secretion of NPs by the ventricle wall increases when HF is accompanied by pressure or

volume overload (35). Only BNP regulates actively. BNP stimulates natriuresis, diuresis, vaso-

dilatation and inhibits the renin-angiotensin-aldosterone system (RAAS) and sympathetic nerv-

ous system. BNP is metabolized actively in blood and cells by neuropeptidases (36). The half-

life time of BNP is 20 minutes. These characteristics render BNP appropriate as marker of HF

(30). NT-proBNP has an about six times longer half-life, viz. 1-2 hours (37). As opposed to

BNP, NT-proBNP is passively filtered in the kidney (38), actively reabsorbed by the proximal

24

tubulus brush border cells and catabolised to amino acids. This process of NT-proBNP degra-

dation is usually nearly complete. Thus, only minor amounts of NT-proBNP are recovered in

urine (39).

Both NPs are released into the circulation in an 1:1 molar ratio. The secretion of BNP exhibits a

pulsatile pattern in the circulation (40) and it may be assumed that NT-proBNP is split-off

pulse-shaped as well. Consequently both NPs exhibit high intra-individual biological variations

which hamper their use for treatment optimization of an individual (34). Mainly experimentally,

NPs are determined in urine using immunoassays. Urine NPs may exhibit less intra-individual

biological variation compared to plasma NPs, assuming that the pulsatile secretion becomes

attenuated by NT-proBNP accumulation in the bladder (41-43). Thus, the experimental use of

urine NP is prompted by the high intra-individual biological variation of plasma NPs (34).

1.11 THERAPIES FOR HEART FAILURE.

Following the diagnosis of HF, treatment is based on CV risk management (CVRM) (2) and

correction of underlying causes (10). CVRM for HF differs in the time of intervention (2) from

other HF guidelines. The current trend is not to wait until HF symptoms have developed, but to

start intensive risk management for patients with increased risk of HF (www.RIVM.nl). It is as-

sumed that if HF could be recognized before symptoms emerge, lifestyle adaptations or appro-

priate medication could reduce HF symptoms or the physical decline caused by HF. An example

of lifestyle effects on CV risk factors is presented by a five-year lasting Dutch study. This study

shows that 1% of adults has changed from physically inactive into moderately-active and 1-2%

from overweight/obese or underweight to normal weight (6). Whether such endpoints can also

be reached in HF patients belonging to the elderly and whether such interventions could prevent

HF progression or even accomplish a reverse, is unknown.

25

In addition, it is advised to treat other diseases precipitating to HF. It is assumed that treatment

of anemia, lung diseases, renal dysfunction, thyroid-gland dysfunction, T2DM and intoxications

reduces the risks of HF (9). Treatment of underlying causes of HF consists, among others, of pharmacological therapy. Most pharmacological therapies are specific for HFrEF (Table 1).

26

Table 1. Treatment of Heart Failure with reduced left ventricular ejection fraction (HFrEF)

(26).

HF with reduced ejection fraction treatment start diuretics to relieve signs/symptoms

and ACE-i (or ARB) When clinically stable, add beta-blocker

Titrate to the target doses of When complaints are not persisting

ACE-i (or ARB) and beta-blocker no complementary therapy indicated

When complaints are persisting, add MR antagonist

When instable (SR and HR ≥ 70/min), add Ivabradine

When complaints does not persist, EF≤35% No further specific treatment

When complaints are persisting and QRS≥120ms Consider CRT-P/CRT-D

When complaints are persisting and QRS ≤120ms consider ICD

When complaints are persisting Consider digoxin and/or hydralazine or isosorbide nitrate

Angiotensin Converting Enzyme inhibitor, ACE-i; Angiotensin Receptor II Blocker, ARB; ejection fraction, EF; sinus rhythm, ARB; heart rate, HR; mineralocorticosteroid receptor, MR; cardiac resynchronization therapy pacemaker, CTR-P/CRT-D, defibrillator.

Since patients with advanced CHF (EF ≤35 - 40%) are at risk of sudden death, an implantable

cardioverter defibrillator (ICD) is placed, for preventive reasons according to guidelines. Finally,

non-pharmacological treatments to improve systolic function of the left ventricle function com-

prise, among others, revascularization procedures, valvular or ventricular surgery, rhythm opti-

mization (electrocardioversion, ablation), cardiac resynchronization therapy (CRT) and heart

transplantation. For HFpEF no specific therapy is available, yet. For these patients only the un-

27

derlying cause such as hypertension is treated (10). For the elderly, the same treatment is

available provided that dosages are adjusted to decreased kidney function and the life expec-

tancy is at least 0.5 years with good quality of life (10).

1.12 THE RESIDENTIAL ELDERLY.

The residential elderly belong to the frailest elderly (chapter 1.2). They reside in care or nursing

homes (LTCFs). Care and nursing homes differ in intermittently or continuously provided sup-

port and ADL-help, respectively. During the final stage of writing this thesis in 2015, financing of

Dutch care homes had just ended. As a consequence most care home elderly had to move to

private houses with care at home or to a nursing home.

In primary care elderly medicine, preferences of the residential elderly are inventoried and taken

as the starting point for further treatment decisions in case they become incompetent in future.

These preferences concern diagnostics, treatment and hospital referral in the event of emer-

gency. These choices or preferences are set out in agreements that are evaluated twice a year,

or more often when indicated. These agreements may include renouncing invasive and intensive

care treatments and hospital referral. As an exception, hospital referral for (surgical) treatment

of bone-fractures are seldom renounced.

1.13 HEART FAILURE IN RESIDENTIAL ELDERLY.

The increasing HF prevalence on the one hand, and lack of specific HF research data on the

other hand challenge healthcare professionals in the provision of optimal patient care in the

residential elderly. We expected the HF diagnostics to be less accurate due to high age, female

sex, multimorbidity (44;45), ’treatment agreements’ and absence of a specific HF guideline. At

the start of this study, it was unknown whether general HF guidelines could be applied to the

residential elderly (9;10;24). Therefore, we set out to study HF prevalence and diagnostic ac-

28

curacy in this particular group of elderly people. We planned also to investigate whether HF

diagnostics could be improved by adding NPs to the usual HF diagnostics.

Furthermore, HF guidelines, the CVRM guideline and family members recommend residential

elderly to be physically active (2;9;10;24) aiming to maintain or improve physical health. The

residential elderly, however, seems to have little motivation to train physically on a regular ba-

sis. He or she faces a limited life expectancy wondering whether his exertion will provide bene-

fits. Not much is known as to whether non-elective exercise might be beneficial in residential

elderly (46). Staffs of nursing and care homes doubt whether they should invest in regular ex-

ercise programmes and whether such interventions will be safe and effective. Therefore, we

experimented with an exercise program based on a review specific to the residential elderly

(47).

In addition, when facing a limited life expectancy, accurate information on prognosis will become

more important. Prognostic information allows residential elderly to take their own decisions so

that loss of Quol at least in the area of autonomy is reduced (chapter 3). Natriuretic peptides

have been shown a powerful prognostic marker in adults with HF and may have similar prog-

nostic value in residential elderly. However, the prognostic properties of NPs have not been not

examined in this particular group. Moreover, many confounders influence NP levels in such a

way that they hamper to inform on prognosis in this residential target group. Therefore we

aimed to study whether NPs could be used as a prospective and independent predictor in those

frail elderly with HF.

Independency of care takers determines Quol to a high degree. It was surprising to learn, at the

start of this study, that the HF diagnosis was not on the list of geriatric chronic diseases in the

‘Resident Assessment Instrument’ (RAI). The RAI is in use in American and Dutch nursing

homes for care-planning at an individual, organizational and state level (48). Nevertheless, we

presumed a relationship between HF state and help-with-care in residential elderly. In that line

of reasoning, NP concentrations might be related to help-with-care as well. However a relation

29

between HF and or NPs with ADL-assistance was not observed until now. Therefore, we set out

to test these research questions.

1.14 AIM AND OUTLINE OF THE THESIS.

In a broad perspective, the studies in this thesis aim to determine whether application of parts of

the HF guidelines can be accurately applied to the specific group of the residential elderly. The

doubts on the applicability of HF guidelines in residential elderly is based on the systematic ex-

clusion of this group from HF research. There is also dubiety based on the observed higher cut-

off values of NPs at a higher age, in renal dysfunction and comorbidity. Evidence as to what

extent HF guidelines can be applied is important to elderly care physicians. These questions

have undermined the accuracy of HF diagnostics resulting in no treatment or inappropriate

treatment, and less well-being. Therefore, we aimed to obtain insight into certain HF aspects,

such as prevalence, diagnostic accuracy, diagnostic and prognostic values of NPs and the rela-

tion of HF with help-at-care. To this end, we screened the residential elderly of a single nursing

home in Groningen on the presence of CHF. To examine its external validity we repeated this

screening program in Aruban nursing homes.

Furthermore, we sought to investigate whether physical activity was to the benefit of care home

elderly on the basis of cardio-metabolic endpoints. These comprise waist circumference, blood

pressure, glucose-homeostasis and lipids. We were also interested to see whether complaints

and NPs ameliorate in the care home elderly with HF. With both aims, we conducted a random-

ized controlled trial employing a 16-weeks lasting supervised exercise intervention. For this, we

collaborated with investigators examining partly the same care home elderly for other, more

functional, endpoints (46).

At last, aiming to gain insight into the feasibility of NPs for treatment optimization, we estimated

the intra- and inter-individual variations of NPs in urine and plasma of stable CHF patients living

in Curaçao.

30

Specific aims of this study are described in the following paragraphs: In Chapter 1 we provide a general introduction of HF with special emphasis on HF diagnostics

in residential elderly. In Chapter 2.1 we study both prevalence and accuracy of HF diagnostics in the Groningen resi-

dential elderly. In the meantime, we aim to find out whether improvements can be achieved by adding NPs to the diagnostic process in use in a Dutch nursing home. The aim of Chapter 2.2

is similar to that of Chapter 2.1. but now applied to all residential elderly in Aruba. In Chapter

2.3 we study the intra-individual variations of NPs with the purpose of finding out whether urine

NPs may exhibit less variations compared to those in plasma in the Curacao population In Chapter 3 we study the prognostic value of NPs on one year mortality of the Groningen resi-

dential elderly, aiming to better inform those facing a limited life expectancy. The aim of Chap-

ter 4 is to determine the relationship between CHF and the required need for ADL-help in the

residential elderly in Groningen. The aim of Chapter 5 is to examine the effects of a physical

exercise program for the Groningen care home elderly. We investigated whether such a pro-

gram may be beneficial with respect to cardiovascular-metabolic (soft) endpoints. Here, we

could collaborate with an investigator who would perform the same intervention, but aiming at

other, more functional endpoints. In Chapter 6.1 we summarized and in Chapter 6.2 we discussed the results of the above de-

scribed studies and in Chapter 6.3 we provided recommendations for future studies. In Chap-

ter 7 a Dutch summary (“Samenvatting”) is given.

31

Figure 4. Flowchart of studies presented in this thesis with the corresponding populations in

which the studies have been conducted.

ADL, Activities of daily living.

Heart failure diagnostics

Heart failure and 1-year mortality

Heart failure and ADL-assistance

Heart failure and exercise

All nursing homes on Aruba N=51

1 nursing home Groningen N=103; ADL-assist. N=93

Heart failure patients on Curaçao N=25

4 care homes in Groningen N=52

Natriuretic peptides urine versus plasma

32

References

1 Charlson ME, Charlson RE, Peterson JC, Marinopoulos SS, Briggs WM, Hollenberg JP.

The Charlson comorbidity index is adapted to predict costs of chronic disease in primary

care patients. J Clin Epidemiol 2008;61(12):1234-40.

2 Wiersma T, Smulders YM, Stehouwer CD, Konings KT, Lanphen J. [Summary of the

multidisciplinary guideline on cardiovascular risk management (revision 2011)]. Ned

Tijdschr Geneeskd 2012;156(36):A5104.

3 Koopmans RT, Lavrijsen JC, Hoek JF, Went PB, Schols JM. Dutch elderly care physici-

an: a new generation of nursing home physician specialists. J Am Geriatr Soc

2010;58(9):1807-9.

4 Hilmer SN, Mager DE, Simonsick EM, Cao Y, Ling SM, Windham BG, et al. A drug bur

den in dex to define the functional burden of medications in older people. Arch Intern

Med 2007;167(8):781-7.

5 Akner G. Analysis of multimorbidity in individual elderly nursing home residents. Devel-

opment of a multimorbidity matrix. Arch Gerontol Geriatr 2009;49(3):413-9.

6 Wendel-Vos W, Droomers M, Kremers S, Brug J, van Lenthe F. Potential environmental

determinants of physical activity in adults: a systematic review. Obes Rev

2007;8(5):425-40.

7 Hirdes JP, Ljunggren G, Morris JN, Frijters DH, Finne, Soveri H, et al. Reliability of the

interRAI suite of assessment instruments: a 12-country study of an integrated health in-

formation system. BMC Health Serv Res 2008;8:277-88.

33

8 Smets IH, Kempen GI, Janssen-Heijnen ML, Deckx L, Buntinx FJ, van den Akker M.

Four screening instruments for frailty in older patients with and without cancer: a diag-

nostic study. BMC Geriatr 2014;14:26.

9 McMurray JJ, Adamopoulos S, Anker SD, Auricchio A, Bohm M, Dickstein K, et al. ESC

Guide lines for the diagnosis and treatment of acute and chronic heart failure 2012: The

Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure 2012 of

the European Society of Cardiology. Developed in collaboration with the Heart Failure

Association (HFA) of the ESC. Eur Heart J 2012 33(14):1787-847.

10 Voors AA, Walma EP, Twickler TB, Rutten FH, Hoes AW. [Multidisciplinary guideline

'Heart failure 2010']. Ned Tijdschr Geneeskd 2011;155:A2957.

11 National Institute for Public Health and the Environment (RIVM). National Public

Health Compass. 2013 [November 19, 2013]; Available from:

http://www.nationaalkompas.nl/ gezondheid-en-ziekte/ziekten-en- aandoenin- gen/hartvaatstelsel).

12 Vaartjes I, Koopman C, van Dis I, et al. Hart- en vaatziekten in Nederland 2013, cijfers

over leefstijl, risicofactoren, ziekte en sterfte. The Hague: Dutch Heart Foundation; 2013.

13 Braunwald E. Pathophysiological of HF. In: Braunwald’s Heart Disease. A textbook of

Cardio vascular Medicine. 10th ed. 2014.14.

14 Maron BJ, Towbin JA, Thiene G, Antzelevitch C, Corrado D, Arnett D, et al. Contempo-

rary definitions and classification of the cardiomyopathies: an American Heart Association

Scientific Statement from the Council on Clinical Cardiology, Heart Failure and Trans

34

plantation Committee; Quality of Care and Outcomes Research and Functional Genom-

ics and Translational Biology Interdisciplinary Working Groups; and Council on

Epidemiology and Prevention. Circulation 2006;113(14):1807-16.

15 Raphael, Claire; Briscoe, C; Davies, J; Ian Whinnett, Z; Manisty, C, et al (2007).

"Limitations of the New York Heart Association functional classification system and

self-reported walking dis- tances in chronic heart failure." Heart (British Cardiac So-

ciety) 93(4):476–82.

16 Engelfriet PM, Hoogenveen RT, Poos MJJC, Blokstra A, van Baal PHM, Verschuren

WMM. Hartfalen: epidemiologie, risicofactoren en toekomst. Heart failure: epidemiology,

risk factors and future. In: Rijksinstituut voor Volksgezondheid en Milieu, RIVM. Rapport

260401006. 2012.

17 Bleumink GS, Knetsch AM, Sturkenboom MCJM, Straus SM, Hofman A, Deckers

JW, et al. Quantifying the heart failure epidemic: prevalence, incidence rate, lifetime risk

and prognosis of heart failure the Rotterdam study. Eur Heart J 2004;25:1614–1619.

18 Murray, C. J. and Lopez, A. D. Evidence-based health policy-lessons from the Global

Burden of Disease Study. Science 274(5288), 740-743. 1-11-1996.

19 Steptoe A, Deaton A, Stone AA. Subjective wellbeing, health, and ageing.

Lancet 2015;385(9968):640-8.

20 Munyombwe T, Hofer S, Fitzsimons D, Thompson DR, Lane D, Smith K, et al. An

evaluation of the Minnesota Living with Heart Failure Questionnaire using Rasch analy

35

sis. Qual Life Res 2014;23(6):1753-65.

21 Oudejans I, Mosterd A, Zuithoff NP, Hoes AW. Applicability of current diagnostic algo-

rithms in geriatric patients suspected of new, slow onset heart failure. Age Ageing

2012;41(3):309-16.

22 Don-Wauchope AC, Santaguida PL, Oremus M, McKelvie R, Ali U, Brown JA, et al.

Incremental predictive value of natriuretic peptides for prognosis in the chronic stable

heart failure population: a systematic review. Heart Fail Rev 2014;19(4):521-40.

23 Damman K, Voors AA, Navis G, van Veldhuisen DJ, Hillege HL. Current and novel renal

biomarkers in heart failure. Heart Fail Rev 2012;17(2):241-50.

24 NICE Guideline No 5: Chronic Heart Failure;

www.pathways.nice.org.uk/pathways/chronic-heart-failure.

25 Kelder JC, Cramer MJ, van Wijngaarden J, van Tooren R, Mosterd A, Moons KG, et al.

The diagnostic value of physical examination and additional testing in primary care pa

tients with suspected heart failure. Circulation 2011;124:2865–2873.

26 Oudejans I, Mosterd A, Bloemen JA, Valk MJ, van Velzen E, Wielders JP, et al. Clinical

evalua tion of geriatric outpatients with suspected heart failure: value of symptoms, signs,

and additional tests. Eur J Heart Fail 2011;13:518–527.

27 Roalfe AK, Mant J, Doust JA, Barton P, Cowie MR, Glasziou P, et al. Development and

initial validation of a simple clinical decision tool to predict the presence of heart failure in

36

primary care: the MICE (Male, Infarction, Crepitations, Edema) rule. Eur J Heart Fail

2012;14(9):1000-8.

28 Mant J, Doust J, Roalfe A, Barton P, Cowie MR, Glasziou P, et al. Systematic review

and individual patient data meta-analysis of diagnosis of heart failure, with modelling of

implications of different diagnostic strategies in primary care. Health Technol Assess.

2009;13(32):1-207, iii.

29 Paulus WJ, Tschope C, Sanderson JE, Rusconi C, Flachskampf FA, Rademakers FE, et

al. How to diagnose diastolic heart failure: a consensus statement on the diagnosis of

heart failure with normal left ventricular ejection fraction by the Heart Failure and Echo

cardiography Associations of the European Society of Cardiology. Eur Heart J

2007;28(20):2539-50.

30 Januzzi JL, Jr. Natriuretic peptide testing: a window into the diagnosis and prognosis of

heart failure. Cleve Clin J Med 2006;73(2):149-7.

31 Sato T, Yamauchi H, Suzuki S, Yoshihisa A, Yamaki T, Sugimoto K, et al. Distinct prog-

nostic factors in patients with chronic heart failure and chronic kidney disease. Int Heart

J 2013;54(5):311-7.

32 Fent G, Hall I, Watt V, O'Toole L, Al-Mohammad A. 43 Raised Natriuretic Peptides are

not Ex clusively Indicative of Heart Failure: A Service Review of a Diagnostic Heart Fail-

ure Clinic. Heart 2014;100 Suppl 3:A23-A24.

33 De Vecchis R, Esposito C, Di BG, Ariano C, Giasi A, Cioppa C. B-type natriuretic pep

tide-guided versus symptom-guided therapy in outpatients with chronic heart failure: a

37

systematic review with meta-analysis. J Cardiovasc Med (Hagerstown) 2014;15(2):122-

34.

34 Bruins S, Fokkema MR, Romer JW, Dejongste MJ, van der Dijs FP, van den Ouweland

JM, et al. High intraindividual variation of B-type natriuretic peptide (BNP) and amino-

terminal proBNP in patients with stable chronic heart failure. Clin Chem

2004;50(11):2052-8.

35 Iwanaga Y, Nishi I, Furuichi S, Noguchi T, Sase K, Kihara Y, et al. B-type natriuretic

peptide strongly reflects diastolic wall stress in patients with chronic heart failure: com-

parison between systolic and diastolic heart failure. J Am Coll Cardiol 2006;47(4):742-

8.

36 Levin ER, Gardner DG, Samson WK. Natriuretic peptides. N Engl J Med

1998;30;339(5):321-8.

37 Balion CM, Santaguida P, McKelvie R, Hill SA, McQueen MJ, Worster A, et al. Physio-

logical, pathological, pharmacological, biochemical and hematological factors affecting

BNP and NT- proBNP. Clin Biochem 2008;41(4-5):231-9.

38 Gekle M. Renal tubule albumin transport. Annu Rev Physiol 2005;67:573-94.

39 Palmer SC, Endre ZH, Richards AM, Yandle TG. Characterization of NT-proBNP in hu-

man urine. Clin Chem 2009;55(6):1126-34.

40 Pedersen EB, Bacevicius E, Bech JN, Solling K, Pedersen HB. Abnormal rhythmic os-

cillations of atrial natriuretic peptide and brain natriuretic peptide in chronic renal failure.

38

Clin Sci (Lond) 2006;110(4):491-501.

41 Cortes R, Portoles M, Salvador A, Bertomeu V, Garcia de BF, Martinez-Dolz L, et al.

Diagnostic and prognostic value of urine NT-proBNP levels in heart failure patients. Eur

J Heart Fail 2006;8(6):621-7.

42 Song J, Li P, Li H, Wang Q. The clinical significance of a urinary B-type natriuretic pep

tide as say for the diagnosis of heart failure. Clin Chim Acta 2011;412(17-18):1632-6.

43 Michielsen EC, Bakker JA, Kimmenade RR, Pinto YM, Dieijen-Visser MP. The diagnos-

tic value of serum and urinary NT-proBNP for heart failure. Ann Clin Biochem

2008;45(Pt 4):389-94.

44 Daamen MA, Schols JM, Jaarsma T, Hamers JP. Prevalence of heart failure in nursing

homes: a systematic literature review. Scand J Caring Sci 2010;24(1):202-8.

45 Rutten FH, Cramer MJ, Grobbee DE, Sachs AP, Kirkels JH, Lammers JW, et al. Unrec-

ognized heart failure in elderly patients with stable chronic obstructive pulmonary dis

ease. Eur Heart J 2005;26(18):1887-94.

46 Chin APM, van Poppel MN, van MW. Effects of resistance and functional-skills training

on habitual activity and constipation among older adults living in long-term care facilities:

a randomized controlled trial. BMC Geriatr 2006;6:9.

47 Weening-Dijksterhuis E, de Greef MH, Scherder EJ, Slaets JP, van der Schans CP.

Frail institutionalized older persons: A comprehensive review on physical exercise, physi

39

cal fitness, activities of daily living, and quality-of-life. Am J Phys Med Rehabil

2011;90(2):156-68.

48 Grebe C, Brandenburg H. [Resident assessment instrument. Application options and

relevance for Germany]. Z Gerontol Geriatr 2015;48(2):105-13.

40

41

Chapter 2.1

PREVALENCE AND MISDIAGNOSIS OF CHRONIC HEART FAILURE

IN NURSING HOME RESIDENTS: THE ROLE OF B-TYPE NATRI-

URETIC PEPTIDES. Barents M, van der Horst IC, Voors AA, Hillege JL, Muskiet FA, de Jongste MJ.

Neth Heart J. 2008;16(4):123-8.

42

ABSTRACT BACKGROUND/OBJECTIVES. Without knowing the exact CHF prevalence, chronic heart failure

(CHF) occurs frequently in elderly people both inside and outside nursing homes (NH). For a

diagnosis we have to rely on physical examination and additional tests. We therefore run risks

of missing CHF diagnoses or of diagnosing CHF when we should not. Natriuretic peptides as-

says have emerged as a diagnostic test but their use in NH residents is limited. We examined

the number of misdiagnoses, the CHF prevalence and the role of natriuretic peptide. METHOD. Residents in one centre without aphasia, cognitive impairments or metastatic cancer

were screened on CHF. The natriuretic peptide levels were measured separately. RESULTS. Of the 150 residents, 103 (64%) were included (79 ± 11 years). The diagnosis of

CHF was established in 24 of these 103 residents with NT-proBNP 1871 (interquartile range

(IQR) 539-4262) and of BNP 194 (IQR 92-460) pg/mL. A striking result was that of the 24

residents found to have CHF after the screening, 15 (66%) had previously been undetected:

NT-proBNP 1146 (IQR 228-3341) and BNP 200 (IQR 107-433) pg/mL). Moreover, in 13 out

of 22 (62%) residents who had previously been thought to have CHF, the diagnosis was re-

jected: NT-proBNP 388 (IQR 174-719) and BNP 90 (IQR 35-128) pg/mL respectively. Re-

garding the diagnostic accuracy of NT-proBNP and BNP, the optimal cut-off level of NT-

proBNP was 450 pg/mL with a sensitivity 0.71 and a specificity 0.67, and of BNP was 100

pg/mL with a sensitivity 0.71 and a specificity 0.70. CONCLUSION. Both undetected and incorrect CHF diagnoses were largely observed. NT-

proBNP and BNP were moderately accurate at CHF diagnosis. The CHF prevalence was 23%.

43

INTRODUCTION Nursing home (NH) residents are characterized by high age and co-morbidity. Chronic heart failure

(CHF) occurs frequently, but is a relatively neglected diagnosis (1). The main reasons why CHF is less

frequently detected are non-specific complaints, co-morbidity and limited access to echocardiography

(2;3;4). On the other hand, some residents may be diagnosed with CHF, while their symptoms may be

caused by other disorders. Both undetected diagnosis and incorrect diagnosis will lead to inadequate

treatment of the underlying disease, and will impair the quality of life.

Natriuretic peptides are increasingly used to aid the diagnosis of CHF. Its diagnostic value is particularly

strong in dyspnoeic patients admitted to the emergency department but less strong in general practices

(5). NH residents are mostly of high age and have multiple chronic diseases, such as renal dysfunction

and diabetes. The role of natriuretic peptides for the diagnosis of CHF in this specific population may be

limited, but so far it has not extensively been studied. In the present study we therefore aim to look at

the prevalence of CHF, both undetected and incorrect diagnoses of CHF, and the diagnostic accuracy of

NT-proBNP and BNP in NH residents.

METHODS During the course of this study there were 150 persons in the departments for residents with diseases of

somatic origin in NH “het Zonnehuis”. Their impairments are based on cardiovascular, pulmonary, neu-

rodegenerative, skeletal muscle and other disorders. Most of them were receiving long-term care (long-

stay). Some residents with musculoskeletal or cerebrovascular disorders were rehabilitated , one-third of

whom could be discharged (short-stay) but should remain care dependent in primary care. Both long-

stay and short-stay groups were invited to take part between 25 September 2004 and 24 May 2005.

Both groups of residents were included if they understood the impact of the study on themselves (com-

petent) and if they agreed to participate by written informed consent. Residents with aphasia or a cogni-

tive impairments, measured by the mini-mental state estimation (MMSE) test, were excluded. The

MMSE test contains 30 questions and a MMSE ≤ 20 is suspect for cognitive impairments (6). Persons

with metastatic cancer who stayed in a department specialized in palliative terminal care were excluded.

Persons were also excluded if the echocardiographic frames could not be sufficiently visualized or if they

44

refused to have their blood sample taken. CHF was defined according to the definition of the Chronic

Heart Failure Guidelines of the European Society of Cardiology (7). Two experienced heart failure cardi-

ologists independently decided on the diagnosis of CHF, based on medical history, physical examination,

ECG, routine blood tests and echocardiography. In case of disagreement a third cardiologist reviewed

the case and made the final diagnosis. At this stage none of the cardiologist knew the levels of BNP or

NT-proBNP. The study was approved of by the Medical Ethical Committee in Groningen, University

Medical Center, The Netherlands.

MATERIALS

In this cross-sectional study, all data were collected anonymously and within one week (ques-

tionnaire, neurohormone sampling, ECG, echocardiography).

One physician collected the data of the patients’ medical history (CHF with NYHA classification,

coronary artery diseases (CAD), hypertension, peripheral vascular disorders, diabetes mellitus

and other comorbidity), of their symptoms (fatigue and dyspnea at great or light effort and in

rest, when laying down and at night, breast pain, palpitation), and of their medication. He also

examined blood pressure, central jugular pressure (CJP), the heart (third heart sound, gallop

rhythm, cardiomegaly by percussion), the lungs (rales), the liver (liver-percussion), length and

weight and looked at periphery oedema (8). A 12 leads electrocardiogram (ECG) was made

with the electrocardiograph Cardioline delta three plus (Cardioline, Milan, Italy, www.cardioline.it) with the patient in a horizontal position. One blood sample per resident (12

mL) was taken if one was in fasting condition and at rest. Assessments of creatinine, haemo-

globin and mean corpuscular volume (MCV) were performed in the NH laboratory. Since the

coefficient of variation of neurohormones is reported to be about 100% we chose not to repeat

the neurohormone sampling (9). For determination of levels of (NT-pro)BNP, a 5 µl aprotinine

solution was adjusted to the 250 µl plasma samples in EDTA and to 250 µl serum samples. At

the UMCG Clinical Chemical Laboratory (CCL) both were frozen at -20˚C and stored in batches

45

for a maximum of 10 months. The assays were run in one go for both NT-proBNP and BNP.

Both NT-proBNP and BNP were measured by immunoassays (Elecsys®1010/2010/modular

analytics 2004 Roche diagnostics Indianapolis IN US, and AXSYM system® BNP 2003 Axis–

shield diagnostics LTD ABBOT Wiesbaden Germany). NT-proBNP and BNP had coefficients of

variation of 3.3% and 7.8% respectively, and measuring ranges of 5-35000 pg/mL and 0-

3465 pg/mL. The renal function has been defined as glomerular filtration rate (GFR) measured

by the Cockcroft and Goult formula in mL/min.

On the UMCG department of Cardiology, echocardiography was performed by four experienced

echocardiographists using a hand held cardiograph ‘Opti Go’ (Philips, Eindhoven, The Nether-

lands, www.philips.com). The left-ventricular ejection fraction was assessed semi-quantitatively

by the two-dimensional visual estimate method (10). A LVEF ≤ 45% was considered to be a

left ventricular systolic dysfunction (LVSD). If the echocardiogram was of insufficient quality, the

subject was transported to the UMCG hospital where an echocardiogram was produced with the

General Electric Vingmed Ultrasound five (www.gemedicalsystems.com, Zoetermeer, The Neth-

erlands).

Statistics Statistical analyses were performed using SPSS 12.0.1 software (SPSS Inc., 233 S. Wacker,

11th Floor, Chicago IL US). Differences of base characteristics of residents with and without CHF

were submitted at Student t tests, of gender at Chi-square test and the Mann Whitney test for

non-parametric continuous data, as appropriate. ROC curves were made of NT-proBNP and

BNP levels. The Area Under the Curve (AUC) presents the test accuracy and the significance

of the difference of both AUCs was estimated by a Chi-square test. Influences of age, gender,

renin-angiotensin system blocker (ACE-inhibitor or angiotensin receptor blocker) and renal

dysfunction on (NT-pro)BNP cut-off points were determined by Mann Whitney tests. All statisti-

cal comparisons were two-tailed, and a P-value <0.05 was considered to be statistically sig-

nificant.

46

RESULTS

A total of 150 individuals was screened. Fourteen residents refused to participate, 30 were ex-

cluded because of cognitive impairment and aphasia. Three persons were excluded because of

incomplete data (two poor quality echocardiograms and one refusing the blood sample). The

remaining 103 residents were included. The mean age of the population was 78 ± 11 years.

Twenty two persons had CHF before the study. Twenty four persons had CHF after the study

and we found 3 groups: subjects diagnosed with CHF before the study whose diagnosis was

confirmed afterwards (CHF confirmed), subjects diagnosed with CHF before the study whose

diagnosis was rejected afterwards (CHF rejected) and subjects not diagnosed with CHF before

the study but diagnosed with CHF after the study (CHF de novo). Of the 24 CHF, nine had CHF confirmed and 15 had CHF de novo. Thirteen subjects had CHF rejected (Table 1).

The CHF prevalence was 23% (24/103).

Residents with CHF more often had CAD, used a renin-angiotensin system blocker, complained

of fatigue, had a raised CJP, had a cardiac thoracic ratio of more than a half, had a third heart

tone and or gallop rhythm, had pulmonary rales, had left ventricle hypertrophy, suffered more from renal dysfunction and from left ventricular dysfunction (see Table 2).

Residents with CHF after compared to those without CHF had NT-proBNP medians of 1871 and

239 pg/mL (P<0.001) respectively and BNP medians of 194 and 68 pg/mL (P<0.001).

Residents with CHF before compared to those without CHF had NT-proBNP medians of 552

and 241 pg/mL.(P=0.013) and BNP medians of 92 and 76 pg/mL respectively (P=0.270) (see Table 3).

The predictive values of the two neurohormones presented by the AUC of NT-proBNP (0.815)

and of BNP (0.758) did not differ significantly (P= 0.234) as is shown in Figure 1. Table 4

presents the sensitivity, specificity, positive and negative predictive values at several cut-offs

points of NT-proBNP and BNP. In search of threshold values with the best combination of ex-

cluding and including CHF, threshold values of NT-proBNP at 450 pg/mL (sensitivity 0.71,

specificity 0.67, positive and negative predictive values of 0.42 and 0.91) and of BNP at 100

47

pg/mL (sensitivity 0.71, specificity 0.70, positive and negative predictive values of 0.63 and

0.89 respectively) were found. Both BNP and NT-proBNP were influenced by renal function

and estimated in subjects with GFR > 75 mL/min and <75 mL/min: NT-proBNP means and

SDs were 619 ± 1240 compared to 2010 ± 4191 pg/mL (P =0.005). BNP means and SDs

were 110 ± 171 compared to 251 ± 457 pg/mL (P =0.004). BNP and NT-proBNP were influ-

enced significantly by age and renal dysfunction (age and BNP Z=-2,220, P=0.026; renal dys-

function and BNP Z=-2,001, P=0.045; age and NT-proBNP Z=-4,196, P=0.000; renal dys-

function and NT-proBNP Z=-3,663, P=0.000). NT-proBNP and BNP cut-off points were not

significantly influenced by gender nor by use of renin angiotensin blockers.

DISCUSSION

In this NH study, CHF was present in almost one quarter of the residents. We found that the

majority of these residents had previously not been diagnosed with CHF. In addition, there were

several residents, who had been diagnosed with CHF before the study, whose diagnosis had

to be rejected after careful examination. Blinded values of both BNP and NT-proBNP differed

significantly between residents who were diagnosed with CHF and those whose previous CHF

diagnosis was rejected. In initial screening, NT-proBNP and to a lesser extent BNP, were mod-

erately accurate in predicting the presence of CHF.

A prevalence of 23% in this NH cohort is roughly twice as much compared with the prevalence

of CHF in the general population but concomitant with another NH population and with a popu-

lation of 70 years and over (11-13).

Both undetected diagnoses and incorrect diagnoses of CHF were expected, but they exceeded

our expectations to a large extent. Reasons for an incorrect diagnosis of CHF were a history of

atrial fibrillation (5 cases) and the coexistence of COPD (3 cases). COPD as a reason for mis-

diagnosis has earlier been described by Rutten (14). In addition, the majority of individuals with

rejected CHF had NYHA class I and II (12/13) and one had class III. Eleven out of 15 CHF de

48

novo may have been overlooked because of mild complaints (7 NYHA class I and 4 class II).

All 4 other subjects with CHF de novo with NYHA class III and IV were missed because multiple

co-morbidity confused their symptoms of cardio-vascular origin. Thus, both incorrect CHF diag-

noses and undetected CHF diagnoses seemed to have been caused by the existence of co-

morbidity. The unspecific clinical presentation of elderly people and the aspecificity of the diag-

nostic tools add to the number of misdiagnoses.

Moreover, this study shows that thorough physical examination in combination with echocardi-

ography can improve the accuracy of the CHF diagnostics. In addition, natriuretic peptides can

improve the CHF diagnostics if added to the usual diagnostic procedure even without cardiac

ultrasounds.

Failures of the expert panel are possible but unlikely, since standard diagnostic tests were per-

formed, and the panel existed of specialized heart failure cardiologists. In addition, NT-proBNP

and BNP levels supported both undetected diagnoses and incorrect CHF diagnoses.

Both NT-proBNP and BNP were moderately accurate in predicting CHF. However, depending

on the cut-offs, the CHF diagnosis will still be missed in a substantial number of residents.

Moreover, one should realize that (NT-pro) BNP levels will increase not only with CHF but also

in conditions like high age and renal dysfunction. So the use of natriuretic peptides alone will

not be sufficient for an adequate diagnosis, although the results have been significantly im-

proved when compared to the baseline situation, especially for physicians not specialized in

cardiology (15). If natriuretic peptides are used as an initial test, we suggest NT-proBNP and

BNP cut-offs of 450 and 50 pg/mL respectively for excluding CHF, and 900 and 200 pg/mL

for establishing CHF. When values are within the cut-off values, a regular assessment should

be performed. The NT-proBNP cut-off of 450 pg/mL is suggested as an age specific cut-point.

The BNP cut-off of 50 pg/mL is lower than in other studies where 100 pg/mL is proposed.

NT-proBNP and BNP were influenced by age and renal dysfunction in accordance with other

studies (16-19). In contrast with other studies, they were not influenced by gender nor by the

use of renin angiotensin system blockers (20-22).

49

A limitation of the study is the small population of 103 NH-residents in one centre, which could

lead to a selection bias in arriving at the prevalence of CHF being 23%. On the other hand, our

prevalence of chronic heart failure is concomitant with the prevalence in other NH residents and

in the elderly (12;23;24). Moreover, the population studied is representative of other NH popu-

lations, in terms of distribution of gender, age, diabetes and the use of ACE-inhibitor therapy,

although not representative in terms of the presence of hypertension and renal dysfunction

(24;25). Furthermore, despite good criteria, setting the diagnosis of CHF remains difficult be-

cause of the lack of diagnostic accuracy of physical examination and the lack of easy access to

echocardiography.

CONCLUSION

At present the overall accuracy of the diagnosis CHF is limited. In this study both undetected

and incorrect CHF diagnoses were observed in a large number of residents. Moreover, the

study shows that a more thorough investigation of residents on CHF will lead to more accurate

diagnostics, which in turn will improve residents’ quality of life. The use of NT-proBNP and BNP

as additional diagnostic instruments is promising also in nursing home residents but needs fur-

ther evaluation. The CHF prevalence in care-dependent elderly people was almost 25%, which

is twice as much as in old people living independently. The use of NT-proBNP and BNP as

additional diagnostic instruments is promising also in NH residents but needs further evaluation.

ACKNOWLEDGMENTS

This study was subsidised by the “Vereniging het Zonnehuis”, Soesterberg, The Netherlands. Financial disclosure: the authors have nothing to disclose. Author contribution: Dr. Voors was

responsible for the echocardiography, Dr. Muskiet was responsible for the biochemistry, Dr. Hil-

lege was responsible for the statistics, Dr. de Jongste was responsible for the design and Drs.

50

Barents was principal investigator and responsible for the manuscript. (Dr. Voors, Dr. van der Horst and

Dr. de Jongste were also independent panel members and responsible for the proofreading).

51

References

1 Cleland JG, Swedberg K, Follath F, Komajda M, Cohen-Solal A, Aguilar JC, Dietz R,

Gavazzi A, Hobbs R, Korewicki J, Madeira HC, Moiseyev VS, Preda I, van Gilst WH,

Widimsky J, Freemantle N, Eastaugh J, Mason J. The Euro Heart Failure survey pro -

gramme - a survey on the quality of care among patients with heart failure in Europe.

Part 1: patient characteristics and diagnosis. Eur Heart J 2003 March;24(5):442-463.

2 Luchi RJ, Taffet GE, Teasdale TA. Congestive heart failure in the elderly. J Am Geriatr

Soc 1991 August;39(8):810-825.

3 Januzzi JL, Jr., Camargo CA, Anwaruddin S, Baggish AL, Chen AA, Krauser DG, Tung

R, Cameron R, Nagurney JT, Chae CU, Lloyd-Jones DM, Brown DF, Foran-Melanson S,

Sluss PM, Lee-Lewandrowski E, Lewandrowski KB. The N-terminal Pro-BNP investiga-

tion of dyspnea in the emergency department (PRIDE) study. Am J Cardiol 2005 April

15;95(8):948-954.

4 Morrison LK, Harrison A, Krishnaswamy P, Kazanegra R, Clopton P, Maisel A. Utility of

a rapid B-natriuretic peptide assay in differentiating congestive heart failure from lung

disease in patients presenting with dyspnea. J Am Coll Cardiol 2002 January

16;39(2):202-209.

5 Smith H, Pickering RM, Struthers A, Simpson I, Mant D. Biochemical diagnosis of ven-

tricular dysfunction in elderly patients in general practice: observational study. BMJ

2000 April 1;320(7239):906-908.

6 Tombaugh TN, McIntyre NJ. The mini-mental state examination: a comprehensive re -

view. J Am Geriatr Soc 1992 September;40(9):922-935.

52

7 Krum H. The Task Force for the diagnosis and treatment of chronic heart failure of the

European Society of Cardiology. Guidelines for the diagnosis and treatment of chronic

heart failure: full text (update 2005). Eur Heart J 2005 November;26(22):2472-

2474.

8 Shamsham F, Mitchell J. Essentials of the diagnosis of heart failure. Am Fam Physician

2000 March 1;61(5):1319-1328.

9 Bruins S, Fokkema MR, Romer JW, Dejongste MJ, van der Dijs FP, van den Ouweland

JM, Muskiet FA. High intraindividual variation of B-type natriuretic peptide (BNP) and

amino-terminal proBNP in patients with stable chronic heart failure. Clin Chem 2004

November;50(11):2052-2058.

10 Quinones MA, Waggoner AD, Reduto LA, Nelson JG, Young JB, Winters WL, Jr.,

Ribeiro LG, Miller RR. A new, simplified and accurate method for determining ejection

fraction with two-dimensional echocardiography. Circulation 1981 October;64(4):744-

753.

11 Mosterd A, Hoes AW, de Bruyne MC, Deckers JW, Linker DT, Hofman A, Grobbee DE.

Prevalence of heart failure and left ventricular dysfunction in the general population; The

Rotterdam Study. Eur Heart J 1999 March;20(6):447-455.

12 Heckman GA, Misiaszek B, Merali F, Turpie ID, Patterson CJ, Flett N, McKelvie RS.

Management of heart failure in Canadian long-term care facilities. Can J Cardiol 2004

August;20(10):963-969.

53

13 Mair J. Monitoring of patients with heart failure. Scand J Clin Lab Invest Suppl

2005;240:99-106.

14 Rutten FH, Cramer MJ, Grobbee DE, Sachs AP, Kirkels JH, Lammers JW, Hoes AW.

Unrecognized heart failure in elderly patients with stable chronic obstructive pulmonary

disease. Eur Heart J 2005 September;26(18):1887-1894.

15 Cowie MR, Jourdain P, Maisel A, Dahlstrom U, Follath F, Isnard R, Luchner A,

McDonagh T, Mair J, Nieminen M, Francis G. Clinical applications of B-type natriuretic

peptide (BNP) testing. Eur Heart J 2003 October;24(19):1710-1718.

16 Bando M, Ishii Y, Sugiyama Y, Kitamura S. Elevated plasma brain natriuretic peptide

levels in chronic respiratory failure with cor pulmonale. Respir Med 1999

July;93(7):507-514.

17 Buckley MG, Sethi D, Markandu ND, Sagnella GA, Singer DR, MacGregor GA. Plasma

concentrations and comparisons of brain natriuretic peptide and atrial natriuretic peptide

in normal sub jects, cardiac transplant recipients and patients with dialysis-independent

or dialysis-dependent chronic renal failure. Clin Sci (Lond) 1992 October;83(4):437-

444.

18 Sayama H, Nakamura Y, Saito N, Kinoshita M. Why is the concentration of plasma brain

natriuretic peptide in elderly inpatients greater than normal? Coron Artery Dis 1999 Octo

ber;10(7): 537-540.

19 Wallen T, Landahl S, Hedner T, Saito Y, Masuda I, Nakao K. Brain natriuretic peptide in

an elderly population. J Intern Med 1997 October;242(4):307-311.

54

20 Kohno M, Yokokawa K, Yasunari K, Kano H, Minami M, Hanehira T, Yoshikawa J.

Changes in plasma cardiac natriuretic peptides concentrations during 1 year treatment

with angiotensin-con verting enzyme inhibitor in elderly hypertensive patients with left

ventricular hypertrophy. Int J Clin Pharmacol Ther 1997 January;35(1):38-42.

21 Missouris CG, Grouzmann E, Buckley MG, Barron J, MacGregor GA, Singer DR. How

does treatment influence endocrine mechanisms in acute severe heart failure? Effects on

cardiac natriuretic peptides, the renin system, neuropeptide Y and catecholamines. Clin

Sci (Lond) 1998 June;94(6):591-599.

22 Redfield MM, Rodeheffer RJ, Jacobsen SJ, Mahoney DW, Bailey KR, Burnett JC, Jr.

Plasma brain natriuretic peptide concentration: impact of age and gender. J Am Coll

Cardiol 2002 September 4;40(5):976-982.

23 Mair J, Friedl W, Thomas S, Puschendorf B. Natriuretic peptides in assessment of left-

ventricular dysfunction. Scand J Clin Lab Invest Suppl 1999;230:132-142.

24 Gambassi G, Lapane K, Sgadari A, Landi F, Carbonin P, Hume A, Lipsitz L, Mor V,

Bernabei R. Prevalence, clinical correlates, and treatment of hypertension in elderly

nursing home residents. SAGE (Systematic Assessment of Geriatric Drug Use

via Epidemiology) Study Group. Arch Intern Med 1998 November

23;158(21):2377-2385.

25 Kohno M, Yokokawa K, Yasunari K, Kano H, Minami M, Hanehira T, Yoshikawa J.

Changes in plasma cardiac natriuretic peptides concentrations during 1 year treatment

with angiotensin-con verting enzyme inhibitor in elderly hypertensive patients with left

ventricular hypertrophy. Int J Clin Pharmacol Ther 1997 January;35(1):38-42.

55

Table 1. Three chronic heart failure groups found after the study.

n nyha NT-proBNP BNP I II III IV CHF confirmed 9 1 5 3 0 2632 (1044-5737) 118 (59-706) CHF rejected 13 7 5 1 0 388 (174-719) 90 (35-128) CHF de novo 15 7 4 1 3 1146 (228-3341) 200 (107-433) Chronic heart failure, CHF; CHF diagnosed before the study and confirmed afterwards, CHF

confirmed; CHF diagnosed before the study and rejected afterwards, CHF rejected; CHF unde-

tected before the study and diagnosed afterwards, CHF de novo; number, n; New York Heart

Association, nyha; N-terminal-pro B-type natriuretic peptide, NT-proBNP; B-type natriuretic

peptide, BNP; median and interquartile range of pg/mL.

56

Table 2. Baseline characteristics of nursing home residents.

Total CHF+ (%) CHF- (%) P age male (%) immobility wheelchair/bed (%) medical history Hypertension (%) CAD (%) rhythm disorder (%) Diabetes type 2 (%) PAD (%) Medication ACE and ARB (%) Beta blockers (%) Ca-antagonists (%) Anti-coagulants (%) Diuretics (%) Anti-diabetics (%) Major symptoms Fatigue (%) Dyspnea Physical examination vCJP Cardiomegaly Murmur 3th tone, gallop Rales Hepatomegaly edema

78±11 39 (38)

40 (39)

57 (55) 28 (27) 24 (23) 26 (25) 16 (16)

42 (41) 24 (23) 10 (10) 63 (61) 50 (49) 21 (20)

40 (39) 31 (30)

11 (11) 22 (21) 24 (23) 20 (19) 17 (17) 4 (4)

40 (39)

80±9 11 (46)

13 (54)

13 (54) 14 (58) 9 (38) 9 (38) 66 (25)

14 (58) 7 (29) 4 (17) 19 (78) 15 (62) 8 (33)

15 (52) 11 (45)

7 (29) 10 (42) 9 (38) 10 (20) 9 (40) 1 (4)

15 (60)

78±11 28 (35)

27 (34)

44 (49) 14 (18) 15 (19) 17 (22) 10 (13)

28 (35) 17 (22) 6 (8)

44 (53) 35 (44) 13 (16)

25 (32) 20 (25)

4 (5)

12 (15) 25 (32) 10 (13) 8 (8) 3 (3)

35 (44)

0.351 0.358

0.078

0.680 <0.001 0.060 0.114 0.144

0.046 0.438 0.189 0.055 0.118 0.072

0.007 0.055

<.001 0.006 0.587 0.587 <0.001 0.935 0.116

57

systolic blood pressure diastolic blood pressure body mass index Electrocardiogram atrial fibrillation ischemia LVH Laboratory Hemoglobin (SD) eGFR (SD) NT-proBNP md IQR BNP md IQR echocardiogram mn SD

141±22 81±13 25± 5

10 (10) 5 (5)

32(31)

7.7±1 76±32

324 (163-1146) 87 (28-187)

51±9

143 (27) 79 (13) 25 (5)

3 (13) 2 (8)

16(67)

7.7±1 63±26

1871(539-4262) 194 (92-460)

43±9

140 (20) 81 (13) 25 (5)

7 (8) 3 (28) 16(20)

7.7±1 80±32

239 (118-674) 68 (20-123)

54±6

0.534 0.590 0.897

0.598 0.365 <0.001

0.834 0.022 <0.001 <0.001 0.016

Coronary artery diseases, CAD; peripheral artery disease, PAD; angiotensin converting enzyme inhibitor, ACE-inhibitor and alfa 2-receptor blocker, ARB; anti-coagulants and anti-platelets, anti-coagulants; ve-nous central jugular pressure, vCJP; left ventricular hypertrophy, LVH; estimated glomerular filtration rate, eGFR; N-terminal-pro B-type natriuretic peptide, NT-proBNP; B–type natriuretic peptide, BNP; md, median IQR, interquartile range; SD, standard deviation.

58

Table 3. Median NT-proBNP and BNP concentrations in residents with and without

chronic heart failure (CHF + and CHF -), before and after the study.

NT-proBNP CHF + CHF - P-value

median (IQR) (n) median (IQR) (n)

before study

after study

552 (309-2477) (22)

1871(539-4262) (24)

241 (128-892) (81)

239 (118-674) (79)

0.013

<0.001

BNP median (IQR) (n) median (IQR) (n)

before study 92 (33-187) (22) 76 (23-187) (81) 0.270

after study 194 (92-460) (24) 68 (20-123) (79) <0.001

Interquartile range, IQR; number of subjects, n; natriuretic peptides in pg/mL; Before the study, medians of NT-proBNP in residents with and without CHF (CHF+ and CHF -) differed

(P=0.013) and medians of BNP in CHF+ and CHF- did not differ (P=0.270). After the study,

medians of both NT-proBNP and BNP differed convincingly (NT-proBNP in CHF+ 1871 and in

CHF – 239, P<0.001, BNP in CHF+ 194 and in CHF – 68 pg/mL, P<0.001).

59

Table 4. N-terminal-pro B-type natriuretic peptide (NT-proBNP) and B-type natriuretic

peptide (BNP) cut-off values with positive and negative predictive values.

Cut-off pg/mL Sensitivity Specificity PPV NPV

NT-proBNP 450 0.71 0.67 0.42 0.91

NT-proBNP 900 0.67 0.85 0.56 0.89

NT-proBNP 1350 0.58 0.86 0.56 0.86

NT-proBNP 1800 0.46 0.94 0.69 0.85

Cut-off pg/mL Sensitivity Specificity PPV NPV

BNP 50 0.88 0.63 0.42 0.89

BNP 100 0.71 0.70 0.63 0.89

BNP 150 0.54 0.79 0.69 0.82

BNP 200 0.50 0.87 1.00 0.80

PPV , positive predictive value; NPV, negative predictive value.

60

61

Chapter 2.2

UNDERESTIMATED PREVALENCE OF CHRONIC HEART FAILURE

AMONG RESIDENTS OF CARE-HOMES IN ARUBA. Stoutjesdijk E, Brouns RM, Barents M, DeJongste MJ, Besselink HJ, Cheng JD,

Wever R, Muskiet FA.

West Indian Med J. 2014 Nov 4;63(6). doi: 10.7727/wimj.2013.277.

62

ABSTRACT

OBJECTIVES. We aimed to study the frequency and accuracy of chronic heart failure diagno-

ses in a group of stable Aruban residential elderly with low physical activity. METHODS. A total of 235 residents were assessed in a cross-sectional designed pilot study.

Residents with dementia, aphasia, a history of psychiatry or terminal illness were excluded.

Physical examination and electrocardiograms were performed in conjunction with blood samples

to determine a./o. B-type natriuretic peptides (BNP) and renal function. A general practitioner

and a cardiologist established or withdraw the diagnosis CHF. RESULTS. Out of, 235 residents, 184 were excluded. Fifty-one (22%) subjects, 78±8 years of

age, were included in the study. CHF was established in 16 [BNP 156 (72-1029) ng/L] out of

the 51 residents. Eleven residents were not previously diagnosed with CHF [BNP 152 (61-

1029) ng/L]. Of two out of the seven residents diagnosed with CHF before the screening, the

diagnosis CHF was rejected [BNP 71 (59-83) ng/L].

CONCLUSION. In Aruban residential elderly, identification of chronic heart failure is severely

underestimated. The detection of CHF was improved by implementing an appropriate guideline

and by including the determination of BNP.

63

INTRODUCTION

During the last one-and-half-century, in the industrialized society the average age of dying has

doubled. This observation is mainly resulting from introduction of hygienic measures (sewers,

water supply, food, life style) and improved treatments such as for ischemic heart diseases. The

drawback of increased survival from cardiac diseases is a subsequent growth in sequelae, such

as heart failure. In addition, the onset of cardiovascular diseases is adversely affected by the

development of an unhealthy lifestyle, as is the case in more prosperous societies. As a result

of an insalubrious life style, the prevalence is raising of acquired afflictions like hypertension,

diabetes, obesity and hypercholesterolemia more often. In this regard it is worth notifying that

Mexican Americans, to which Arubans belong develop atherosclerosis much more frequently

most likely because of an imbalance between intake and leisure-time physical activities (1;2).

Hence, in Arubans cardiovascular morbidity has become a substantial part of the multimorbidity

in the elderly. Multimorbidity is related to care dependency and more help in activities of daily

living (3). Accordingly, the most care dependent elderly live in a nursing home, here referred to

as residential elderly. It is unknown however likely that a substantial part of Aruban residential

elderly suffer from CHF. In Canada and The Netherlands CHF prevails in one to two% of all

adults, ten% in a community dwelling elderly and 20% in residential elderly (4-6). The life-time

risk to develop CHF is estimated to be 20%. The prognosis of CHF greatly depends on the se-

verity of symptoms (7). For instance, persons with CHF of NYHA class III-IV have an one-year

mortality of 60%, those with NYHA class I-II have 20% and half of the latter is deceased within

five years (8-10).

Heart failure is a syndrome defined as the presence of at least one symptom and a heart disor-

der. Key symptoms are dyspnea and fatigue, both at rest and during exertion. Heart failure may

be caused by hypertension, atherosclerotic heart diseases, valve and rhythm disorders and a

variety of cardiomyopathies. When CHF is accompanied by pressure or volume overload, na-

triuretic peptides, as B-type Natriuretic Peptide (BNP) secretion by the ventricle wall is one of

the regulating mechanisms. BNP stimulates natriuresis, diuresis, vasodilatation and inhibition of

64

the renin-angiotensin-aldosterone system - and sympathetic nervous system. BNP is cleared by

binding to natriuretic peptide receptors and the half-life time of this binding is 20 minutes. The-

se characteristics make BNP appropriate as marker of CHF. However, many factors increase

the secretion of BNP as well (11;12).

In residential elderly, it is even more difficult to identify subjects with CHF since there are no

CHF guidelines for residential elderly (4;13;14). Moreover, BNP is not validated for elderly ≥ 75

years or for elderly with multimorbidity (15). Finally, access to echocardiography is limited. All

these confounders have attributed to regular misjudgment of CHF in Dutch community-dwelling

and residential elderly (4;13). In Aruba, it can be predicted that in a population of residential

elderly, with a presumably higher incidence of not recognized CHF, comparable difficulties are

met. Therefore, we aim to investigate the diagnostic accuracy and role of BNP in diagnosing

CHF in Aruban residential elderly with low physical activity.

METHOD

Design: this study was designed as multi-centre cross-sectional pilot study.

Persons: from February till August 2009, 235 residents live in the “Stichting Algemeen

Bejaardenzorg Aruba” located in San Nicolas, Savaneta and Oranjestad. Half the residents were

immobile for the most part because they had an amputation. Some were admitted because of

lacking family care. All residents were invited to participate in the study. They were included if

they signed the informed consent. They were excluded in case of cognitive impairments as de-

mentia (Korsakov’s syndrome), incompetency (one does not understand the impact of the study

on himself), aphasia, serious psychiatric and neurological disorders and terminal illness.

From the included residents, one of the investigators collected data on medical history and

medication. One general practitioner examined the residents on CHF symptoms. He also meas-

ured blood pressure and heart rate. He examined the central jugular pressure, cardiac size

(percussion) and heart rhythm. He also looked for the presence of murmurs, deviate breath

65

sounds and peripheral edema. A 12-leads electrocardiogram was made at rest (Burdick Eclipse

LE (www.Burdick.com)). The general practitioner and cardiologist decided on the diagnosis

CHF, independently. The general practitioner suggested the diagnosis CHF based on medical

history, physical examination, electrocardiogram and blood tests except BNP. The cardiologist

came to the diagnosis CHF based on the same data as the general practitioner but now with

BNP included. They discussed all cases and in case of disagreement, the cardiologist prevailed.

All data were retrieved anonymously from the resident files.

According to CHF guidelines (16;17), BNP is determined when CHF is suspected. When BNP is

lower than 100ng/L and the electrocardiogram is not deviating, CHF can be excluded in most

cases leaving a small risk of still having CHF of less than 10% (5;18).

Laboratory assistants took non-fasting blood samples in tubes prepared with lithium heparin and

EDTA. They stored all samples in a cooler and analyzed them within four hours only in the La-

boratory. Sodium, potassium, urea, kreatinin, thyroid-stimulating hormone, BNP, hemoglobin,

haematocrit and mean corpuscular volume were determined.

BNP was determined with the Triage BNP assay (www.biosite.com) at a Beckman Coulter

Unicel DxC600i Immunoassay System (www.beckmancoulter.com). BNP had coefficients of

variation of <7% with a range of 5 to 4970ng/L. Kreatinin was measured by a standardized

isotope dilution mass spectrometry. Renal function was defined by the estimated glomerular-

filtration rate (eGFR) measured by the Modification of Diet in Renal Disease formula

(mL/min/1.73 m2) = 175 * (Sc/88.4)-1.154 * (age)-0.203 * 0,742 if female *1.212 if negroïd.

Serious renal dysfunction was defined as eGFR less than 30 mL/min/1.73m2 (19;20).

We requested the laboratory to deliver us the percentage of BNP tests ordered by general prac-

titioners during the study period.

Statistical analyses were performed using PASW Statistics 17 software. We compared the char-

acteristics of residents with and without CHF with the Student’s t-test and Mann-Whitney test for nonparametric continuous data, as appropriate (Table 2). The x2- test was used in parametric

and the Mann-Whitney test in nonparametric binominal variables. The predictive values of the

66

BNP test was calculated at different cut-off points from 50 – 200 ng/L as described in Table

3. Influences of age, gender, angiotensin-converting enzyme inhibitor (ACE-i) and eGFR on

BNP were examined with linear regression analyses with a log transformation. All statistical

comparisons were two-tailed, and a p value less than 0.05 was considered to be statistically

significant.

RESULTS

Out of all 235 residents, 184 (78%) were excluded for a variety of reasons such as dementia in

half of them. Out of 184 residents, 165 were not eligible according the exclusion criteria. And,

19/184 dropped out because of refusal to participate (eight), cognitive impairments (five), ad-

mission to a hospital beforehand (four), incomplete data (one), and for not living in a nursing

home (one). The remaining 51 (22%) were included and screened.

The included residents were aged 78 (range 56-93) years and 29 were females with 22

males. The majority (55%) of the individuals was immobilized and bound to wheelchair or bed

(55%), as a consequence of various disorders, such as cerebrovascular accidents (31%), CHF

(31%) and serious sequelae of T2DM (59%), like amputation of a leg. The estimated GFR was

46±22 mL/min.

Chronic heart failure was diagnosed in sixteen residents (16/51, 31%). We distinguished four

groups: seven residents were judged to have CHF before the screening (CHF before study). In

two of these seven, the CHF diagnoses were not confirmed (CHF rejected). In five of these

seven, the CHF diagnoses were confirmed (CHF confirmed). In addition, eleven CHF diagno-

ses were first established after the screening while they were undetected before (CHF de novo) (Figure 1 and Table 2) The cardiologist overruled the general practitioner in 4 out of 16 resi-

dents in which there was disagreement with regard to CHF (BNP 134, 392, 440, 459ng/mL).

67

Comparing the 16 residents with CHF to those without, the first group used more anticoagulants

and suffered more from dyspnea, fatigue and edema. In addition, the CHF group had higher

BNP levels and a lower renal function, on average.

Regarding the ECG observations, residents with CHF had more LVH and atrial fibrillation com-

pared to those without CHF. And, within the other ECG disorders, residents with and without

CHF contained two and four atrial-ventricular conduction disorders, zero and two rhythm disor-

der and seven and six implanted artificial cardiac pacemaker, respectively and none significant-

ly. One resident with CHF had LVH and an atrial-ventricular conduction disorders.

The predictive values of BNP, calculated at different cut-off points from 50 to 200ng/L with

each 25 ng/L difference in between, were presented in Table 3.

Of the included confounders, we found renal function inversely related to BNP (β= -0.49,

p=0.002) while age, gender and ACE-i did not influence BNP levels, significantly (p=0.452,

p=0.314, p=0.787, respectively). During the study period, general practitioners ordered 6-15%

of all ordered BNP tests.

DISCUSSION

In Aruban residential elderly, the CHF frequency was unknown and undetected CHF was pre-

sumed because of several reasons. We aimed to investigate the frequency and diagnostic accu-

racy in diagnosing CHF in that population. Therefore, all eligible residential elderly were

screened on CHF in three nursing homes. We found that the CHF frequency was about one-

third of the residents. And, that in the majority of them the diagnosis of CHF had not been made

previously. Undetected diagnoses were expected but the number exceeded our expectations.

Reasons for undetected CHF diagnoses were the lack of consensus on the sequence of the

diagnostic procedures, incomplete data documentation in the medical files and limited use of

BNP testing by general practitioners.

68

The main etiology of CHF existed of atherosclerosis. Arguments for this assumption were high

coexistence of T2DM in residents with CHF and physical inactive lifestyle.

Quite a percentage of residential elderly had T2DM which was well documented. This occur-

rence corresponded to Palloni et al. who found an high or higher frequency of T2DM in residen-

tial elderly of 60+ years living in the Caribbean compared to those from the U.S.A.. And, the

high T2DM frequency corresponded to a study in the Netherland Antilles. In that study, T2DM

was also related to gender (more females), social status and a Western lifestyle.

Near the high T2DM frequency, the studied population was characterized, surprisingly, by a

residents’ mean age comparable to the Aruban life expectancies of 75 and 76 years in 2009

and 2011, respectively (www.indexmundi.com). So, it is likely that our population represented

the Aruban elderly as regards T2DM occurrence and age. However, our population represented

only the Aruban residential elderly as regards the distribution over the Island.

Although the population was too small to speak of predictive values of BNP, the calculated sen-

sitivity and specificity were in line with a comparable study on CHF diagnosis in NH residents of similar age (4) (Table 3). And, from the estimated confounders, BNP was only inversely influ-

enced by renal function which was to be expected from the moderate to severe renal dysfunc-

tion (23;25). Thus, these findings do not limit the additional value of BNP to the CHF diagnos-

tics in residential elderly.

The sense of improving CHF diagnoses for residential elderly was a better quality of life. Since,

more correct CHF diagnoses will lead to adequate treatment, ADL help and planning of care

givers (3;26). And, from another view, less unjust CHF diagnoses save unnecessary treatments

and costs.

Altogether, we observed that the carefully medical work-up according to the applied guideline

improved the CHF diagnostics in residential elderly (17).

This study had some shortcomings. The studied residents did not represent all since widely half

of them were demented and the characteristics of all excluded could not have been uncovered

because of ethical reasons. However, the studied residents represented Aruban elderly as re-

69

gards T2DM occurrence and age. Therefore, we could not rule out a selection bias. So, we cit-

ed a frequency of CHF rather than prevalence. The lack of a panel as standard of diagnosing

CHF was another shortcoming. This could have influenced the CHF frequency but to a limited

extent because the cardiologist was specialized in heart failure and the diagnosis CHF was

made up according to the guideline including BNP (17). Main limitation of this pilot study was

the small population.

In the future, a study is recommended on the detection of CHF in a representative cohort of

residential elderly.

CONCLUSION

In Aruban residential elderly, identification of chronic heart failure is severely underestimated.

The detection of CHF was improved by implementing an appropriate guideline including deter-

mination of BNP.

ACKNOWLEDGMENTS

We thank Stichting Algemeen Bejaardenzorg Aruba, dr. Horacio E. Oduber Hospitaal and Cen-

tro Medico: dr. Rudy Engelbrecht for their participation in this study. None of the authors re-

ported conflicts of interest.

70

References

1 Bielak LF, Yu P, Ryan KA, Rumberger JA, Sheedy PF, Turner ST, et al. Differences in

preva lence and severity of coronary artery calcification between two non-Hispanic

white populations with diverse lifestyles. Atherosclerosis 2008;196(2):888-95.

2 Kurian AK, Cardarelli KM. Racial and ethnic differences in cardiovascular disease risk

factors: a systematic review. Ethn Dis 2007;17(1):143-52.

3 Barents M, Boersma F, Van Der Horst ICC, Hillege HH, de Jongste MJL. Increase of

Help at Activities of Daily Life Given to Frail Elderly With Chronic Heart Failure in a

Nursing Home. Activities, Adaptation & Aging 2011;35(2):98-110.

4 Barents M, van der Horst IC, Voors AA, Hillege JL, Muskiet FA, de Jongste MJ.

Prevalence and misdiagnosis of chronic heart failure in nursing home residents: the role

of B-type natriuretic peptides. Neth Heart J 2008;16(4):123-8.

5 Rutten FH, Cramer MJ, Grobbee DE, Sachs AP, Kirkels JH, Lammers JW, et al. Un rec-

ognized heart failure in elderly patients with stable chronic obstructive pulmonary dis

ease. Eur Heart J 2005;26(18):1887-94.

6 Heckman GA, Misiaszek B, Merali F, Turpie ID, Patterson CJ, Flett N, et al. Manage

ment of heart failure in Canadian long-term care facilities. Can J Cardiol

2004;20(10):963-9.

7 Barents M, Hillege HH, van der Horst IC, de Boer RA, Koster J, Muskiet FA, et al. BNP

and NT-proBNP, predictors of 1-year mortality in nursing home residents. J Am Med Dir

Assoc 2008;9(8):580-5.

71

8 Cowie MR, Fox KF, Wood DA, Metcalfe C, Thompson SG, Coats AJ, et al. Hospitaliza-

tion of patients with heart failure: a population-based study. Eur Heart J

2002;23(11):877-85.

9 Bleumink GS, Knetsch AM, Sturkenboom MC, Straus SM, Hofman A, Deckers JW, et

al. Quantifying the heart failure epidemic: prevalence, incidence rate, lifetime risk and

prognosis of heart failure The Rotterdam Study. Eur Heart J 2004;25(18):1614-9.

10 Ho KK, Anderson KM, Kannel WB, Grossman W, Levy D. Survival after the onset of

congestive heart failure in Framingham Heart Study subjects. Circulation

1993;88(1):107-15.

11 Januzzi JL, Jr. Natriuretic peptide testing: a window into the diagnosis and prognosis

of heart failure. Cleve Clin J Med 2006;73(2):149-7.

12 Beleigoli A, Diniz M, Nunes M, Barbosa M, Fernandes S, Abreu M, et al. Reduced

brain natriuretic peptide levels in class III obesity: the role of metabolic and cardiovas-

cular factors. Obes Facts 2011;4(6):427-32.

13 Rutten FH, Grobbee DE, Hoes AW. Differences between general practitioners and car-

diologists in diagnosis and management of heart failure: a survey in every-day

practice. Eur J Heart Fail 2003;5(3):337-44.

14 Barents M, van der Horst IC, Voors AA, Hillege JL, Muskiet FA, de Jongste MJ.

Prevalence and misdiagnosis of chronic heart failure in nursing home residents: the role

of B-type natriuretic peptides. Neth Heart J 2008;16(4):123-8.

72

15 Voors AA, Walma EP, Twickler TB, Rutten FH, Hoes AW. [Multidisciplinary guideline

'Heart failure 2010']. Ned Tijdschr Geneeskd 2011;155:A2957.

16 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 endorsed by the European

Society of Intensive Care Medicine (ESICM). Eur Heart J 2008;29(19):2388-442.

17 van LJ, Rutten FH, Walma EP, Wiersma T, Goudswaard AN. [Summary of the prac-

tice guide line 'Heart failure' (first revision) from the Dutch College of General Practitio-

ners]. Ned Tijdschr Geneeskd 2005;149(48):2668-72.

18 Cleland JG, Swedberg K, Follath F, Komajda M, Cohen-Solal A, Aguilar JC, et al. The

EuroHeart Failure survey programme- a survey on the quality of care among patients

with heart failure in Europe. Part 1: patient characteristics and diagnosis. Eur Heart

J 2003;24(5):442-63.

19 Kenchaiah S, Evans JC, Levy D, Wilson PW, Benjamin EJ, Larson MG, et al. Obesity

and the risk of heart failure. N Engl J Med 2002;347(5):305-13.

20 Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, III, Feldman HI, et al. A new

equation to estimate glomerular filtration rate. Ann Intern Med 2009;150(9):604- 12.

73

21 Palloni A, McEniry M. Aging and health status of elderly in Latin America and the Car-

ibbean: preliminary findings. J Cross Cult Gerontol 2007;22(3):263-85.

22 Grievink L, Alberts JF, O'Niel J, Gerstenbluth I. Waist circumference as a measure

ment of obe- sity in The Netherlands Antilles; associations with hypertension and dia-

betes mellitus. Eur J Clin Nutr 2004;58(8):1159-65.

23 DeFilippi C, van Kimmenade RR, Pinto YM. Amino-terminal pro-B-type natriuretic

peptide test ing in renal disease. Am J Cardiol 2008;101(3A):82-8.

24 deFilippi CR, Christenson RH. B-type natriuretic peptide (BNP)/NT-proBNP and renal

function: is the controversy over? Clin Chem 2009;55(7):1271-3.

25 van Kimmenade RR, Januzzi JL, Jr., Bakker JA, Houben AJ, Rennenberg R, Kroon

AA, et al. Renal clearance of B-type natriuretic peptide and amino terminal pro-B-type

natriuretic peptide a mechanistic study in hypertensive subjects. J Am Coll Cardiol

2009;53(10):884-90.

74

51 included

2 CHF rejected 5 CHF confirmed After the study 11 CHF de novo

184 excluded

7 CHF before the study

16 (31%) CHF 35 (69%) no CHF

Figure 1. Flow chart of included and excluded residents, and of CHF diagnoses before

and after the study in Aruba.

Chronic heart failure, CHF.

235 residents

75

Table 1. Characteristics of 51 residents.

CHF+ n=16

(%) CHF- n=35

(%) p-value

Age (years) mn (SD) 79 (9) 78 (8) 0.460 Female 11 69 18 51 0.248 Wheelchair/bed 7 44 21 60 0.288 Medical history Cerebrovascular accident 4 25 12 32 0.517 Coronary artery disease 2 13 1 3 0.298 Diabetes Mellitus type 2 12 75 18 54 0.151 Medication Antidiabetics 12 75 14 40 0.018 Diuretics 8 50 11 26 0.211 Calcium-ion antagonists 2 13 6 23 0.680 Beta-blockers 4 25 4 9 0.287 ACE-i or AT2Ri 6 38 10 29 0.533 Heart glycosides 2 13 3 9 0.665 Anticoagulants 13 81 19 37 0.049 Symptoms, signs Dyspnea 13 81 11 31 0.020 Fatigue 14 87 19 54 0.009 Edema 8 50 5 14 0.020 Systolic tension mm Hg 135 28 133 17 0.782 Diastolic tension mmHg 72 16 76 10 0.378 Heart rate beats/min. 74 11 79 9 0.133 Increased CVP 10 63 5 14 0.002 Cardiomegaly 2 13 0 0 0.164 Pacemaker 3 19 2 13 0.243 Electrocardiogram Normal electrocardiogram 3 19 19 54 0.011

76

Left ventricular hypertrophy 4 25 0 0 0.041 Atrial fibrillation 2 0 2 0 0.413 Ischemia 2 0 4 0 0.917 Other 6 56 12 34 0.343 Laboratory data BNP ng/L md (range) 156 72-1029 59 9-191 0.008 Hb mmol/L mn (SD) 7(1) 8(1) 0.290 MCV femtoL mn (SD) 91 (6) 90 (6) 0.763 eGFR mL/min/1.73m² n(SD) 33(20) 58(24) 0.001 TSH mE/L mn (SD) 1(1) 2(1) 0.346

Numbers, n and percentage unless other indicated; mn, mean and SD, standard deviation ;

CVP, central jugular pressure; ACE-I, angiotensin-converting enzyme inhibitor; AT2Ri, angio-

tensin II receptor inhibitor; other, rhythm or conduction disorder, pacemaker or no-specified dis-

order.

77

Table 2. CHF diagnoses after the screening with matching BNP values.

N residents B-type natriuretic peptide median (ng/L) chronic heart failure rejected 2 71 (59 – 83)

chronic heart failure confirmed 5 105 (84 – 392)

chronic heart failure de novo 11 152 (61 – 1029)

Table 3. Predictive values of BNP at different cut-off points. BNP cut-off point ng/L sensitivity specificity PPV NPV 50 1.00 0.37 0.42 1.00 75 1.00 0.64 0.54 1.00 100 0.75 0.69 0.52 0.86 125 0.63 0.74 0.53 0.81 150 0.56 0.89 0.69 0.82 200 0.44 1.00 1.00 0.80 PPV, positive predictive value; NPV, negative predictive value; cut-off points are rounded num-bers.

78

79

Chapter 2.3

HIGH INTRA-INDIVIDUAL VARIATION OF AMINO-TERMINAL PRO-

BRAIN NATRIURETIC PEPTIDE (NT-PROBNP) IN URINE OF PA-

TIENTS WITH STABLE CHRONIC HEART FAILURE.

COMPARISON WITH PLASMA. Schimmel AM, Barents M, de Jongste MJL, Römer JWP, Steward HN, Muskiet FAJ.

Accepted for publication in part in Clinical Chemistry.

80

ABSTRACT

BACKGROUND. The large intra-individual variation (CVi) of plasma N-terminal prohormone-

brain natriuretic peptide (NT-proBNPp) limits its applicability for guided therapy optimization in

individual patients with chronic heart failure (CHF). We investigated whether urine NT-proBNP

(NT-proBNPu) has lower CVi compared with NT-proBNPp. METHODS. CVis were measured in 25 patients with stable CHF (NYHA classes I-III). Blood

and urine samples were collected on a single day (6 2-h blood samples and spontaneously

voided urines during 24-h), 5 consecutive days (5 blood samples, 5 enforced urine samples, 5

full 24-h urines) and 6 consecutive weeks (6 blood samples, 6 enforced urine samples, 6 full

24-h urines). NT-proBNP was measured by immunoassay. Within-day, day-to-day and week-

to-week total CVs (CVts), CVis and reference change values (RCVs) were calculated. RESULTS. Median age was 61 y (range 36-80), 60% was male and mean±SD left ventricular

ejection fraction was 36±15%. Median CVis for NT-proBNPp were 9 (within-day), 18 (day-to-

day) and 30% (week-to-week). For NT-proBNPu they were 34, 21 and 28%, respectively.

RCVs of NT-proBNPu, NT-proBNPu/creatinineu and NT-proBNPu/h were higher than, or equal

to, those of NT-proBNPp. NT-proBNPu did not correlate with NT-proBNPp up to 310 pg/mL NT-

proBNPp, but related almost linearly from thereon. NT-proBNPp increased during the day, while

NT-proBNPu tended to decrease. CONCLUSION. Measurement of NT-proBNP in urine provides no advantages over plasma NT-

proBNP for therapy optimization in CHF patients.

81

INTRODUCTION

Improvements in diagnostics and treatment have contributed considerably to the increasing sur-

vival of patients with cardiac diseases (1). As a consequence, also the prevalence and hence

the impact on society of chronic heart disease, notably chronic heart failure (CHF), has grown

tremendously (2). CHF has become a major health care issue in therapy, hospital admissions

and costs. Measurements of biomarkers may favourably affect the cost-benefit ratio. An ideal

biomarker has high diagnostic sensitivity and specificity, should be a true reflection of

(patho)physiological changes, has low analytical variation (CVa), and is applicable in

(sub)groups irrespective of their demographic background (3). Natriuretic peptides, notably

brain natriuretic peptide (BNP) and amino-terminal pro-brain NP (NT-proBNP) are well estab-

lished CHF biomarkers. They are widely used for CHF diagnosis, follow-up and prognosis (4),

while they also provide opportunities for biomarker-guided treatment optimization (5-8). Meta-

analyses have shown that natriuretic peptide (NP)-guided therapy is superior to symptom-

guided therapy in groups (5-7). However, because of high intra-individual variations compared

to expected changes, it is not possible to establish whether longitudinal changes of BNP or NT-

proBNP in individual patients differ with 95% certainty following therapy-adjustment (8).

BNP has natriuretic, diuretic and vasodilator properties (9;10). NT-proBNP becomes co-

secreted in equimolar amounts, but is physiologically inactive (11). Both are released by cardiac

atria and ventricles upon myocyte stretch, mostly due to volume overload (9). Like many, if not

all, hormones, natriuretic peptides exhibit pulsatile secretion patterns (12). Plasma BNP has a

half-life of about 20 min, while that of NT-proBNP is about 120 min. BNP is rapidly deactivated

in the circulation and tissues by neuropeptidases (10). In contrast, NT-proBNP becomes pas-

sively filtered in the kidneys to become excreted in urine, notably in the form of smaller peptides

that are at least in part immunoreactive (13). In the brush border of the renal tubules low-

molecular-weight proteins may become hydrolyzed to smaller peptides and amino acids that are

subsequently reabsorbed. Usually, this process is nearly complete, suggesting that only minor

82

amounts reach the finally voided urine (14;15). Another mechanism is by endocytosis followed

by lysosomal degradation, secretion via either apical or basolateral membranes, or by transcy-

tosis of the intact protein (16). With increasing plasma, and thereby filtered, concentrations all

enzymatic and active processes may ultimately be expected to reach saturation. From this

threshold, the renal clearance may be expected to increase to finally reach the glomerular filtra-

tion rate.

We have previously established that plasma NT-proBNP (NT-proBNPp) in patients with CHF

exhibits high within-day (mean CVi 9%; range 0-20%), day-to-day (CVi 20%; 3-80%) and

week-to week (CVi 35%; 8-103%) intra-individual variations (CVis) (17). This high CVi, calcu-

lated by correcting total variation (CVt) for analytical variation (CVa, 3%), hampers the useful-

ness of NT-proBNPp for treatment optimization of individual HF patients. Establishment of a

genuine difference between pre and post-treatment NT-proBNP concentrations requires infor-

mation of the minimal change value (also named reference change value; RCV). The RCV for

95% certainty is calculated by 2.8*CVt in which CVt is composed mostly of CVi in case of NT-

proBNPp (17).

Measurement of immunoreactive NT-proBNP in urine voidings (abbreviated NT-proBNPu) may

provide us with a tool that is less affected by the pulsatile secretion of NT-proBNP into plasma.

As a long term plasma ultrafiltrate, NTproBNPu may exhibit less intra-individual variation, at

least theoretically (18-20), but there may be many confounding factors. Measurements in urine

may especially be useful in CHF patients with NTproBNPp concentrations beyond the renal re-

absorption threshold. In this study we calculated CVis for NT-proBNP in both plasma and urine

of patients with stable CHF. To do so we collected corresponding plasma and urine samples for

the establishment of the within-day, day-to-day and week-to-week CVt, CVa, CVi and RCV. The

ultimate goal was to predict whether measurements in urine would allow treatment optimization

for the individual patient at the level of 95% certainty.

83

PATIENTS AND METHODS

The study was conducted from March–July 2008 in the island of Curaçao (12° 12' N, 68° 5'

W). The protocol was approved by the Medical Ethics Committee of the St. Elisabeth Hospital

and was in agreement with local ethical standards and the Helsinki Declaration of 1975, as re-

vised in 2008. All patients were assured that their data were anonymized. Written informed

consent was obtained from all of them.

Study design Within a study period of six weeks we collected blood and urine samples during home visits.

Participants were asked whether they preferred to participate in one, two or all sampling proto-

cols. These protocols were: 1) a within-day protocol; 6 blood samples (taken every 2-h starting

at 08:00 and ending at 18:00) and collection of spontaneously voided urine portions during 24-

h (starting 08:00 after discarding an enforced voiding and ending with a collected enforced

voiding at 08:00 the next day), 2) a day-to-day protocol; one blood sample (collected from

08:00 to 10:00), an enforced urine portion (immediately after blood sampling) and a full 24-h

urine taken during 5 consecutive days, and 3) a week-to-week protocol; one blood sample

(collected from 08:00 to 10:00 each week, an enforced urine portion (immediately after blood

sampling) and a full 24-h urine, on the same day of the week during 6 consecutive weeks.

Clock times of urine voidings were documented. The sampling protocols were integrated where

possible, to limit the number of sampling moments. For example, the 08:00 samples of the

within-day protocol were also part of the day-to-day protocol and samples of one sampling day of the day-to-day protocol was also part of the week-to-week protocol (Figure 1).

84

Patients CHF patients of 18 years and older were recruited from two cardiological practices. The inclu-

sion criteria were: stable CHF (defined as unchanged CHF treatment during the last two

months) living in the vicinity of the hospital and speaking Dutch or English. The exclusion crite-

ria were obstructive heart valve diseases, obstructive or restrictive heart diseases, potentially

transient causes of HF (e.g. acute myocarditis), myocardial infarction, unstable angina, uncon-

trolled arrhythmias and pregnancy or lactation. Information on date of birth, body mass index

and use of medication was collected at baseline. The most recent New York Heart Association

(NYHA) classes and the left ventricular ejection fractions (LVEF) were retrieved from the medi-

cal files. In addition to the NYHA classification, we established whether there was systolic or

diastolic CHF, defined as having a LVEF ≥45%, or >45%, respectively (21).

Samples and analyses Ten mL EDTA-anticoagulated blood samples were collected by venapuncture. They were stored

at 4°C until further handling in the laboratory. Urine voidings and full 24-h urines were collected

in containers and immediately stored in cool boxes. Blood samples were centrifuged within 2 h

after sampling (1,500 g-force, 15 min, 4 °C). Volumes of the spontaneous voidings, enforced

voidings and full 24-h urines were measured. Ten mL of each urine portion was taken and cen-

trifuged (1,500 g-force, 15 min, 4 °C). Plasma and urine samples were aliquoted and stored in

Eppendorf tubes at -80 °C until transportation to The Netherlands. NT-proBNP was analyzed

within nine months after collection in the Clinical Chemical Laboratory of the University Medical

Center Groningen, The Netherlands. Plasma and urine samples of a single patient were ana-

lyzed in a single series to minimize the influence of the CVa. For measurements of NT-proBNP

we applied the sandwich technique [Elecsys 2010 immunoassay system, Roche Diagnostics,

Mannheim, Germany (22)]. Creatinine was measured in all urine samples (creatinineu).

Creatinine was also measured in the firstly collected plasma sample of each patient to calculate

the estimated glomerular filtration rate (eGFR). In these calculations we used the factor of 1.112

85

for patients with African-American backgrounds (23). NT-proBNPu concentration (pg/mL) was

also expressed as NT-proBNPu/creatinineu (pg/nmol) and NT-proBNPu/h (pg/h). NT-

proBNPu/NT-proBNPp represents the apparent filtration fraction of NT-proBNP.

The detection limit of the NT-proBNP immunoassay was 5 pg/mL. An NT-proBNP <5 pg/mL

was documented as 0 pg/mL. The within-series CVa of NT-proBNPp was 3% (17). Since NT-

proBNPu ranges from about 5-50 pg/mL, we used 4 quality-control samples with NT-proBNPu

concentrations of 8, 10, 42 and 48 pg/mL. These samples were selected from the total patient

sample collection. They were analyzed six times to calculate the CVa of NT-proBNPu. The CVa

of creatinineu was calculated from all quality control measurements of creatinineu in the labora-

tory as performed in the month preceding the NT-proBNP analyses.

Statistical analysis Data of participants sampled three times or less within a sampling protocol were excluded from

evaluation in that specific protocol. The remaining data were processed with Microsoft Office

Excel 2007 and statistically analyzed using SPSS (SPSS 20.0; SPSS, Inc). Results were ex-

pressed as mean±SD or medians (range). The relations of NT-proBNPp and NT-proBNPu with

the patients’ characteristics were analyzed by Spearman’s correlation tests. Spearman’s correla-

tion tests were also employed for correlating NT-proBNPp with NT-proBNPu, using all data from

the day-to-day and week-to-week samples.

For comparison of within-day NT-proBNPu with NT-proBNPp, we employed the spontaneously

voided urine portion within 2 h following venapuncture. For example, the urine portion of 10:00

should be voided from 10:00 to 12:00, the portion of 12:00 from 12:00 to 14:00 etc. In case of

two spontaneously voided urine portions obtained within 2 h, the mean NT-proBNPu concentra-

tion of those samples was taken. If no urine was voided within 2 h, the NT-proBNPu of 10:00

was labelled as missing. The Shapiro-Wilk test was used to examine the distribution of NT-

86

proBNP at different clock times for normality. NT-proBNPp, NT-proBNPu and NT-proBNPu/NT-

proBNPp ratio at 8:00 and 18:00 were compared with Wilcoxon rank-sum test. NT-proBNP lev-

els at six different clock times within one day were analyzed for differences between those six

measurements using Friedman’s’ test. To identify differences between NT-proBNPp, NT-

proBNPu and NT-proBNPu/NT-proBNPp measurements from the same patient at different clock-

times, we used the Wilcoxon signed-rank test. The outcomes were corrected for type 1 errors

(p<0.01).

The plasma and urine total variations (CVt) of NT-proBNP for each patient and each of the 3

sampling protocols were calculated. CVi was calculated using (CVi)2 = (CVt)2 – (CVa)2 (17).

Reference change values (RCVs) at 95% probability were calculated from median CVas and

CVis, according to: RCV = Z × 21/2 × (CVa2/na + CVi

2/ns)1/2 (22;24;25), where Z is 1.96

(i.e. Z-score for 95% CI); na is the number of replicate assays; and ns the number of patient

samples needed to estimate each of the two homeostatic set points, ns=1 in the present study.

RCVs reflect the minimum percentage change in serial results that is different from the com-

bined analytical and biological variation, with 95% confidence. The CVis of the within-day, day-

to-day and week-to-week data were compared with the Wilcoxon signed-rank test. Only for

these comparisons the significance was set at p<0.01. All other significances were set at

p<0.05 (see above).

RESULTS

Patients and samples Twenty-five patients (mean age 61±10 years, 60% males) were included into the study (Table

1). Fifteen patients (60%) participated in all three sampling protocols. The main reasons for not

participating in all protocols were interference with work or holidays. All together, we were able

to collect samples from 20 patients participating in the within-day protocol, 18 patients following

87

the day-to-day protocol and 22 patients completing the week-to-week protocol (Figure 2). This

translated into 329 out of the maximum of 342 blood samples (96%), 187 spontaneously

voided urine portions, 212 of the maximum of 222 enforced urine voidings (95%), and 207 of

the maximum of 222 full 24-h urines (93%). Four participants (16%) exhibited NT-proBNPu

levels below the detection limit of 5 pg/mL. These levels were set at 0 pg/mL. In two of these

four patients, all NT-proBNPu values were set at 0 pg/mL. The NYHA classes of these two

patients were II and unknown. In the other two patients 89% and 50% of the NT-proBNPu val-

ues were below the detection limit and their NYHA classes were II and III, respectively.

Within-day, day-to-day and week-to-week intra-individual coefficients of variation

The calculated CVas of NT-proBNPu quality-control samples were 12% at 8 pg/mL, 15% at 10

pg/mL, 3% at 42 pg/mL and 3% at 48 pg/mL. The CVa of NT-proBNPu was inversely related

to NT-proBNPu (r= -0.983, p<0.05). The CVa for NT-proBNPu levels between 5 and 48

pg/mL was derived from CVa (%) = 119.67*(NT-proBNPu)-1.0016. For NT-proBNPu levels

above 48 pg/mL, a CVa of 3% was used. The CVa of creatinineu, calculated from all measure-

ments in a single month, was 2%.

The median CVis for NT-proBNPp in the within-day, day-to-day and week-to-week protocols were 9, 18 and 30%, respectively (Table 2). The corresponding median CVis for NT-proBNPu

were 34, 21 and 28%, respectively. The within-day CVi of NT-proBNPu was higher than the

corresponding CVi of NT-proBNPp (p=0.003). This was also the case for the CVis of NT-

proBNPu/creatinineu (p=0.002) and NT-proBNPu/h (p<0.001). The within-day CVi of NT-

proBNPu/h (pg/h) was higher than that of NT-proBNPu (pg/mL; p=0.031). For the day-to-day

and week-to-week protocols, the CVis of NT-proBNPu did not differ from the CVis of NT-

proBNPp. For the day-to-day protocol, the CVi of NT-proBNPu/h was higher compared to the

CVi of NT-proBNPu (p=0.044). The same statistical differences as for the CVis are applicable

for the comparisons between RCVs, because of the proportionality between these two parame-ters (Table 2).

88

Correlations Baseline NT-proBNPp (i.e. firstly collected blood sample of each patient) was inversely related

to body mass index (r= -0.548, p=0.005). Inversely related to LVEF were: baseline NT-

proBNPp (r= -0.599, p=0.009), baseline NT-proBNPu (i.e. the firstly voided urine) (r= -0.596,

p=0.009), baseline NT-proBNPu/creatinineu (r= -0.481, p=0.043), and baseline NT-

proBNPu/h (r= -0.538, p=0.021). Baseline NT-proBNPp and baseline NT-proBNPu were not

related to other patient characteristics, such as age, eGFR and NYHA class. Related to baseline

NT-proBNPp were: baseline NT-proBNPu (r=0.841, p=<0.001), baseline NT-

proBNPu/creatinineu (r=0.804, p=<0.001), and baseline NT-proBNPu/h (r=0.835,

p=<0.001).

When the data of the day-to-day and week-to-week protocols were pooled, NTproBNPp corre-

lated with NT-proBNPu in the corresponding enforced urine voidings (r=0.739, p<0.001) and

also with NT-proBNPu in the full 24-h urines (r=0.735, p<0.001). NT-proBNPu correlated also

with NT-proBNPu/creatinineu (r=0.932, p<0.001) and NT-proBNPu/h (r=0.950, p<0.001).

Figure 3 shows the relation between NT-proBNPp and NT-proBNPu in the corresponding en-

forced urine voidings collected in the day-to-day and week-to-week protocols. Initially, NT-

proBNPp did not correlate with NT-proBNPu. However, there was a clear relation from an NT-

proBNPp of about 500 pg/mL, obeying the equation NT-proBNPu = 0.0226 × (NT-

proBNPp)1.1563. The median NT-proBNPu in the lower NT-proBNPp region was 17 pg/mL.

Substituting this value into the equation gave an estimated intersection at an NT-proBNPp of

310 pg/mL.

Analysis of a subgroup with NT-proBNPp of 310 pg/mL We investigated whether the RCVs for NT-proBNPp were better for concentrations beyond 310 pg/mL, i.e. the concentration from which NT-proBNPu correlated with NT-proBNPp (Figure 3).

89

This appeared not to be the case. Except for a decrease in RCVs for week-to-week NT-

proBNPp (from 83 to 59%) and NT-proBNPp/creatinineu (from 113 to 107%), and day-to-day

NT-proBNPu/h (87 to 72%), all other RCVs increased while that of the day-to-day NT-proBNPp remained constant (Supplemental Table 1).

Changes of NT-proBNP in plasma and urine during the day The within-day courses, relative to baseline, for NT-proBNPp, NT-proBNPu and NT-

proBNPu/NT-proBNPp, all calculated from the concentrations in pg/mL, are shown in Figure

4A-4C. Concomitant p-values are in Supplemental Table 2. Data were expressed relative to

levels at 08:00 h (set at 100%). Within-day NT-proBNPp increased from 08:00 to 18:00 (Wil-coxon rank-sum p=0.006; Friedman p<0.001). No other changes were detected (Figure 4A).

No concomitant changes were observed for the within-day NT-proBNPu. The apparent decrease of NT-proBNPu from 08:00 to 18:00 was insignificant (Figure 4B). There were also no

changes in NT-proBNPu/creatinineu and NT-proBNPu/h from 08:00 to 18:00. NT-proBNPu/h

fluctuated from 12:00 to 16:00: there was a decrease from 12:00 to 14:00 and an increase

from 14:00 to 16:00 (Wilcoxon signed-rank test p=0.006 and p=0.007, respectively). There

was no change of the NT-proBNPu/ NT-proBNPp ratio from 8:00 to 18:00, but it decreased from 8:00 to 10:00 (Wilcoxon signed-rank test p=0.009; Figure 4C).

DISCUSSION

We investigated whether the within-day, day-to-day (5-days) and week-to-week (6-weeks;

collection at the same day of week) CVis of NT-proBNPu were more favorable compared to the

corresponding NT-proBNPp in patients with stable CHF. In contrast to expectation, we found

that the within-day CVi of NT-proBNPu was higher than that of NT-proBNPp. The day-to-day

and week-to-week CVis of NT-proBNPu did not differ from the CVis of NT-proBNPp. This con-

clusion also applies for NT-proBNPu levels adjusted for dilution (by creatinine) or expressed as

90

an excretion rate (per h). The CVis did not improve by restricting the analyses to NT-proBNPp

concentrations ≥310 pg/mL, i.e. the levels from which there was an almost linear relation be-

tween NT-proBNPu and NT-proBNPp.

In the present study the within-day, day-to-day and week-to-week CVis of NT-proBNPp

amounted to 9, 18 and 30%. These outcomes were similar to those of our previous study,

where Bruins et al. (17) reported 9, 20 and 35%, respectively. They are also in line with the

week-to-week CVis of 30% for natriuretic peptides reported by others (26;27). High NT-

proBNPp CVis are accompanied by proportionally higher RCVs that in the present study were 25

(within-day), 51 (day-to-day) and 83% (week-to-week). The latter RCV was comparable to the

week-to-week (8-weeks) RCV of 90% for NTproBNPp as reported by Wu et al. (26). The

nowadays well established high RCVs for NT-proBNP in both plasma and urine preclude the

use of these parameters for guided therapy of individuals, if longitudinal differences are to be

detected with 95% certainty (17).

It is as yet unclear why the CVis of (immunoreactive) NT-proBNPu are higher than those of

(immunoreactive) NT-proBNPp. NT-proBNPu reflects NT-proBNPp over an extended period and

thereby dampens fluctuation of NT-proBNPp, caused by its pulsatile secretion together with

BNP in a stoichiometric manner (12). Fluctuations of filtration fraction, eGFR, glomerular per-

meability of proteins and proximal tubule function are candidates. Permeability may not be limit-

ing for low molecular weight proteins that are usually filtered without difficulty. Larsson et al.

(28) observed no consistent changes of the eGFR during the day in healthy subjects. Other

determinants might relate to proteolytical cleavage in tubular microvillar membranes (14;15) and

receptor-mediated endocytosis by proximal tubule cells. In the latter, the megalin/cubilin recep-

tor complex may play a central role by its ability to bind structurally very different proteins (29),

but other mechanisms of protein internalization, like the caveolin-pathway and fluid-phase en-

docytosis have also been described (16). Megalin is involved in proximal tubular uptake of

91

glomerular-filtered albumin, other low-molecular weight proteins and their cargos (such as vita-

mins, minerals), including albumin, hemoglobin, transcobalamin-vitamin B12, and vitamin D-,

retinol- and folic acid-binding proteins, but also many peptides including parathormone, insulin,

epidermal and insulin-like growth factor and leptin (16;29;30). Following endocytosis, the cap-

tured proteins may become cycled to the plasma or tubular fluid by exocytosis, either as the

intact protein (transcytosis) or after (partial) degradation in lysosomes (16).

Our data suggest that (immunoreactive) NT-proBNPu is not related to (immunoreactive) NT-proBNPp (Figure 3) up to an NT-proBNPp threshold of about 310 pg/mL. From this threshold

filtration seems to be the dominant determinant, suggesting that at lower levels either filtration,

reabsorption, or both, might be limiting. Given the passiveness of filtration, NT-proBNPp above

310 pg/mL might be consistent with saturation of tubular reabsorption/metabolism, either by

co-filtered proteins like albumin, or the reach of maximum reabsorption/metabolism capacity of

NT-proBNP itself. Whatever the cause, we found that CVis for NTproBNPu at NTproBNPp levels

above 310 pg/mL were not more favorable than those applicable for the entire range. More

detailed studies on the renal handling of plasma proteins and notably NT-proBNPp are needed

to elucidate the factors causing the presently encountered high CVis of NT-proBNPu.

Consistent with our previous study (17), we found that NTproBNPp steadily increased during the day (Figure 4A). This seems to coincide with a decrease of NTproBNPu, notably between

08:00 and 10:00 (Figure 4B), although it did not reach significance. Increased volume load

(e.g. salt and fluids) related to breakfast and other meals, but also the prescribed medication

(Table 1) might contribute to the observed pattern, which is at least in part responsible for the

observed high within-day CVis.

The strength of this study was its prospective design with simultaneous collections of blood

samples and urines. This resulted in a reliable comparison of NT-proBNP levels in both plasma

92

and urine, collected in a single day (10-h), from day-to-day (5-days) and from week-to-week

(6-weeks), all taken during usual daily life activities. A limitation was some shortcomings in the

sampling within a single day. The number of urine portions differed per patient. Our solution was

to categorize the samples in periods of 2 h to provide the most reliable figure. The patients

noted the times of voiding themselves and these were verified by us during every visit to mini-

mize potential bias. Second, participants may not have collected the entire 24-h volumes, de-

spite our frequent visits. Correction for creatinine and expression as an excretion rate at least

partially overcame this problem. Finally, most participants used diuretics which may have caused higher water and salt excretion from 08:00 to 14:00 h (Figure 4A and 4B).

CONCLUSION

We conclude that measurement of urine NT-proBNP provides no advantages over plasma NT-

proBNP for therapy optimization in patients with chronic heart failure.

ACKNOWLEDGEMENTS

We thank all patients for their kind cooperation and Roche Diagnostics Netherlands b.v. (Mr.

Ben Aalderink) for partial financial support.

93

References

1 Leening MJ, Siregar S, Vaartjes I, Bots ML, Versteegh MI, van Geuns RJ, et al. Heart

disease in The Netherlands: a quantitative update. Neth Heart J 2014;22(1):3-10.

2 Heidenreich PA, Albert NM, Allen LA, Bluemke DA, Butler J, Fonarow GC, et al. Fore

casting the impact of heart failure in the United States: a policy statement from the

American Heart Association. Circ Heart Fail 2013;6(3):606-19.

3 Morrow DA, de Lemos JA. Benchmarks for the assessment of novel cardiovascular

biomarkers. Circulation 2007;115(8):949-52.

4 McMurray JJ, Adamopoulos S, Anker SD, Auricchio A, Bohm M, Dickstein K, et al.

ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure

2012: The Task Force for the Diagnosis and Treatment of Acute and Chronic Heart

Failure 2012 of the European Society of Cardiology. Developed in collaboration with the

Heart Failure Association (HFA) of the ESC. Eur Heart J 2012;33(14):1787-847.

5 Pfisterer M, Buser P, Rickli H, Gutmann M, Erne P, Rickenbacher P, et al. BNP-guided

vs symptom-guided heart failure therapy: the Trial of Intensified vs Standard Medical

Therapy in Elderly Patients With Congestive Heart Failure (TIME-CHF) randomized trial.

JAMA 2009;301(4):383-92.

6 De Vecchis R, Esposito C, Di BG, Ariano C, Giasi A, Cioppa C. B-type natriuretic pep-

tide-guided versus symptom-guided therapy in outpatients with chronic heart failure: a

systematic review with meta-analysis. J Cardiovasc Med (Hagerstown )

2014;15(2):122-34.

94

7 Savarese G, Trimarco B, Dellegrottaglie S, Prastaro M, Gambardella F, Rengo G, et al.

Natriuretic peptide-guided therapy in chronic heart failure: a meta-analysis of 2,686 pa

tients in 12 randomized trials. PLoS One 2013;8(3):e58287.

8 Fent G, Hall I, Watt V, O'Toole L, Al-Mohammad A. 43 Raised Natriuretic Peptides are

not Exclusively Indicative of Heart Failure: A Service Review of a Diagnostic Heart Fail-

ure Clinic. Heart 2014;100 Suppl 3:A23-A24.

9 Rodeheffer RJ. Measuring plasma B-type natriuretic peptide in heart failure: good to go

in 2004? J Am Coll Cardiol 2004;44(4):740-9.

10 Levin ER, Gardner DG, Samson WK. Natriuretic peptides. N Engl J Med

1998;339(5):321-8.

11 Balion CM, Santaguida P, McKelvie R, Hill SA, McQueen MJ, Worster A, et al.

Physiological, pathological, pharmacological, biochemical and hematological factors af-

fecting BNP and NT- proBNP. Clin Biochem 2008 Mar;41(4-5):231-9.

12 Bentzen H, Pedersen RS, Pedersen HB, Nyvad O, Pedersen EB. Abnormal rhythmic

oscillations of atrial natriuretic peptide and brain natriuretic peptide in heart failure. Clin

Sci (Lond) 2003;104(3):303-12.

13 Palmer SC, Endre ZH, Richards AM, Yandle TG. Characterization of NT-proBNP in hu

man urine. Clin Chem 2009;55(6):1126-34.

14 Kenny AJ, Maroux S. Topology of microvillar membrance hydrolases of kidney and in-

testine. Physiol Rev 1982;62(1):91-128.

95

15 Maack T, Johnson V, Kau ST, Figueiredo J, Sigulem D. Renal filtration, transport, and

metabolism of low-molecular-weight proteins: a review. Kidney Int 1979;16(3):251-70.

16 Dickson LE, Wagner MC, Sandoval RM, Molitoris BA. The proximal tubule and albu-

minuria: really! J Am Soc Nephrol 2014;25(3):443-53.

17 Bruins S, Fokkema MR, Romer JW, Dejongste MJ, van der Dijs FP, van den Ouwe-

land JM, et al. High intraindividual variation of B-type natriuretic peptide (BNP) and

amino-terminal proBNP in patients with stable chronic heart failure. Clin Chem

2004;50(11):2052-8.

18 Ng LL, Geeranavar S, Jennings SC, Loke I, O'Brien RJ. Diagnosis of heart failure us

ing urinary natriuretic peptides. Clin Sci (Lond) 2004;106(2):129-33.

19 Cortes R, Portoles M, Salvador A, Bertomeu V, Garcia de BF, Martinez-Dolz L, et al.

Diagnostic and prognostic value of urine NT-proBNP levels in heart failure patients. Eur

J Heart Fail 2006;8(6):621-7.

20 Michielsen EC, Bakker JA, Kimmenade RR, Pinto YM, Dieijen-Visser MP. The diag

nostic value of serum and urinary NT-proBNP for heart failure. Ann Clin Biochem

2008;45(Pt 4):389-94.

21 Solomon SD, Anavekar N, Skali H, McMurray JJ, Swedberg K, Yusuf S, et al. Influ-

ence of ejection fraction on cardiovascular outcomes in a broad spectrum of heart fail-

ure patients. Circulation 2005;112(24):3738-44.

96

22 Barnes SC, Collinson PO, Galasko G, Lahiri A, Senior R. Evaluation of N-terminal

pro-B type natriuretic peptide analysis on the Elecsys 1010 and 2010 analysers. Ann

Clin Biochem 2004;41(Pt 6):459-63.

23 Stevens LA, Coresh J, Feldman HI, Greene T, Lash JP, Nelson RG, et al. Evaluation

of the modification of diet in renal disease study equation in a large diverse population.

J Am Soc Nephrol 2007;18(10):2749-57.

24 Steele BW, Wang E, Palomaki GE, Klee GG, Elin RJ, Soldin SJ, et al. An evaluation

of analytic goals for assays of drugs: a College of American Pathologists Therapeutic

Drug Monitoring Sur vey Study. Arch Pathol Lab Med 2001;125(6):729-35.

25 Petersen PH, Sandberg S, Fraser CG, Goldschmidt H. Influence of index of individual-

ity on false positives in repeated sampling from healthy individuals 162. Clin Chem Lab

Med 2001;39(2):160-5.

26 Wu AH, Smith A, Wieczorek S, Mather JF, Duncan B, White CM, et al. Biological varia-

tion for N-terminal pro- and B-type natriuretic peptides and implications for therapeutic

monitoring of patients with congestive heart failure. Am J Cardiol 2003;92(5):628-31.

27 Melzi dG, Tagnochetti T, Nauti A, Klersy C, Papalia A, Vadacca G, et al. Biological

variation of N-terminal pro-brain natriuretic peptide in healthy individuals. Clin Chem

2003;49(9):1554-5.

28 Larsson A, Akerstedt T, Hansson LO, Axelsson J. Circadian variability of cystatin C,

creatinine, and glomerular filtration rate (GFR) in healthy men during normal sleep and

after an acute shift of sleep. Chronobiol Int 2008;25(6):1047-61.

97

29 Shankhajit De 1 SK2aAS2. The Endocytic Receptor Megalin and its Associated Pro-

teins in Proximal Tubule Epithelial Cells. 4[membranes 4030333], 333-355. 2014.

ww.mdpi.com/journal/membranesw, Membranes. 11-7-2014.

30 Christensen EI, Birn H, Storm T, Weyer K, Nielsen R. Endocytic receptors in the renal

proximal tubule. Physiology (Bethesda ) 2012;27(4):223-36.

98

Week-to-week protocol

week 1 week2 week 3 week 4 week 5 week 6

Day-to-day protocol

day 1 day 2 day 3 day 4 day 5

08:00

10:00

12:00

14:00

16:00

18:00 Within-day protocol

Figure 1. Protocol for blood and urine sampling: an example.

The total study duration was six weeks. We employed three different protocols for establish-

ment of within-day, day-to-day and week-to-week variation, respectively. The samples of the

week-to-week protocol were collected on the same day of the week.

99

Figure 2. Overview of collected and missing blood samples and urine portions of within-day, day-to-day and week-to-week sampling protocols. Spont., spontaneously; p.p., per person.

20 patients 6 times 18 patients 5 days 22 patients 6 weeks

maximum: 120 samples maximum: 90 samples maximum: 132 samples

116 plasma samples 4 missing 89 plasma samples 1 missing 124 plasma samples 8 missing

187 spont.voided urine portions 89 forced urine portions 1 missing1 missing 123 forced urine portions 9 missing

mean 9 (6-14) urine portions p.p. 84 full 24-h urine portions 6 missing 123 full 24-h urine portions 9 missing

patient recruitment from

2 cardiologists' practices

25 participants

Within-day Day-to-day Week-to-week

100

Figure 3. Relation between NT-proBNP in plasma and NT-proBNP in the enforced

urine voidings collected in the day-to-day and week-to-week protocols.

Data derive from 22 patients and in total 182 comparisons of NT-proBNP in plasma and urine.

Samples were from the day-to-day and week-to-week protocols. For urine we used the en-

forced voidings. NT-proBNPp was unrelated to NT-proBNPp up to a concentration of 310 pg/mL

(corresponding with NT-proBNPp of 17 pg/mL). The relation from NT-proBNPp of 310 pg/mL

was: NT-proBNPp= 0.0226*(NT-proBNPp)1.1563.

101

102

Figure 4. Within-day courses for NT-proBNP in plasma, urine and the NT-proBNP

urine/plasma ratio. Data are medians (interquartile range, IQR) relative to 08:00. Numbers

underneath clock time refer to numbers of patient samples. #significance of 8:00 – 18:00 time-

period; *significance of two hours time-period; NT-proBNP, amino terminal pro-brain natriuretic

peptide. All p-values are reported in Table 3.

103

Table 1. Baseline characteristics of the study group.

Participants, numbers 25

Age, mean±SD, years 61±10

Gender (male/female), numbers 15/10

Body mass index mean±SD, kg/m2 28±6

eGFR, mean±SD, mL/min/1.73 m2 81±31

left ventricular ejection fraction, mean±SD, % 36±15

New York Heart Association classes, number (%)

I 2 (8)

II 14 (56)

III 8 (32)

Medication, number (%)

Beta-blocking agent 19 (76)

Angiotensin-converting enzyme inhibitor 10 (40)

Diuretics 22 (100)

Digoxin 5 (20)

NT-proBNPp, median (range)

NT-proBNPp, pg/mL 205 (12-3,404)

NT-proBNPub, median (range)

NT-proBNPu, pg/mL 19 (0-173)

NT-proBNPu/NT-proBNPp 2 (0-49)

NT-proBNPu/h, pg/h 1 (0-18)

NT-proBNPu/NT-proBNPp, median (range) 0.06 (0.00-1.86)

eGFR, estimated glomerular filtration rate; NT-proBNP, amino terminal pro-brain natriuretic peptide; p, plasma; u, urine; a Retrieved from medical files. Seven out of 25 LVEF data were missing. One out of 25 NYHA classifications was missing. b full 24-h urine portions.

104

Table 2. Analytical variations (CVa), intra-individual variations (CVi), total variations (CVt)

and reference change values (RCV), for plasma and urine NT-proBNP.

Within-day Day-to-day Week-to-week

NT-proBNPp, pg/mL

CVa, % 3 3 3

CVi, % 9 (0-21) c 18 (5-90) 30 (6-215)

CVt, % 9 (3-22) 19 (6-90) 30 (7-215)

RCV, % 25 51 83

NT-proBNPu, pg/mL

CVa, % 5 (0-15) 7 (0-14) 4 (0-87) e

CVi, % 34 (0-71) c 21 (0-60) d 28 (0-180) e

CVt, % 35 (0-71) 23 (0-60) 29 (0-200) e

RCV, % 96 62 80

NT-proBNPu/creatinineu,

pg/nmol

CVa, % 6 (2-15) 8 (2-15) 4 (2-87) e

Cvi, % 29 (0-93) c 27 (0-112) 40 (0-180) e

CVt, % 29 (0-93) 27 (0-113) 40 (0-200) e

RCV, % 81 77 113

NT-proBNPu/hour, pg/h

CVa, % 5 (0-15) 7 (0-14) 4 (0-87) e

Cvi, % 47 (0-174) c 31 (0-67) d 33 (0-180) e

CVt, % 48 (0-174) 32 (0-67) 34 (0-200) e

RCV, % 132 87 92

Data are medians (ranges). For within-day urine amino-terminal pro-brain natriuretic peptide

(NT-proBNPu) we used spontaneously voided urine portions; for day-to-day and week-to-week

105

NT-proBNPu we used the full 24-h urine portions. c Within-day CVis of NT-proBNPu, NT-

proBNPu /creatinineu, and NT-proBNPu/h were higher than the CVi of NT-proBNPp (p=0.003

for pg/mL; p=0.002 for pg/nmol and p<0.001 for pg/h). Within-day CVi of NT-proBNPu/h (pg/h) was higher than that of NT-proBNPu (pg/mL; p=0.031). d Day-to-day CVi of NT-

proBNPu/h (pg/h) was higher than that of NT-proBNPu (pg/mL; p=0.044). e Analytical variation (CVa) increased with decreasing NT-proBNP outcome; the upper-limit of

the presented CVa was determined by one patient.

106

Table 3. Within-day time-period differences for NT-proBNP in plasma, urine and the

urine/plasma ratio.

p-values

time-period

W

08-18

F

08-18

WSR

08-10

10-12

12-14

14-16

16-18

NT-proBNPp 0.006f 0.000f 0.015 0.019 0.327 0.586 0.557

NT-proBNPu 0.064 0.261 0.100 0.650 1.000 0.701 0.790

NT-proBNPu/creatu 0.638 0.286 0.036 0.861 0.347 0.028 0.657

NT-proBNPu/h 0.695 0.054 0.088 0.152 0.006g 0.007g 0.286

NT-proBNPu/NT-

proBNPp

0.091 0.215 0.009f 0.152 0.790 0.480 0.445

W, Wilcoxon test; F, Friedman test; WSR, Wilcoxon Signed Ranks test using p<0.01 for signifi-

cance; NT-proBNP, amino-terminal pro-B-type natriuretic peptide; f significant at p<0.05; g significant at p<0.01.

107

Supplemental Table 1.

Analytical variations (CVa), intra-individual variations (CVi), total variations (CVt) and refer-

ence change values (RCV), for plasma and urine NT-proBNP in the subgroup with NT-

proBNPp>310 pg/mL.

Within-day Day-to-day Week-to-week NT-proBNPp, pg/mL

CVa, % 3 3 3 CVi, % 12 (7-21) h 18 (9-90) 21 (6-66) i

CVt, % 12 (7-22) 19 (10-90) 21 (7-66) RCV, % 34 51 59

NT-proBNPu, pg/mL CVa, % 2.55 (2.55-8) 2.55 (2.55-9) 2.55 (1.76-8) CVi, % 48 (36-71) h 36 (1-60) 36 (19-101) i

CVt, % 49 (39-71) 36 (4-60) 36 (19-101) RCV, % 135 100 100

NT-proBNPu/creatinineu, pg/nmol

CVa, % 3 (3-9) 3 (3-9) 3 (3-8) Cvi, % 32 (10-93) h 27 (4-86) 38 (22-134) i

CVt, % 32 (11-93) 27 (5-86) 38 (22-140) RCV, % 89 75 107

NT-proBNPu/hour, pg/h CVa, % 2.55 (2.55-8) 2.55 (2.55-9) 2.55 (1.76-8) Cvi, % 56 (38-107) h 26 (14-66) 55 (26-94) i

CVt, % 56 (39-107) 28 (14-66) 55 (26-94) RCV, % 157 72 154

Data are median (range). For within-day urine amino-terminal pro-brain natriuretic peptide (NT-

proBNPu), spontaneously voided urine portions were used; for day-to-day and week-to-week NT-proBNPu, full 24-h urine portions were used. h The within-day CVis of NT-proBNPu were

108

higher than the CVi of NT-proBNPp (p=0.018 for pg/mL; p=0.028 for pg/nmol and p=0.018

for pg/h). i The week-to-week CVis of NT-proBNPu were higher than the CVi of NT-proBNPp (p=0.013 for

pg/mL; p=0.020 for pg/nmol and p=0.028 for pg/h). Analytical variation, CVa; intra-individual

variation, CVi; total variation, CVt; p, plasma; u, urine; reference change value, RCV.

109

110

111

Chapter 3

BNP AND NT-PROBNP, PREDICTORS OF 1-YEAR MORTALITY IN

NURSING HOME RESIDENTS. Barents M, Hillege HH, van der Horst IC, de Boer RA, Koster J, Muskiet FA, de Jongste MJ.

J Am Med Dir Assoc. 2008;9(8):580-5. Epub 2008 7.

112

ABSTRACT

OBJECTIVES. To investigate on one-year mortality prediction of B type natriuretic peptide

(BNP) and N terminal-proBNP (NT-proBNP) in institutionalized elderly with multiple-morbidity. METHODS. A prospective, cross-sectional study in one nursing home was designed. Partici-

pants: 93 residents (mean age 81 ± 3 years, 66% female). Residents with serious cognitive

impairments, aphasia or metastatic cancer were excluded.

Measurements: clinical assessment, immobilization, medical history, electrocardiogram (ECG),

echocardiogram, blood samples. One general geriatrician assessed non-cardiovascular dis-

eases, a cardiologist panel established the diagnosis of chronic heart failure (CHF). Subjects

were tracked for 1 year as far as status of death. RESULTS. 18/93 enrolled individuals died. BNP was significant higher in non-survivors com-

pared to survivors (138 (49-753) versus 87 (27-162) P 0.029), NT-proBNP was higher but

did not reach significance 1382 (193-5683) versus 335 (175-900) pg/mL (interquartile range

(IQR), P 0.059). The adjusted value on one-year mortality, of six predefined chronic diseases,

immobilization, age, sex, NT-proBNP and BNP was estimated by means of Cox proportional

hazard regression analyses. Finally, both for NT-proBP and BNP, a mutually adjusted multivari-

ate Cox proportional hazard analysis with the covariates presented that BNP and NT-proBNP

predicted one-year mortality significantly (HR 1.16 and P 0.003, HR 1.02 and P 0.001, respec-

tively).The mortality risk increased at rising BNP and NT-proBNP levels. CONCLUSION. BNP and NT-proBNP are predictors of one-year mortality independently of age,

gender and morbidity.

The mortality risk increases at elevating natriuretic peptide concentrations. We postulate that

plasma levels of BNP and NT-proBNP are also of use to predict prognosis in institutionalized

elderly with multiple morbidity.

113

INTRODUCTION

Elderly with multiple morbidity admitted to a nursing home are faced with questions concerning

their health, living conditions and social life. Information about the prognosis of their diseases is

essential for them to make decisions on these questions and therefore of importance for the

quality of their life. Aside existing chronic diseases and their progression, also body mass index

(BMI), waist-hip ratio, biomarkers as albumin and natriuretic peptides predict 1-year mortality

(1-11). Natriuretic peptides (BNP and NT-proBNP) have become available as tools for the di-

agnosis of heart failure and for the prognosis (4;7-9). Their prognostic role was evaluated in-

dependently of the presence of chronic heart failure and found to be of importance. However,

most of these data are obtained from specialized clinics treating relatively younger patients with

heart failure without comorbidity. In addition, BNP levels also increase in noncardiac conditions

such as old age, being female, pulmonary diseases (pulmonary hypertension or embolism,

chronic obstructive pulmonary disease) and renal dysfunction (12-15). Although nearly all nurs-

ing home residents suffer from multiple morbidity, data on the value of use of natriuretic pep-

tides in them are limited. We therefore investigated whether BNP and NT-proBNP plasma levels

remain of prognostic value in elderly patients with multiple morbidity.

METHODS

During the course of this study there were 140 persons in the departments for residents with

diseases of somatic origin in NH “het Zonnehuis”. Their impairments are based on cardiovascu-

lar, pulmonary, endocrine, neurodegenerative, skeletal muscle, renal dysfunction (16) and other

disorders. Most of them receive long-term care. Some residents with skeletal muscle or cere-

brovascular disorders were reactivated, one-third of whom could be discharged (short stay) but

who should remain care dependent in primary care. Both groups were invited to take part if they

were 65 years of over. The long-term and short-stay residents were included if they understood

114

the impact of the study on themselves (competent) and if they agreed to participate by written

informed consent.

Persons with aphasia or a cognitive impairment measured by the mini-mental state estimation

(MMSE) test ≤20 were excluded. The MMSE test contains 30 questions and a MMSE ≤20 is

suspect for cognitive impairments (17). Persons with metastatic cancer who stayed in a de-

partment specialized in palliative terminal care were excluded. Persons were also excluded if

the echocardiographic frames could not be sufficiently visualized or if they refused to have their

blood sample taken.

The study was approved of by the Medical Ethical Committee in Groningen, University Medical

Center Groningen, The Netherlands (METc number 2004.107).

Diagnostics In this cross-sectional study, all data were collected anonymously and within one week (ques-

tionnaire, neurohormone sampling, ECG, echocardiography). One physician collected data of

the patients’ medical history and (chronic) diseases of the cardiovascular, pulmonary, endo-

crine, neurodegenerative and skeletal muscle system, of their symptoms and medication. He

also examined blood pressure, height and weight and performed full physical examination (18).

Aside number and type of chronic disorder, we registered immobilization defined as being

wheelchair dependent or bedridden. We regarded immobilization as a consequence of the

chronic diseases on the individual and therefore as a parameter of progression of the present

chronic diseases. The diagnosis of chronic heart failure (CHF) was made by two experienced

cardiologists. A third colleague decided in cases of disagreement. Of note, the cardiologists

were unaware of resident’s BNP and NT-proBNP levels. A 12-lead electrocardiogram (ECG)

was made with the electrocardiograph Cardioline delta three plus (Cardioline, Milan, Italy,

www.cardioline.it), with the patient in a horizontal position.

The left-ventricular ejection fraction (LVEF) was assessed semi-quantitatively by the two-

dimensional visual estimate method (19). A LVEF ≤ 45% was considered to be a left-

115

ventricular systolic dysfunction (LVSD) or diminished LVEF(dLVEF). The hand held cardiograph

‘Opti Go’ (Philips, Eindhoven, The Netherlands, www.philips.com) was used. One blood sample

per resident (12 mL) was taken if one was fasting and at rest. Assessments of creatinine, hae-

moglobin and mean corpuscular volume (MCV) were performed in the NH laboratory. Since the

neurohormone levels may vary about 100% we choose not to repeat the neurohormone sam-

pling. For determination of levels of (NT-pro)BNP, a 5 µl aprotinine solution was adjusted to the

250 µl plasma samples in EDTA and to 250 µl serum samples. At the UMCG Clinical Chemi-

cal Laboratory (CCL) both were frozen at -20˚C and stored in batches for a maximum of 10

months. The assays were run in one go for both NT-proBNP and BNP. Both NT-proBNP and

BNP were measured by immunoassays (Elecsys®1010/2010/modular analytics 2004 Roche

diagnostics Indianapolis IN US, and AXSYM system® BNP 2003 Axis–shield diagnostics LTD

ABBOT Wiesbaden Germany). NT-proBNP and BNP had coefficients of variation of 3.3% and

7.8%(ranges of 5-35000 pg/mL and 0-3465 pg/mL respectively).

Renal function has been defined as estimated glomerular filtration rate (eGFR) measured by

the Cockcroft -Gault formula in mL/min.

Anemia was defined as haemoglobin (Hb) ≤ 7.5 mmol/l ≈12 g/dL (7). The status of death or

alive of each resident was recorded after one year following initial data collection.

Statistical analysis To address as to BNP and NT-proBNP predict one-year mortality in a cohort of old age with

multiple chronic diseases we used different statistic tests. Differences in basic characteristics of

survivors and non-survivors were analysed using Students ’t test, X 2 test or the Fisher exact

test for categorical data and the Mann-Whitney test for non-parametric continuous data, as ap-

propriate. The association between underlying diseases or immobilization and NT-proBNP and

BNP levels divided into tertiles, was tested with X2 statististics.

The adjusted value of the six predefined chronic diseases, immobilization, NT-proBNP and

BNP, unadjusted and adjusted for age and sex were estimated by means of Cox proportional

116

hazard regression analyses. The six chronic diseases were CHF all cases (or CHF with dLVEF

or CHF pLVEF), COPD, diabetes, renal dysfunction, neurodegenerative diseases, skeletal-muscle system diseases (Table 3).

Finally, both for NT-proBP and BNP, a mutually adjusted multivariate Cox proportional hazard

analysis was made with the covariates in the equation. The quantitative relationship between NT-proBNP and BNP levels and one-year mortality was expressed in Figure 1 and 2, respec-

tively.

Statistical analyses were performed using SPSS 12.0.1 software (SPSS Inc., 233 S. Wacker,

11th Floor, Chicago IL US). All statistical comparisons were two-tailed, and a p value <0.05 was

considered to be statistically significant.

RESULTS

A total of 93 (62%) individuals out of 140 residents were included. Individuals were excluded

because of cognitive impairment [30 (21%)], missing ECG data [2 (1.4%)], missing laboratory

values [1 (0.7%)], and unwillingness to participate [14 (10%)].

The mean age of the studied population was 81 ± 3 years and approximately 66% of the resi-dents were females (Table 1).

After a follow-up period of one year, 18/93 individuals died. Patients who survived did not differ

in age from non-survivors (80 ± 7 years and 81 ± 9 versus, p 0.796). In comparison with sur-

vivors, non-survivors had more often CHF (all cases), diabetes and higher BNP levels.

Out of the 10 non-survivors with CHF, two subjects had symptoms and eight had no symptoms

of heart failure, 4/10 had pLVEF and 6/10 dLVEF. BNP and NT-proBNP were elevated in

CHF compared to no-CHF, but only BNP differed significantly (median BNP in CHF 194, IQR

92-460 versus BNP in no-CHF 87, IQR 28-187 pg/mL (p<0.001) and NT-proBNP in CHF

1871, IQR 539-4262 versus NT-proBNP in no-CHF 324, IQR 163-1146 pg/mL (p< 0.001)).

117

Renal dysfunction (p 0.046 and p 0.004), CHF all cases (p 0.014 and p 0.004) and CHF

with dLVEF (p 0.018 and p 0.001) were more observed in the highest BNP and NT-proBNP tertiles (p-values respectively) (Table 2).

CHF all cases (P=0.014 and P=0.004) and CHF with dLVEF (P=0.018 and P=0.001) were

observed more in the highest BNP and NT-proBNP tertiles (P values respectively). BNP and

NT-proBNP predicted 1-year mortality significantly after adjustment for age, sex, the 6 chronic

diseases, and immobilization (with CHF all cases, the analysis had a hazard ratio [HR] of 1.67

and P=0.000, HR 0.60 and P=0.000; with CHF with pLVEF HR 1.76 and P=0.000, HR 2.06

and P=0.000; with CHF with dLVEF HR 1.03 and P=0.000, HR 0.93 and 4.86, respectively) (Table 3). The mortality risk was increasing at rising NT-proBNP and BNP levels. An increase

of 10 pg/mL BNP was related to a 2.2% rise of the HR of the mortality risk. Moreover, an in-

crease of 10 pg/mL NT-proBNP was associated with a 0.27% rise of the HR of mortality risk (Figures 1 and 2, Table 3).

DISCUSSION

The major finding of this study is that in NH-residents BNP and NT-proBNP are independent

predictors of mortality, after adjustment for age, sex, chronic diseases and immobilization. One

of the main characteristics of NH-residents is the presence of multiple morbid conditions. Resi-

dents have an average of four chronic diseases each. Moreover, CHF all cases and diabetes are more frequently represented in those who died, compared to those who did survive (Table

1). We had to consider the skewed distribution of BNP and NT-proBNP caused by extended

morbidity in a relatively small population. Therefore, medians instead of means of neurohor-

mones were used and an adjustment for all confounders of neurohormone levels was made as is presented in Table 2 and 3. After adjustment for all confounders, the neurohormones pre-

dicted one-year mortality significantly in NH-residents. Another aspect of morbidity, aside of

number and type is the progression of each chronic disease. In this study however we did not

register the progression of chronic diseases. To address this shortcoming we determined immo-

118

bilization as parameter of consequences of morbidity (22-24). Aside mortality risks as age and

sex, we had to count with chronic diseases and their consequences influencing individuals. So,

after adjustment for age, sex and morbidity, BNP and NT-proBNP still remained predictors of one-year mortality in nursing home residents (Table 3). Moreover, the quantitative relation be-

tween natriuretic peptides and mortality revealed high levels of natriuretic peptides required for prognosis compared to their use for diagnosis (Figure 1 and 2 ). For instance, an increase of

BNP or NT-proBNP with 10 pg/mL is related to an increase of mortality risk of 2,2% or 0.27%,

respectively. To date, there is a scarcity of reports on the value of BNP and NT-proBNP in eld-

erly in nursing homes. Comparisons with other populations should therefore only be made,

taken these flaws into account.

In NH residents BNP and NT-proBNP are of prognostic value at higher levels, compared to a

non-elderly population (25;26). In a study of ambulatory CHF patients (mean age 76 ±11

years) with preserved LVEF and with readmission or death for cardiac reasons as endpoints,

Valle et al. found predictive values of BNP at 200-499 and >500 pg/mL (HRs of 2.2 and

5.8) after six months follow-up (27). Though there were differences in design and population

between Valle’s and the present study, both studies suggest that BNP levels for prognostic use

were found to be two- to fourfold higher than advised for diagnostic purpose. Bibbins et al. found NT-proBNP to be a marker of long-term mortality, independently of other prognostic

markers. In subjects with stable coronary heart disease with a mean age of 67 to 72 ± 9 year,

they also found increasing mortality risks at incremental NT-proBNP levels (28). McKie et al. studied a younger community-based cohort without heart and renal failure (mean age of 62

years), and found BNP and NT-proBNP to be biomarkers for mortality at much lower levels

(biosite assay of BNP was 63 vs. 22 and NT-proBNP 206 vs. 63 pg/mL median levels, non-

survivors vs. survivors) (29). Regarding a different range, the relationship between increasing

mortality risk and elevating natriuretic peptide levels was comparable to our results (30).

Altogether, we have demonstrate in this study that BNP and NT-proBNP are of prognostic value

also in care dependent elderly with multiple morbidity. However, the cut-off points for prognostic

119

purpose of neurohormones seem higher and should be identified in future studies keeping track

on comorbidity. A limitation of the study is the small population of 93 nursing home residents, in

one centre. On the other hand, the population studied is representative of other nursing home

populations in terms of distribution of age, gender, diabetes, CHF, the use of ACE-inhibitor

therapy and number of chronic diseases, although not representative in terms of the presence of

hypertension and renal dysfunction (31-33).

CONCLUSION

In a cohort of NH-residents, BNP and NT-proBNP levels are independent predictors of one-year

mortality, after adjustment for age, sex, chronic morbidity and immobilization. In the respect of

the prognosis, mortality risk increases with elevated natriuretic peptide levels. We postulate that

plasma levels of BNP and NT-proBNP are also of use to predict prognosis in institutionalized

elderly with multiple co-morbidities.

ACKNOWLEDGEMENTS

All authors had no conflict of interest. Fund: This study was subsidised by the “Vereniging het

Zonnehuis”, Soesterberg, The Netherlands. Financial disclosure: All contributors have no con-

flict of interest to disclose.

120

Table 1. Baseline characteristics of NH residents, survivors compared to non-survivors.

Survivors Non-survivors P value

Number (% of 93) 75 (81) 18 (19)

Age (year) mn±SD 80 ± 7 81 ± 9 0.796

Female 49 (53) 12 (13) 0.740

BP diastolic mmHg mn±SD 80 ± 12 74 ± 11 0.050

BP systolic mmHg mn±SD 142 ± 26 133 ± 23 0.155

Hemoglobin mmol/l mn±SD 7.7 ± 1 7.4 ± 1 0.352

Creatinine umol/l mn±SD 76 ± 27 102 ± 78 0.175

Body mass index kg/m2 mn±SD 26 ± 6 23 ± 3 0.780

LVEF % mn±SD 52 ± 8 49 ± 10 0.254

CHF all cases 13 (14) 10 (10) 0.007

CHF reduced ejection fraction 10 (11) 6 (6) 0.053

CHF preserved ejection fraction 3 (3) 4 (4) 0.190

chronic obstructive pulmonary d. 17 (18) 7 (7) 0.161

Diabetes mellitus type 2 16 (17) 9 (9) 0.038

renal dysfunction 11 (12) 7 (7) 0.066

neurological disorders 39 (41) 7 (7) 0.231

skeletal-muscle disorders 35 (37) 4 (4) 0.060

immobilization 26 (27) 9 (9) 0.232

BNP pg/mL md (IQR) 87 (27 -162) 138 (49 – 753) 0.029

NT-proBNP pg/mL md (IQR) 335 (175-900) 1382 (193-5683) 0.059

Data are numbers (% of the 93 subjects) unless otherwise indicated. BP, blood pressure; im-

mobilization, subjects in wheelchair or bed; md IQR, median with interquartile range; NT-

proBNP, N-terminal-pro B-type natriuretic peptide; BNP, B-type natriuretic peptide.

121

Table 2. BNP and NT-proBNP concentrations divided in tertiles and linear related to chronic

diseases and immobilization.

Tertiles

total

% of 93

BNP

I

%

II

%

III

%

P

NT-

pro

I

%

BNP

II

%

III

%

P

COPD 25 8 13 4 0.276 11 8 6 0.507

diabetes 26 12 6 8 0.433 13 5 8 0.254

RD 19 2 8 9 0.046 2 5 12 0.004

MD 41 17 15 9 0.089 16 13 12 0.443

ND

CHF all

CHFpEF

CHFrEF

Immobilization

48

24

6

18

36

15 4

1

3

11

18 6

2

4

15

15 14

3

11

10

0.904 0.014

0.559

0.018

0.870

18 3

2

1

10

11 6

1

5

12

19 15

3

12

14

0.800 0.004

0.586

0.001

0.434

BNP, B-type natriuretic peptide; NT-proBNP, N-terminal-proBNP; COPD, chronic obstructive

pulmonary disease; RD, renal dysfunction; MSD, musculoskeletal disorder; ND, neurological

disorders; CHF, chronic heart failure; CHFpEF, CHF with preserved ejection fraction; HFrEF,

CHF with reduced.

122

Table 3. Association of BNP and NT-proBNP with 1-year mortality after adjustment for mor-tality risk factors. Neuro- Hormone, NH

HR

BNP 95% CI

P

HR

NT-proBNP 95% CI

P

NH, unadjusted 1.22 1.13-1.32 0.000 1.03 1.02-1.04 0.000 NH, age and sex adj. 1.23 1.13-1.33 0.000 1.03 1.02-1.04 0.000 NH, age and sex adj.+ each of the following added 1 at a time:

CHF all* 1.18 1.08-1.29 0.000 1.02 1.01-1.04 0.000 CHFpEF* 1.20 1.10-1.30 0.000 1.03 1.01-1.04 0.000 CHFrEF* 1.19 1.09-1.30 0.003 1.03 1.01-1.04 0.001 diabetes mellitus type2 1.20 1.09-1.31 0.000 1.03 1.01-1.04 0.000 COPD 1.18 1.08-1.29 0.000 1.02 1.01-1.04 0.000 Renal disorders 1.18 1.08-1.29 0.000 1.02 1.01-1.07 0.000 muscular-skeletal d. 1.17 1.07-1.30 0.001 1.02 1.01-1.06 0.000 neurological disorders 1.15 1.04-1.27 0.004 1.02 1.01-1.06 0.001 Immobilization 1.16 1.05-1.28 0.003 1.02 1.01-1.07 0.001 All diseases 1.16 1.05-1.28 0.003 1.02 1.01-1.07 0.001

CHF, chronic heart failure; * only 1 of them is in the model; CHFpEF, CHF with preserved

ejection fraction; CHFrEF, CHF with reduced EF; COPD, chronic obstructive pulmonary disease

; d. disorders; HR, hazard ratios; CI, confidence interval; BNP, B-type natriuretic peptide; NT-

proBNP, amino-terminal pro natriuretic peptide.

123

BNP pg/ml

0 100 200 300 400 500 600 700 800

Fitte

d Ha

rzar

d Ra

tio

0.0

1.0

2.0

3.0

4.0

5.0

6.0

Figure 1. Hazard ratio’s (HR) of BNP and 1-year mortality risk. HR of B-type natriuretic pep-tide, BNP (X) and mortality risk (Y) with confidence intervals (Cl 3 and Cl 4) presenting the quantitative relation of mortality at increasing BNP. The BNP range is limited to fit in this graph. HR 1 is at BNP level 87 pg/mL.

NT-proBNP pg/ml

0 1000 2000 3000 4000 5000

Fitte

d Ha

rzar

d Ra

tio

0.0

1.0

2.0

3.0

4.0

5.0

6.0

Figure 2. Hazard ratio’s (HR) of N-terminal-proB-type natriuretic peptides, NT-proBNP and

1-year mortality risk. HR according to Nt-proBNP range from 0 to 5000 pg/mL are depicted.

Maximum Nt-proBNP level is fitted to this graph. HR 1 is at Nt-proBNP 450 pg/mL.

124

References

1 Price GM, Uauy R, Breeze E, Bulpitt CJ, Fletcher AE. Weight, shape, and mortality risk

in older persons: elevated waist-hip ratio, not high body mass index, is associated with

a greater risk of death. Am J Clin Nutr 2006;84(2):449-60.

2 Sullivan DH, Roberson PK, Bopp MM. Hypoalbuminemia 3 months after hospital

discharge: significance for long-term survival. J Am Geriatr Soc 2005;53(7):1222-6.

3 van Dijk PT, Mehr DR, Ooms ME, Madsen R, Petroski G, Frijters DH, et al.

Comorbidity and 1 year mortality risks in nursing home residents. J Am Geriatr Soc

2005;53(4):660-5.

4 Wallen T, Landahl S, Hedner T, Nakao K, Saito Y. Brain natriuretic peptide predicts

mortality in the elderly. Heart 1997;77(3):264-7.

5 Yan LL, Daviglus ML, Liu K, Stamler J, Wang R, Pirzada A, et al. Midlife body mass

index and hospitalization and mortality in older age. JAMA 2006;295(2):190-8.

6 Taylor JA, Christenson RH, Rao K, Jorge M, Gottlieb SS. B-type natriuretic peptide and

N-terminal pro B-type natriuretic peptide are depressed in obesity despite higher left

ventricular end diastolic pressures. Am Heart J 2006;152(6):1071-6.

7 Richards M, Troughton R, Lainchbury J, Doughty R, Wright S. Health gains by using

natriuretic peptides in diagnosis, prognosis and treatment. Scand J Clin Lab Invest

Suppl 2005;240:129-37.

8 Valle R, Aspromonte N, Feola M, Milli M, Canali C, Giovinazzo P, et al. B-type

natriuretic peptide can predict the medium-term risk in patients with acute heart failure

and preserved systolic function. J Card Fail 2005;11(7):498-503.

125

9 Wang R, Mouliswar M, Denman S, Kleban M. Mortality of the institutionalized old-old

hospitalized with congestive heart failure. Arch Intern Med 1998;158(22):2464-8.

10 Witham MD, Gillespie ND, Hutcheon SD, Struthers AD, McMurdo ME. B-type

natriuretic peptide is associated with mortality in older functionally impaired patients. J

Am Geriatr Soc 2005;53(11):1991-5.

11 deFilippi CR, Seliger SL, Maynard S, Christenson RH. Impact of renal disease on

natriuretic peptide testing for diagnosing decompensated heart failure and predicting

mortality. Clin Chem 2007;53(8):1511-9.

12 Redfield MM, Rodeheffer RJ, Jacobsen SJ, Mahoney DW, Bailey KR, Burnett JC, Jr.

Plasma brain natriuretic peptide concentration: impact of age and gender. J Am Coll

Cardiol 2002;40(5):976-82.

13 Kiely DG, Kennedy NS, Pirzada O, Batchelor SA, Struthers AD, Lipworth BJ. Elevated

levels of natriuretic peptides in patients with pulmonary thromboembolism. Respir Med

2005;99(10):1286-91.

14 Bozkanat E, Tozkoparan E, Baysan O, Deniz O, Ciftci F, Yokusoglu M. The

significance of elevated brain natriuretic peptide levels in chronic obstructive pulmonary

disease. J Int Med Res 2005;33(5):537-44.

15 Robinson BE. Epidemiology of chronic kidney disease and anemia. J Am Med Dir

Assoc 2006;7(9 Suppl):S3-S6.

16 Tombaugh TN, McIntyre NJ. The mini-mental state examination: a comprehensive

review. J Am Geriatr Soc 1992;40(9):922-35.

17 Shamsham F, Mitchell J. Essentials of the diagnosis of heart failure. Am Fam

Physician 20001;61(5):1319-28.

126

18 Quinones MA, Waggoner AD, Reduto LA, Nelson JG, Young JB, Winters WL, Jr., et

al. A new, simplified and accurate method for determining ejection fraction with two-

dimensional echocardiography. Circulation 1981;64(4):744-53.

19 K/DOQl clinical practice guidelines for chronic kidney disease: Evaluation,

classification, and stratification Am J Kidney Dis 2002;39:S1-266.

20 Izaks GJ, Westendorp RG, Knook DL. The definition of anemia in older persons. JAMA

1999;281(18):1714-7.

21 Badley EM. An introduction to the concepts and classifications of the international

classification of impairments, disabilities, and handicaps. Disabil Rehabil

1993;15(4):161-78.

22 Johnston M, Pollard B. Consequences of disease: testing the WHO International

Classification of Impairments, Disabilities and Handicaps (ICIDH) model. Soc Sci Med

2001;53(10):1261-73.

23 Porock D, Oliver DP, Zweig S, Rantz M, Mehr D, Madsen R, et al. Predicting death in

the nursing home: development and validation of the 6-month Minimum Data Set

mortality risk index. J Gerontol A Biol Sci Med Sci 2005;60(4):491-8.

24 Galasko GI, Lahiri A, Barnes SC, Collinson P, Senior R. What is the normal range for

N-terminal pro-brain natriuretic peptide? How well does this normal range screen for

cardiovascular disease? Eur Heart J 2005;26(21):2269-76.

25 Valle R, Aspromonte N, Feola M, Milli M, Canali C, Giovinazzo P, et al. B-type

natriuretic peptide can predict the medium-term risk in patients with acute heart failure

and preserved systolic function. J Card Fail 2005;11(7):498-503.

127

26 Valle R, Aspromonte N, Barro S, Canali C, Carbonieri E, Ceci V, et al. The NT-

proBNP assay identifies very elderly nursing home residents suffering from pre-clinical

heart failure. Eur J Heart Fail 2005;7(4):542-51.

27 Bibbins-Domingo K, Gupta R, Na B, Wu AH, Schiller NB, Whooley MA. N-terminal

fragment of the prohormone brain-type natriuretic peptide (NT-proBNP),

cardiovascular events, and mortality in patients with stable coronary heart disease.

JAMA 2007;297(2):169-76.

28 McKie PM, Rodeheffer RJ, Cataliotti A, Martin FL, Urban LH, Mahoney DW, et al.

Amino-terminal pro-B-type natriuretic peptide and B-type natriuretic peptide: biomarkers

for mortality in a large community-based cohort free of heart failure. Hypertension

2006;47(5):874-80.

29 Heckman GA, Misiaszek B, Merali F, Turpie ID, Patterson CJ, Flett N, et al.

Management of heart failure in Canadian long-term care facilities. Can J Cardiol

2004;20(10):963-9.

30 Mair J, Friedl W, Thomas S, Puschendorf B. Natriuretic peptides in assessment of left-

ventricular dysfunction. Scand J Clin Lab Invest Suppl 1999;230:132-42.

31 Mann JL, Evans TS. A review of the management of heart failure in long-term care

residents. Consult Pharm 2006;21(3):222-8.

32 Nygaard HA, Naik M, Ruths S, Kruger K. Clinically important renal impairment in

various groups of old persons. Scand J Prim Health Care 2004;22(3):152-6.

128

129

Chapter 4

CHRONIC HEART FAILURE, RELATED TO HELP AT ACTIVITIES OF

DAILY LIVING IN NURSING HOME RESIDENTS. Barents M, Hillege HH, Muskiet FA, de Jongste MJ.

Activities, Adaptation & Aging 2011;35(2):98-110.

130

ABSTRACT

It is unknown whether CHF is associated with ADL–functioning in nursing home residents there-

fore we studied that association.

METHOD. 103 residents were screened on age, sex, comorbidity, and CHF. ADL-dependence

and ADL-help were scored with the use of Minimum Data Set (MDS)-items of the Resident

Assistant Instrument. Logistic multivariate regressions of ADL-dependence and ADL-help were

performed with age, sex, CHF and comorbidity as independent variables.

RESULTS. age 78 ± 11 years. 24/103 had CHF (23%). CHF was associated with ADL-help

(OR 4.68 and p=0.015) after multivariable adjustment.

CONCLUSION. CHF is associated with increase of ADL-help given to nursing home residents.

CHF is of importance for both residents (loss of autonomy and life satisfaction) and organiza-

tion (workforce planning).

131

INTRODUCTION

Chronic heart failure (CHF) is a major public health care problem in the developed world be-

cause it prevails mostly at high age and takes up quite a part of public health care expenses.

Chronic heart failure is an underestimated diagnosis.

In the United States (US), nearly 5 million patients are suffering of CHF (1). CHF is primarily a

disease of the elderly. In the US and Europe, CHF affects about 1% of persons in their 50s and

rises progressively with age to afflict 10% of persons in their 80s (2). In Canadian and Dutch

nursing homes, the CHF prevalence is higher than in independent living elderly and valued at

15-23% (3;4). In North America, CHF is responsible for large health care costs (1;4;5). In de-

veloped countries, health care expenses on CHF consumes 1-2% of the entire health care

budget. In Sweden and The Netherlands, the costs of public health care are increasing gradu-

ally (5). Health care expenses consist primarily of hospitalization costs followed by costs of

care.

As noted, CHF is often undetected or incorrect diagnosed for several reasons. The key symp-

toms fatigue and dyspnea are unspecific. Older adults attribute fatigue to their age or chronic

diseases in case of multimorbidity. Fatigue and effort intolerance are gradually different symp-

toms. Oxygen poverty related to the body’s need causes fatigue and effort intolerance. Dyspnea

is often taken for chronic obstructive pulmonary disease (COPD) (7). So the lack of symptom

specificity and the contribution of symptoms to aging lead often to under diagnosis of heart fail-

132

ure. Another problem is that diagnostic tests (B-type natriuretic peptide (BNP) exclude but does

not determine CHF in older adults with multimorbidity (3;8). At last, echocardiography is not

easy accessible.

From literature, we know that CHF patients need care at activities of daily life (ADL) because

decreased effort tolerance (9;10). In independent elderly ADL-ability is vital in providing for life

satisfaction (11). Thus, care givers need to know whether persons suffer from CHF not only

from treatment but also from ADL-ability and quality of life point of view.

ADL-ability can be measured by the Resident Assisted Instrument (RAI). RAI is a core set of

assessments developed to provide a picture of, under other things, each nursing home resi-

dent’s ADL-ability (12). Care attendants observe and score specific ADL-settings. They also fill

in standardized diagnoses. RAI is widely used in American and European nursing homes as aid

for staff and care attendants (13). Remarkably, RAI does not contain CHF as standardized di-

agnosis so nursing home staffs which use RAI do not take CHF into account for care planning

and might overestimate ADL-ability of residents with CHF.

In The Netherlands, three questions gave rise to a study of CHF in nursing home residents

(residents): first, we expected under- and incorrect CHF diagnoses also in residents according

to literature in primary care (7). Second and third, the CHF prevalence in residents and the

CHF consequences on ADL-ability were unknown. Results of the first two questions studied

133

were described elsewhere (3;8). Third, we knew that Dutch nursing home staffs didn’t take

CHF into account for care level calculations. Consequently residents with CHF might be under-

valued concerning ADL-functioning and the care planning might be underestimated concerning

residents suffering CHF. We didn’t know whether CHF is associated with ADL–functioning in

residents. Subsequently, we sought to study the association of CHF to ADL-functioning in this

nursing home. We screened residents of one nursing home on CHF, scored ADL-ability with

RAI and compared residents with and without CHF on ADL-ability.

METHODS

Design

This mono-centre study had a non-randomized, prospective screening study design. The Medi-

cal Ethical Commission of the University Medical Centre of Groningen (UMCG) approved the

study under number 2004.107.

Subjects

The study population consisted of 150 residents who stayed in nursing home the “Zonnehuis-

groep Noord” in Zuidhorn, Netherlands between September 2004 and May 2005. The inclusion

criteria were competency (the resident understood the impact of study participation on himself),

age 65 years and over and signing the informed consent. The Mini Mental State Estimation

(MMSE 30) was administered with a required MMSE score of at least 21 when the resident's

134

competency was in doubt (14). The exclusion criteria were age under 65 years, serious apha-

sia, MMSE < 21, metastatic carcinomas and absence of echocardiogram or blood sample.

Data collection

Data of all subjects consisted of a) medical history from medical records, b) standardized

physical examination, c) electrocardiogram and echocardiogram, d) blood tests and e) ADL-

scores.

We collected all data of each person within one week. One physician-investigator summarized

medical history except CHF and performed physical examinations. We categorized the chronic

diseases (comorbidity) on the basis of six organ systems to limit the number of variables. The

six disease categories were chronic obstructive pulmonary diseases (COPD), diabetes, neuro-

logical diseases (ND), muscular-skeletal diseases (MSd), chronic renal dysfunction (CRD) and

blindness. ND comprised of cerebrovascular accidents (CVA), multiple sclerosis, amyotrophic

lateral sclerosis, Parkinson’s disease, ataxia, spasms, neuropathies and paresis /paralysis.

MSd included artrosis, hip artroplasty and all fractures in the past six months, (rheumatism)

arthritis, deformations of body and extremities and myopathies. CRD was defined as creatinine-

clearance ≤ 50 mL/min (Cockgroft-Gault formula). A panel of cardiologists diagnosed CHF on

the basis of medical history, symptoms, electrocardiogram, echocardiogram and blood tests.

This diagnostic method was the golden standard for CHF.

135

Standardized physical examination b) comprised examination of the cardiac pulmonary and

skeletal muscle system, height and weight.

c) Physician assistants made 12-lead-electrocardiograms (ECG) and echocardiographists

echocardiograms on location. The echocardiogram [Opti Go (www.philips.com)] estimated the

left ventricular ejection-fraction (LVEF, percentages) two-dimensionally and visually. We de-

fined a LVEF≤ 45% as left ventricular systolic dysfunction (LVSD) (15). Blood tests d) con-

sisted of a one-time-assessment of haemoglobin (Hb), mean corpuscular volume (MCV) and

creatinine.

Care attendants scored e) ADL divided in ADL-dependence (ADL-D) and ADL-help (ADL-H).

They observed dependence (D) and help (H) during four specified ADL-settings as toilet using,

eating, performing personal hygiene and walking on the ward. They scored dependence level

(D) of the resident and amount of help (H) given to the resident during these four ADL-settings

in the last seven days. ADL-D had a six-point scale so sum scores varied from 0 to 24. ADL-H

had a three-point scale so sum scores varied from 0 to 12.

Higher scores indicated more dependence and help (Table 1). Afterwards, we condensed ADL-

D and ADL–H scores each to ranges from 0 to 6, as advised by experts (www.prismant.nl). In

this way, a data set arose which is known as Minimal Data Set of the Resident Assistant In-

strument (RAI) (16-18). We selected the RAI for ADL measurements because its relative reli-

ability and validity in nursing homes. And, because the care attendants of this nursing home had

experience with RAI (8-10;13;19;20). Blood samples d) were taken from residents at rest and

136

in fasting condition, stored at -20 C° and analyzed in one go by the Laboratory Noord

(www.laboratoriumnoord.nl) to optimize validity.

Statistics

Statistical analyses were performed using SPSS 15.0.01. The question was whether CHF was

associated to more ADL-dependence and ADL-help compared to residents without CHF. Differ-

ences in base characteristics of residents with and without CHF were calculated using Student’s

t, chi-square and the Mann-Whitney test for nonparametric continuous data, as appropriate

(Table 2). Correlations between variables were calculated by Spearman’s correlation coefficient

ro (Table 3). Univariate and multivariate logistic regression analyses were made of ADL-D and

ADL-H. Age, sex, CHF and 6 chronic diseases were put in the analyses. We used the Enter-

method since this method matched theory testing best (Table 4 and 5). All statistical compari-

sons were two-tailed, and a p value <0.05 was considered to be statistically significant.

RESULTS

In total, 150 residents were recruited of whom 47 were excluded. Of the 47 excluded, 14 re-

fused participation, 30 had cognitive impairment or aphasia, 2 had inaccessible echocardiogram

frames and 1 refused to let take blood (1). Thus, 103 residents were recruited of whom 39 men

and 74 women. The distribution of age and sex did not differ between the included and ex-

cluded residents (age 78 ± 11 versus 79 ± 12, p=0.703 and sex male/female 37/63 versus

137

30/70, p=0.340 respectively). Each resident had 2 ± 1 chronic diseases in average with a

distribution of 0-6 not including CHF. Two percentages residents had no other chronic disease

(only CHF), 18% had one chronic disease, 40% had two, 28% had three and 12% had four to

six chronic diseases. The CHF group had more often chronic renal dysfunction as compared to

the non-CHF group (p=0.032). CHF involved more ADL-D and ADL-H compared to residents

without CHF (p=0.005 and p=0.009)(Table 2). For significant correlations see Table 3.

In univariate analyses, CHF was significantly associated with ADL-H after adjustment for sex

and age [odds ratio (OR)=4.44 confidence interval (CI)1.45 – 13.60; p=0.009 and OR=4.46

(CI 1.44-13.88); p=0.015, respectively].

In multivariate analyses, CHF was significant associated with ADL-H after adjustment for sex,

age and 6 chronic diseases [OR=4.68 (CI 1.35- 16.17); p=0.015](Table 5).

DISCUSSION

The question of this study is whether chronic heart failure is associated with ADL-dependence

and ADL-help in nursing home residents. We found that CHF is clearly associated with ADL-

help after adjustment for age, sex and comorbidity [OR=4.68 (CI 1.35- 16.17), p=0.015]. We

also found that CHF is not associated with ADL-dependence. The somatic chronic diseases

examined are neither associated with ADL-dependence nor ADL-help in this population.

138

What does an increase of ADL-help imply for resident and organization? At first, reception of

ADL-help implies loss of autonomy, also for residents. Second, ADL-ability is a determinant for

life satisfaction probably also in residents. Therefore, for the resident point of view, each in-

crease of ADL-help makes the difference between an autonomic or assisted performance of

ADL-activities, such as use of toilet. Thus, a resident will experience each increase of ADL-help

as loss of autonomy and life satisfaction. Each increase of ADL-help is therefore clinically most

relevant (11;12).

For the organization point of view, an increase of ADL-help means that the resident need one

or two additional care attendants. This affects the workforce planning. Thus, an increase of

ADL-help is most important for resident and organization.

Why are the six chronic diseases examined not associated with ADL-dependence and ADL-help

while some of them occur frequently? The disease category might be underestimated when

residents have more than one disease within the category. In such cases the disease merely

count as one chronic disease. Moreover, the categorization of chronic diseases used does not

represent the client’s disease burden. Generalization for the sake of categorization may have

led to the absence of significant associations of each of the six chronic diseases with ADL-

dependence and ADL-help.

How does the result that ADL-help is determined by CHF relate to literature? In his primary

care study on ADL and motor skills, Norberg finds that senior citizens with CHF are not de-

139

pendent on their personal hygiene but are in 75% of the cases dependent and get help with

putting on and taking off their compression stockings and shoes (21). Chen finds an increase in

ADL-help at 7 ADL-items from MDS, in institutionalized CHF residents (mean age 91 years), in

their final year of life (22). Scilley and Owsley prove the relation between impaired eyesight and

ADL-dependence in nursing home residents (23). Altogether, for the greater part, the results of

this study are supported by other studies. Is it possible that the etiological categorization of

chronic disorders influences its results? Is one of the chronic disorders related to CHF or are

they interrelated by a common pathophysiology such as vascular disorders? Kriegsman shows

indeed that, in the case of several chronic disorders, a common etiology leads to less loss of

function and a different etiology leads to more loss of function (9).This mono-centre study is

limited in its small population (# 103). However, when compared with two other Dutch nursing

home populations, our study population was representative in terms of distribution of age, sex,

hypertension and diabetes but not in terms of renal dysfunction (24;25). A bias on the part of

one physician-investigator who collected the data is possible. But this seems unlikely since she

did not assess electrocardiograms. Moreover, she was blinded for the final diagnoses made by

the cardiologist’ panel. Another possible limitation of the study might be that depression and

psychotropic drug use are left out of the analysis. The potential influence of these variables on

ADL-help has to be taken in consideration in further studies. The relevance of this study is that

CHF is associated to increase of ADL-help in this nursing home population. Therefore, the di-

agnosis of CHF is of importance both for resident (loss of autonomy and life satisfaction) and

organization (workforce planning). Further studies on CHF in relation to ADL are needed in lar-

140

ger NH populations. Also, subgroup analysis of CHF related to ADL are needful because it is

unknown whether only CHF with decreased LVEF is related to more ADL-help and not CHF

with preserved LVEF.

CONCLUSION

CHF is associated with increase of ADL-help given to nursing home residents. CHF is of impor-

tance for both residents (loss of autonomy and life satisfaction) and organization (workforce

planning).

ACKNOWLEDGEMENTS

The “Vereniging het Zonnehuis”, Bilthoven, The Netherlands granted the study. This study was

subsidised by the “Vereniging het Zonnehuis”, Bilthoven, The Netherlands. The authors have no

conflict of interest.

141

Table 1. Scores of ADL-dependency (ADL-D) and ADL-help (ADL-H) of MDS-items (Mini-

mum Data Set) from the Resident Assistant Instrument (RAI).

scores definition

Activities of daily living-

dependence 0 independent

1 care attendance, relatively independent

2 some help, mild restrictions

3 extensive help (with lifting) by 1 person

4 extensive help (with lifting) by 2 persons

5 dependent

6 total dependent

Activities of daily living-help 0 no help, no care needed

1 care attendance, verbal support

2 physical help by 1 person

3 physical help by 2 or more persons

Activities of daily living is scored from 0-6; Activities of daily living-help is scored from 0-3.

142

Table 2. Baseline characteristics of nursing home residents (n=103).

variables all % CHF+ % 103 CHF- % 103 P

number 103 (100) 24 (23) 79 (77)

age years, mean±SD

sex male (%)

Comorbidity COPD (%)

diabetes m. type 2(%)

neurological d. (%)

muscular-skeletal(%)

Renal dysfunction (%)

blind (%)

Physical examination tension systolic mn±SD

tension diastolic mn±SD

BMI kg/cm2, mean±SD

Blood tests Hb mmol/L, mean±SD

ADL scores ADL-D median, iqr

ADL-H median, iqr

78

39

25

26

52

44

77

3

141

80

25

7.7

2

5

(11)

(37)

(24)

(25)

(50)

(43)

(32)

(3)

(22)

(13)

(5)

(1.0)

(1-4)

(3-7)

80

11

6

9

15

7

12

0

143

78

26

7.8

4

6

(9)

(46)

(25)

(38)

(47)

(29)

(12)

(0)

(27)

(13)

(5)

(1.3)

(2-5)

(3-8)

78

18

19

17

37

37

21

3

140

81

25

7.7

2

4

(11)

(23)

(24)

(22)

(47)

(47)

(20)

(4)

(21)

(13)

(5)

(0.9)

(0-4)

(2-7)

0.351

0.358

0.925

0.114

0.181

0.096 0.032

0.083

0.534

0.590

0.897

0.834

0.005

0.009

Characteristics in numbers (percentage) unless otherwise indicated. COPD, chronic obstructive

pulmonary disorders; ND, neurological disorders; MD, musculoskeletal disorders; BMI, body

mass index; ADL, activities of daily living; ADL-D, ADL-dependence; ADL-H, ADL-help.

143

Table 3. Significant inter-variable correlations of age, sex, chronic disorders and CHF.

variable with variable correlation r p

chronic heart failure renal dysfunction 0.21 0.031

chronic heart failure ADL-dependence 0.28 0.004

chronic heart failure ADL-help 0.29 0.030

renal dysfunction age 0.42 0.000

neurological disorders ADL- dependence 0.21 0.032

neurological disorders musculoskeletal disorders 0.36 0.000

neurological disorders sex 0.25 0.010

ADL- dependence ADL-help 0.80 0.000

sex sex 0.23 0.020

ADL, activities of daily living.

144

Table 4. Associations of ADL-dependence (ADL-D) and age (univariate), sex (univariate),

CHF (age and sex adjusted) and each of the six chronic diseases (age and sex adjusted).

ADL-D

uni va-

riate OR

95% CI

P

age sex adj. OR

95% CI

P

multi varia-

te

OR

95% CI

P

age 1.00 0.96- 1.06 0.808 1.00 0.95-1.06 0.861

sex 0.72 0.26- 2.01 0.527 1.07 0.32-0.59 0.913

CHF 2.55 0.86- 7.54 0.092 2.45 0.82-7.35 0.110 2.43 0.73-8.14 0.149

COPD

0.87 0.26- 2.93 0.823 0.81 0.24-2.80 0.744 0.85 0.22-3.30 0.813

T2DM 1.63 0.54- 4.88 0.387 1.64 0.54-4.94 0.383 1.49 0.45-4.93 0.516

CRD 1.95 0.60-6.35 0.267 1.84 0.51-6.70 0.352 1.04 0.23-4.65 0.958

ND 0.93 0.32-2.75 0.897 1.12 0.32-3.87 0.862 1.93 0.47-7.93 0.361

MSd 0.32 0.10- 1.06 0.062 0.33 0.10-1.09 0.069 0.48 0.13-1.79 0.274

blind 10.50 0.90-122.81 0.061 9.72 0.82-115.67 0.072 10.73 0.73-158.83

0.084

CHF, chronic heart failure; COPD, chronic obstructive pulmonary disorders; T2DM, diabetes

mellitus; CRD, chronic renal dysfunction; ND, neurological disorders; MD, musculoskeletal dis-

orders; OR, odds ratio; CI, confidence interval; P, P-value.

145

Table 5. Associations of ADL-help (ADL-H) and age (univariate), sex (univariate), CHF (age and sex adjusted) and each of the six chronic diseases (age and sex adjusted).

ADL-H uni vari-

ate OR

95% CI

P

age sex adj. OR

95% CI

P

multi vari-

ate OR

CI

P

age 0.97 0.94-1.09 0.722 1.01 0.95-1.07 0.736 sex 1.07 0.23-4.51 0.982 0.88 0.24-3.21 0.848 CHF 4.44 1.45-13.60 0.009 4.46 1.44-13.88 0.010 4.68 1.35-16.17 0.015 COPD 0.68 0.18-2.62 0.577 0.65 0.17-2.53 0.534 0.60 0.14-2.62 0.497 T2DM 2.01 0.65-6.21 0.225 2.04 0.66-6.31 0.218 1.80 0.52-6.32 0.353 RD 1.60 0.45-5.65 0.465 1.52 0.39-6.00 0.548 0.97 0.21-4.56 0.968 ND 1.32 0.45-3.85 0.616 0.39 0.07-2.19 0.285 0.50 0.06-4.16 0.524 MS 0.56 0.18-1.75 0.317 0.56 0.78-1.75 0.316 0.75 0.20-2.83 0.673 blind 2.83 0.24-3.24 0.407 2.82 0.23-34.16 0.415 6.73 0.41-110.65 0.182 CHF, chronic heart failure; COPD, chronic obstructive pulmonary disorders; T2DM, diabetes

mellitus ; CRD, chronic renal dysfunction; ND, neurological disorders; MD, musculoskeletal dis-

orders; OR, odds ratio; CI, confidence interval; P, P-value.

146

Figure 1. ADL-dependence of residents with and without chronic heart failure.

M, median; IQR, interquartile range; CHF, chronic heart failure; ADL, activities of daily living.

.

Figure 2. ADL-help of residents with and without chronic heart failure.

M, median; IQR, interquartile range; CHF, chronic heart failure; ADL, activities of daily living.

147

References

1 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 endorsed by the European Society of

Intensive Care Medicine (ESICM). Eur Heart J 2008;29(19):2388-442.

2 Kelder JC, Cowie MR, McDonagh TA, Hardman SM, Grobbee DE, Cost B, et al.

Quantifying the added value of BNP in suspected heart failure in general practice: an

individual patient data meta-analysis. Heart 2011;97(12):959-63.

3 Wright SP, Doughty RN, Pearl A, Gamble GD, Whalley GA, Walsh HJ, et al. Plasma

amino-terminal pro-brain natriuretic peptide and accuracy of heart-failure diagnosis in

primary care: a randomized, controlled trial. J Am Coll Cardiol 2003;42(10):1793-800.

4 Cowie MR, Mosterd A, Wood DA, Deckers JW, Poole-Wilson PA, Sutton GC, et al. The

epidemiology of heart failure. Eur Heart J 1997 Feb;18(2):208-25.

5 Hobbs RE. Guidelines for the diagnosis and management of heart failure. Am J Ther

2004;11(6):467-72.

6 Zaphiriou A, Robb S, Murray-Thomas T, Mendez G, Fox K, McDonagh T, et al. The

diagnostic accuracy of plasma BNP and NTproBNP in patients referred from primary

care with suspected heart failure: results of the UK natriuretic peptide study. Eur J Heart

Fail 2005;7(4):537-41.

7 Izaks GJ, Westendorp RG, Knook DL. The definition of anemia in older persons. JAMA

1999;281(18):1714-7.

148

8 Achterberg WP, Holtkamp CC, Kerkstra A, Pot AM, Ooms ME, Ribbe MW.

Improvements in the quality of co-ordination of nursing care following implementation of

the Resident Assessment Instrument in Dutch nursing homes. J Adv Nurs

2001;35(2):268-75.

9 Holtkamp CC, Kerkstra A, Ooms ME, van CC, Ribbe MW. Effects of the implementation

of the Resident Assessment Instrument on gaps between perceived needs and nursing

care supply for nursing home residents in The Netherlands. Int J Nurs Stud

2001;38(6):619-28.

10 Morris JN, Murphy K, Nonemaker S. Interrai resident assesment instrument versie 2.0

gebruikershandleiding. Utrecht: SIG zorginformatie Mailiebaan 50 3508 SC Utrecht The

Netherlands; 96 A.D.

11 Mutasingwa DR, Ge H, Upshur RE. How applicable are clinical practice guidelines to

elderly patients with comorbidities? Can Fam Physician 2011;57(7):e253-e262.

12 Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, et al.

Diagnosis and management of the metabolic syndrome: an American Heart

Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation

2005;112(17):2735-52.

13 Barents M, van der Horst IC, Voors AA, Hillege JL, Muskiet FA, de Jongste MJ.

Prevalence and misdiagnosis of chronic heart failure in nursing home residents: the role

of B-type natriuretic peptides. Neth Heart J 2008;16(4):123-8.

14 Voors AA, Walma EP, Twickler TB, Rutten FH, Hoes AW. [Multidisciplinary guideline

'Heart failure 2010']. Ned Tijdschr Geneeskd 2011;155:A2957.

149

15 Barents M, van der Horst IC, Voors AA, Hillege JL, Muskiet FA, de Jongste MJ.

Prevalence and misdiagnosis of chronic heart failure in nursing home residents: the role

of B-type natriuretic peptides. Neth Heart J 200;16(4):123-8.

16 Mason JM, Hancock HC, Close H, Murphy JJ, Fuat A, de BM, et al. Utility of biomarkers

in the differential diagnosis of heart failure in older people: findings from the heart failure

in care homes (HFinCH) diagnostic accuracy study. PLoS One 2013;8(1):e53560.

17 Oudejans I, Mosterd A, Bloemen JA, Valk MJ, van VE, Wielders JP, et al. Clinical

evaluation of geriatric outpatients with suspected heart failure: value of symptoms, signs,

and additional tests. Eur J Heart Fail 2011;13(5):518-27.

18 Fernhall B. Long-term aerobic exercise maintains peak VO(2), improves quality of life,

and reduces hospitalisations and mortality in patients with heart failure. J Physiother

2013 ;59(1):56.

19 Dontje ML, van der Wal MH, Stolk RP, Brugemann J, Jaarsma T, Wijtvliet PE, et al.

Daily Physical Activity in Stable Heart Failure Patients. J Cardiovasc Nurs 2013 Feb 14.

150

151

Chapter 5

PHYSICAL EXERCISE PERFORMED BY RESIDENTIAL ELDERLY OF

HIGH AGE, IS SAFE HOWEVER DOES NOT AFFECT HEART FAIL-

URE SYMPTOMS, NT-PROBNP AND CARDIOVASCULAR-

METABOLIC INDICES; A PILOT, SINGLE-BLINDED RCT. Barents M, Weening B, de Greef MH, Muskiet FAJ, Hillege HL, DeJongste MJ.

Submitted.

152

ABSTRACT

Does exercise benefit care-home residents afflicted with cardiovascular-metabolic diseases who

are leading a sedentary life? Exercise effects were investigated on chronic heart failure (CHF)

and the metabolic syndrome. METHODS. For 16 weeks a twice weekly exercise was performed by the exercise group and a

non-physical program by the control group. Cardiovascular-metabolic markers were measured

[(N-terminal-pro B-type-natriuretic peptides(NT-proBNP), waist, HbA1c, diabetes mellitus type

2 (T2DM), plasma lipids and blood pressure]. RESULTS. Analysis of aged 85 years showed neither significant changes in exercise groups

(24) compared to controls (22) nor lasting injury from falls. Only exercise group attendance is

inversely related to non-fasting triglycerides (p=0.024). CONCLUSION. In care-home residents, guided physical exercise is safe in the short term but

may be less effective in reducing chronic heart failure symptoms, NT-proBNP concentrations

and the metabolic syndrome. We have explained the neutral results by the high age (85 year

on average) accompanied by low muscle endurance, and by a training schedule that might be

not enough personalized.

153

INTRODUCTION

The Dutch guideline for cardiovascular risk management (CVRM) states that the 10-year risk of

cardiovascular diseases (CVD) of 70+ elderly should exceed 20% and that the elderly should

be eligible for medication when lifestyle changes fail (1).

Care home elderly (residents) are even more at cardiovascular (CV) risk when compared to

70+ elderly because the first group (of elderly) are sedentary most often and they are suffering

from multimorbidity. Examples of multimorbidity are diabetes mellitus type 2 (T2DM), CV dis-

eases, notably heart failure (CHF), cerebrovascular ischemic attack (CVA) (2). In the daily practice of geriatric medicine, these risks are addressed pharmacologically rather

than by a change of life style such as physical activity (1;3;4). Changes of lifestyle require mo-

tivation to adjust behaviour while drug intake barely takes extra effort. This is one of the reasons

why residential elderly remain sedentary. Their daily physical activity is three to eleven minutes.

However, this inactivity of residents is inconsistent with the recommendations of the CVRM

guideline (5-9). This inconsistency is also based on perspectives of a limited lifespan and fear

of falling. These anxieties prevent residents from becoming physically active. Chronic heart

failure (CHF) is defined as a syndrome resulting from myocardial muscle dysfunction with usual

exposure of characteristic symptoms (10). Natriuretic peptides, notably B-type natriuretic pep-

tide (BNP) and amino-terminal pro-B-type NP (NT-proBNP) are well established CHF bio-

markers. They are widely used for CHF diagnosis, follow-up and prognosis (11), while they also

provide opportunities for biomarker-guided treatment optimization (12).

The metabolic syndrome was defined as the presence of at least three out of five of the follow-

ing symptoms (13): central obesity, compromised glucose homeostasis or diagnosed T2DM,

atherogenic dyslipidemia and hypertension (14;15). Both CV and metabolic derailments have

atherosclerosis as a central pathophysiological phenomenon that is eligible for lifestyle changes

as physical exercise.

In the present study we sought to investigate two aspects of physical exercise with/in residents,

its safety and changes of CHF symptoms and a marker, and markers of the metabolic syn-

154

drome. We aimed to study whether participation in a predefined exercise program is safe and

whether symptoms and a marker of CHF and markers of the metabolic syndrome, will change

after such a program. x

METHODS

This pilot study was designed as a multicenter single-blinded randomized controlled trial. The

end points were CHF symptoms, a marker, and the metabolic syndrome. This study was part of

the Benefit study (). The Benefit study took place in 14 care homes. The care home residents

in the present study participated in the same intervention program as those in the Benefit study.

This program consisted of either physical exercise or a program without exercise. The Benefit

and present study differed in size (Benefit: 14 care homes, present study: 4 care homes) and

end points. The Benefit study end points consisted of physical fitness, functional performance,

activities of daily life and quality of life (16).

Subject selection. From March 2010 to December 2011, we invited all elderly of four residential homes in Gronin-

gen to participate in the study. The inclusion criteria were: no dementia, the ability to walk at

least ten meters, understanding the impact of the study. Potential participants had to sign an

informed consent. Exclusion criteria were age <70 years, disturbed cognition (Mini-Mental

State Examination test <20; range 0-30), serious aphasia, heart failure New York Heart Asso-

ciation (NYHA)-class IV, terminal phase, and metastatic cancer. To the best of our knowledge

there was no literature available about effects of exercise programs on patients with CHF and

CV-metabolic indices among the elderly with multimorbidities. Since a meaningful estimation of

the number of participants needed for a power analysis was debatable we decided to perform a

pilot study.

155

Endpoint definitions (Table 1) Chronic heart failure was defined as a syndrome resulting from myocardial muscle dysfunction

with the usual exposure of characteristic symptoms such as shortness of breath (dyspnoea) and

fatigue, both in rest, during activity and during exertion (in three conditions)(10). Symptoms of

CHF were estimated as changed if dyspnea (scores 0-3) and or fatigue (scores 0-3) de-

crease with at least one score point. The metabolic syndrome was defined as the presence of at

least three out of five of the following symptoms (17): central obesity (waist circumference ≥102

males or ≥88 females, compromised glucose homeostasis (increased glycated hemoglobin >41

mmol/mol or the diagnosis T2DM), atherogenic dyslipidemia (high triglycerides >150 mg/dL,

low HDL-cholesterol <40 mg/dL for males or <50 for females) and increased systolic blood

pressure (>130 mm Hg) (14;15). Smoking is not included in this enumeration. Participation in

the exercise program was defined as safe if falls or unfavourable events during exercise ses-

sions occurred in <4% [16 out of 32 followed sessions x 24 (number of) exercising resi-

dents=384 occasions/100%=around 4% accidents]. The chosen changes of endpoint variables

are represented in Table 1.

Motives for non-participation. As expected (17), a relatively low participation was taken into

account. Participants were randomly assigned to the two groups at each location. Intending to

analyse motives for not participating, we inquired those who refused participation about their

reasons. In order to compare characteristics of non-participants to those of participants, biomet-

ric data of the first group (age, weight, height, heart rate and blood pressure) were collected by

the investigator. Medical history and medication were obtained from general practitioners. The

reasons for refusal were grouped into intrinsic motives (‘no realization of benefits of exercise’

and ‘does not want to adhere to the program’) and extrinsic motives. The latter were defined as

‘too busy with other activities’, illness, hospital admission, moving and impending death.

156

Subjects

During two visits before and after the intervention, all participants were subjected to a structured

interview, specific physical examination and biometric measurements. Blood samples were

taken within two hours after breakfast and lunch. Participants were physically examined for car-

diopulmonary symptoms, blood pressure (Heine gamma-G5®; www.Heine.com), symptoms or

complaints of the abdomen, joints, back, visual ability and hearing. Multimorbidity was scored

using the Charlson Comorbidity Index (18;19). Blood samples were taken two hours after

breakfast or lunch. For logistic reasons, it was not feasible to collect fasting blood samples. The

Medical Ethical Committee of the UMCG approved of the study (number 2010.178).

Interventions

The exercise program included moderate to high-intensity training sessions (16).

In each session, the participants performed progressive resistance training of upper and lower

extremities and trunk, static and dynamic balance training, and functional training. Elastic bands

(20) were used for resistance training. The following muscle groups were trained: biceps, tri-

ceps, shoulder, back, abdominal and hip muscles, quadriceps femoris, gastrocnemic, peroneus

and tibialis as well as foot muscles. The volume of resistance training was gradually increased

from one set of eight repetitions in the first five weeks to two sets of eight repetitions in weeks

six to ten and finally to three sets of eight repetitions in weeks eleven to sixteen. Balance train-

ing was adjusted to individual abilities. The level of difficulty was established by observing the

participants performing a balance task, e.g. walking on a fitness mat. The participants’ balance

was progressively challenged by increasing the level of difficulty every two weeks. Functional

training comprised exercises like chair stands, walking, turning, walking on a course with obsta-

cles. Frequency and duration of both group sessions were one hour twice weekly/two times a

week during sixteen weeks, 32 in total. The controls followed a non-physical social program

with information, games and videos in the same weeks. The participants were not allowed to

participate in other physical activity programs except current physiotherapy.

157

Clinical chemical analyses

Hemoglobin, estimated glomerular filtration rate (eGFR), HDL-cholesterol, triglycerides and

HbA1c were measured by standard procedures. NT-proBNP was measured by immune-assay

(www.roche-diagnostics.com).

Statistics All statistical analyses were performed with SPSS version 18.0 for Windows. Results were ex-

pressed as means ± SD or medians [interquartile range (IQR)] for parametric and non-

parametric data, respectively. Available data of non-participants were compared to data of par-

ticipants. Characteristics of exercise and control groups were compared at baseline. Differences

within groups between end-point minus baseline of each variable were tested by a paired t-test

or Wilcoxon-signed-ranks-test. Differences across both groups were tested by unpaired tests (t-test or Mann-Whitney U test). A sub-analysis was made, aimed at those participants who com-

pleted 16 or more (≥50%) of all sessions, who were defined as ‘as treated’. In case of insignifi-

cant differences in the outcomes of the ‘intention-to-treat’ and ‘as treated’ statistical analyses,

we only report outcomes of the ‘intention-to-treat’ analysis. All statistical comparisons were two-

tailed. A p<0.05 was considered statistically significant.

RESULTS Subjects’ characteristics

The number of subjects in four care homes consisted of 434 residents. Of these, 163 (38%)

were eligible to participate. There were 52/434 (12%) subjects willing to join the study who

were included (participants), and 111/434 (26%) who refused to take part (non-participants).

Of the non-participants, 47/111 (42%) were willing to provide us some biometric data only

once. They also explained their motives not to take part in the study:23/47 repulsed participa-

tion because of intrinsic and 24/47 because of extrinsic motives. Fifty-two participants were

158

assessed at baseline, of which 28/52 were randomized to the exercise group and 24/52 to

the controls. During the study, 6/434 (1%) participants dropped out because of not exercise-

related reasons: one died after a CVA, two moved to other institutions and three discontinued

their participation for intrinsic reasons. Their data were excluded from the final analysis. Out of

52 participants, 46 completed the program, 24/28 of the exercise group and 22/24 of the

controls. No differences were observed between characteristics of the interviewed non-participants

(n=47) and participants (n=46), except that the non-participants had a more disturbed cogni-

tion compared to the participants (MMSE 30 22 vs. 25; p=0.011). The participants were 85 years on average and 65% of them were female (Table 1). T2DM

occurred in 17% with a mean HbA1c of 43 ± 8 mmol/mol. Half of them used antihyperten-

sives. Nine (20%) had CHF. More females than males and more residents with CHF than without CHF were enrolled in the

exercise group compared to controls, respectively (79% vs. 50%; p=0.041 and 8 vs. 1;

p=0.028). In the exercise group, median NT-proBNP (446 vs. 251 pg/mL), systolic blood

pressure (143 versus 127 mm Hg) and antihypertensive use (83% versus 55%) respectively,

were significantly higher compared to controls. The exercise group differed from the control

group neither in fatigue nor in dyspnea. One fall without permanent injury was reported during an exercise session (Table 3). The usual physical activity consisted of one walk a day to an

indoor restaurant. Residents seldom visited the garden outside. For distances longer than 50-

150 m a wheelchair was used.

Comparison of exercise and control groups before and after intervention Within the exercise group, the female waist changed significantly [-8(-15- -2) cm]. The triglyc-

erides changed insignificantly [-6(-19-8)] within the exercise group and significantly [-8(-17-

0)] in the controls. Between-group comparison revealed relevant changes in female waist and

159

systolic BLOOD PRESSURE, though the findings were not statistically significant. The other variables did show neither significant within-group nor between-group differences (Table 4).

Sub-analysis. In the sub-analysis of participants who attended ≥16 of the sessions (‘as treated’)

we found a decrease of female waist within the controls [-11(-22-0)cm, p=0.049)], and within

the exercise group a decrease of triglycerides [-11(-19--1) mg/dL, p=0.032] and an increase

of NT-proBNP 81(8-155), p=0.33], both insignificant when compared between exercise and control group (supplemental Table 4a). There were no other significant or worthwhile between-

group differences detected in this ‘as treated’ sub-analysis. Attendance and endpoints. The

number of attended sessions was not associated with age, gender and comorbidity. However, it

was inversely related with changes of triglycerides (p=0.024) in the exercise group only.

DISCUSSION

In this study, we questioned whether participation in a predefined exercise program was safe

and effective to change symptoms, a marker of CHF, and markers of the metabolic syndrome.

Besides, we regarded participation rates in the interventions. In this pilot study, we found the

applied exercise program to be safe for residents (1 fall, no lasting injuries). We observed that

neither metabolic markers nor a marker and symptoms of CHF changed after completing a one

hour moderate-to-high intensity exercise program, followed twice weekly over a period of 16-

weeks, when compared to controls. It was found that the number of attended sessions in the

exercise group was inversely related to the change of non-fasting triglycerides from baseline to

study end. This relation was however not confirmed by a triglycerides decrease after exercise.

The change in triglycerides suggest that some metabolic change occurred, but this was at most

genuinely minor. Of the eligible residents, we found one-third (32%) motivated, one-third not

willing for intrinsic and extrinsic reasons and one-third unmotivated to participate with unknown

motives .

160

To explain the results, we assessed population characteristics, exercise frequency, safety, par-

ticipation and attendance. At last, the results of the Benefit study were added to our outcome.

Firstly, plasma levels of Hb, HbA1c, triglycerides, HDL-cholesterol, and systolic blood pressure

were all within reference limits before the intervention. The renal function was only slightly af-

fected. Therefore, it was estimated that participants were properly set up on medications for CVD, CHF and metabolic syndrome (e.g. the majority used antihypertensives (Table 3). It was

noted that doses of antidiabetics, antihypertensives and digitalis remained unchanged during the

study period. Secondly, we also found that participants had fewer complaints. Only 27% of the

exercise and 41% of the control group complained of fatigue and or dyspnea. Thirdly, the high

age of participants might explain partly why symptom and marker changes did not occur. In

general, people aged 50 yrs and older, lose muscle strength (mass) and muscle endurance

(mitochondria), especially when they lead a sedentary life (21). Endurance training, which is

the best exercise to increase or maintain mitochondrial concentration when aging, has generally

resulted in relatively small functional benefits for care home residents (22;23). This implies that

the high-age related changes in muscular and CV physiology, augmented by the CHF disease

state in some of our participants, limited the exercise capacity and its corresponding benefits.

Surprisingly, the studied population was stable thanks to medication. Fourthly, the high intra-

individual variation of NT-proBNP (24) explained the lack of results, which precluded the detec-

tion of minor changes, if any, in a small study group.

Although the exercise program used was specifically developed for our study population (16), it

was questioned whether a (three-times weekly) frequency of three times a week might have

provided more changes. However, we preferred a twice weekly frequency for fear of more drop-

outs and based on promising results of others (25).

Measures taken to ensure safety consisted not only of the individual exertion level but also of

guidance of the training and the own environment where the training sessions were held. Till

now, effective management of (the) fear of falling is lacking (26).

161

The participation rate of our study consisting of 32% inclusion and 68% refusal, was compared

with that of another group of 65 years old residents with 49% inclusion and 6% refusal (17).

This comparison suggested that the higher age of our participants was due to the lower partici-

pation rate. Another reason for the found lower participation rate might be interference of other

social activities with our interventions. In addition, the reported non-participants were related to

those with a lower cognition. This suggested that a certain level of cognition may be required to

participate in a study design as is applied in our study.

One positive aspect was noted. Most participants remained in the program. Their motivation

appeared from fewer drop-outs (12%) compared to the number of drop-outs (>20%) found in

another group of 72 year old exercising residents (17).

In this study, the attendance (up to 60%) was lower compared to that reported by others (31).

Accordingly, we analyzed the relations between attendance and all outcome variables in a sub-

group of participants with the best attendance ( ≥50% attendance). Compared to the entire

group, this subgroup did not exhibit more changes in metabolic markers and symptoms and a marker of CHF (supplemental Table 4a). It is therefore unlikely that our results are based on

lower attendance. Other reasons must account for this.

Using the same exercise program in comparable cohorts, it is rational to summarize the conclu-

sions of the present study and the Benefit study (20). Both studies conclude that the exercise

program is ineffective in reducing disability, care-dependency and cardiovascular-metabolic

endpoints in care home residents. However, on the basis of these two studies, the question

whether we should advise physical exercise to residents in order to promote health and func-

tionality or not, cannot be answered. As a consequence, we still do not know whether the

CVRM recommendations on physical exercise are applicable to care home residents. Subse-

quently, studies in care home residents are warranted. The limitations of the present study were

the skewed distribution of CHF and gender following randomization and the non-fasting triglyc-

erides measurement, which could, unfortunately, not be avoided because of logistical reasons.

162

CONCLUSION

In care-home residents, guided physical exercise is safe in the short term but may be less ef-

fective in reducing chronic heart failure symptoms, NT-proBNP concentrations and the metabolic

syndrome. We have explained the neutral results by the high age (85 year on average) accom-

panied by low muscle endurance and by a training schedule that might be not enough personal-

ized.

ACKNOWLEDGEMENTS

The “Hanze University of Applied Sciences” and “Stichting de Hoven”, both in Groningen, The Netherlands, granted the study. Conflict of interest statement: none.

163

References

1 Wiersma T, Smulders YM, Stehouwer CD, Konings KT, Lanphen J. [Summary of the

multidisciplinary guideline on cardiovascular risk management (revision 2011)]. Ned

Tijdschr Geneeskd 2012;156(36):A5104.

2 Marengoni A, Rizzuto D, Wang HX, Winblad B, Fratiglioni L. Patterns of chronic

multimorbidity in the elderly population. J Am Geriatr Soc 2009;57(2):225-30.

3 Mancia G, De BG, Dominiczak A, Cifkova R, Fagard R, Germano G, et al. 2007 ESH-

ESC Practice Guidelines for the Management of Arterial Hypertension: ESH-ESC Task

Force on the Management of Arterial Hypertension. J Hypertens 2007;25(9):1751-62.

4 Kroon AA. [Dutch guideline for cardiovascular risk management: a consensus model].

Ned Tijdschr Geneeskd 2012;156(36):A5196.

5 Nelson ME, Rejeski WJ, Blair SN, Duncan PW, Judge JO, King AC, et al. Physical

activity and public health in older adults: recommendation from the American College of

Sports Medicine and the American Heart Association. Med Sci Sports Exerc

2007;39(8):1435-45.

6 Jans M.P. De ontwikkeling van een beweegnorm voor ouderen in verpleeg- en

verzorgingshuizen. Vreede de P., Tak E.C.P.M., Meeteren van N.L.U., editors. TNO

kwaliteit van leven. 046[KvL/B&G], 9-15. 1-9-2008. www.tno.nl, TNO.

7 Christmas C, Andersen RA. Exercise and older patients: guidelines for the clinician. J

Am Geriatr Soc 2000;48(3):318-24.

8 Taylor AH, Cable NT, Faulkner G, Hillsdon M, Narici M, Van Der Bij AK. Physical activity

and older adults: a review of health benefits and the effectiveness of interventions. J

Sports Sci 2004;22(8):703-25.

164

9 Mutasingwa DR, Ge H, Upshur RE. How applicable are clinical practice guidelines to

elderly patients with comorbidities? Can Fam Physician 2011;57(7):e253-e262.

10 Mosterd WL, Rosier PF. [Guideline 'Chronic heart failure']. Ned Tijdschr Geneeskd

2004;148(13):609-14.

11 McMurray JJ, Adamopoulos S, Anker SD, Auricchio A, Bohm M, Dickstein K, et al. ESC

Guidelines for the diagnosis and treatment of acute and chronic heart failure 2012: The

Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure 2012 of

the European Society of Cardiology. Developed in collaboration with the Heart Failure

Association (HFA) of the ESC. Eur Heart J 2012;33(14):1787-847.

12 Morello AM, Januzzi JL. Amino-terminal pro-brain natriuretic peptide: a biomarker for

diagnosis, prognosis and management of heart failure. Expert Rev Mol Diagn

2006;6(5):649-62.

13 Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, et al.

Diagnosis and management of the metabolic syndrome: an American Heart

Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation

2005;112(17):2735-52.

14 Executive Summary of The Third Report of The National Cholesterol Education Program

(NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood

Cholesterol In Adults (Adult Treatment Panel III). JAMA 2001;285(19):2486-97.

15 Brunzell JD, Davidson M, Furberg CD, Goldberg RB, Howard BV, Stein JH, et al.

Lipoprotein management in patients with cardiometabolic risk: consensus statement from

the American Diabetes Association and the American College of Cardiology Foundation.

Diabetes Care 2008;31(4):811-22.

165

16 Weening-Dijksterhuis E, de Greef MH, Scherder EJ, Slaets JP, van der Schans CP. Frail

institutionalized older persons: A comprehensive review on physical exercise, physical

fitness, activities of daily living, and quality-of-life. Am J Phys Med Rehabil

2011;90(2):156-68.

17 Nyman SR, Victor CR. Older people's recruitment, sustained participation, and

adherence to falls prevention interventions in institutional settings: a supplement to the

Cochrane systematic review. Age Ageing 2011;40(4):430-6.

18 Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity

index. J Clin Epidemiol 1994;47(11):1245-51.

19 Charlson ME, Charlson RE, Peterson JC, Marinopoulos SS, Briggs WM, Hollenberg JP.

The Charlson comorbidity index is adapted to predict costs of chronic disease in primary

care patients. J Clin Epidemiol 2008;61(12):1234-40.

20 Baum EE, Jarjoura D, Polen AE, Faur D, Rutecki G. Effectiveness of a group exercise

program in a long-term care facility: a randomized pilot trial. J Am Med Dir Assoc

2003;4(2):74-80.

21 Rogers MA, Evans WJ. Changes in skeletal muscle with aging: effects of exercise train-

ing. Exerc Sport Sci Rev 1993;21:65-102.

22 Fiatarone MA, O'Neill EF, Ryan ND, Clements KM, Solares GR et al. Exercise training

and nutritional supplementation for physical frailty in very elderly people. N Engl J Med

1994;330(25):1769-1775.

23 Tseng BS, Marsh DR, Hamilton MT, Booth FW. Strength and aerobic training attenuate

muscle wasting and improve resistance to the development of disability with aging. J

Gerontol A Biol Sci Med Sci 1995;50 Spec No:113-9.

166

24 Bruins S, Fokkema MR, Romer JW, Dejongste MJ, van der Dijs FP, van den Ouweland

JM, et al. High intraindividual variation of B-type natriuretic peptide (BNP) and amino-

terminal proBNP in patients with stable chronic heart failure. Clin Chem

2004;50(11):2052-8.

25 Chin APM, van Poppel MN, Twisk JW, van MW. Effects of resistance and all-round,

functional training on quality of life, vitality and depression of older adults living in long-

term care facilities: a 'randomized' controlled trial [ISRCTN87177281]. BMC Geriatr

2004;4:5.

26 Parry SW, Deary V, Finch T, Bamford C, Sabin N, McMeekin P, et al. The STRIDE

(Strategies to Increase confidence, InDependence and Energy) study: cognitive

behavioural therapy-based intervention to reduce fear of falling in older fallers living in

the community - study protocol for a randomised controlled trial. Trials 2014;15:210.

27 Rydwik E, Gustafsson T, Frandin K, Akner G. Effects of physical training on aerobic

capacity in frail elderly people (75+ years). Influence of lung capacity, cardiovascular

disease and medical drug treatment: a randomized controlled pilot trial. Aging Clin Exp

Res 2010;22(1):85-94.

167

Table 1. Endpoints

Primary endpoints At risk for

all Male Female

waist circumference (cm)a >102 >88 triglycerides (mg/dL)a >150 high-Density Lipoprotein-cholesterol (mg/dL)a <40 <50 systolic blood pressure (mm Hg)a >130 diastolic blood pressure (mm Hg)a >95 non-fasting glucose (mg/dL)a <140 glycated Hemoglobin (mmol/mol) >41 Secondary endpoints fatigue in 3 activity levels (1 – 3)

>1

dyspnea in 3 activity levels (1 – 3) >1 N-terminal pro B-type natriuretic peptide( pg/mL)b

>450

a Based on definition of the metabolic syndrome. The severity of fatigue and dyspnea are based on the definition of the NYHA classification: class 1: no symptoms; class 2: symptoms occur at ordinary activity; class 3: symptoms occur ordinary activity. b NT-proBNP cut-off value for the diagnosis of heart failure (11).

168

Table 2. Sample description. Participants are compared to non-participants at baseline.

Variable Participants

N = 46

Non-participants

N=47

P

male/female 16/30 9/38 0.151

age (years) 85 ± 6 84 ± 8 0.658

body mass index kg/m2 27 ± 5 25 ± 65 0.258

mini-mental state estimation 30 25 ± 4 22 ± 4 0.011

heart rate (beats/min) 73 ± 11 71 ± 25 0.522

smoking median (interquartile range) 3 (1-4)

heart failure n (%) 9 (%)

Cerebro-vascular accident n (%) 16 (%)

comorbidity 11 (%)

residents with antidiabetics n (%) 6 (13)

residents with antihypertensives n(%) 32 (70)

Data are presented as means ± SD. Comorbidity, Charlson Comorbidity Index (sum of current

chronic diseases out of 17 chronic diseases).

169

Table 3. Characteristics of exercise and control groups of care home residents. Variable Exercise group

(Ex) n=24 Controls (C) n=22

Ex vs. C P-value

age years mn SD 85±5 84±6 0.298 male/female number/number 5/19 11/11 0.041 waistm cm mn SD 107±15 113±28 0.692 waistf cm mn SD 105±21 97±12 0.690 mini-mental state examination mn SD 26±3 25±4 0.421 smoking number 4 7 0.273 fatigue n (range 0-3) 1±1 1±1 0.757 dyspnea n (range 0-3) 1±1 1±1 0.859 heart rate (beats/min) 71±12 76±9 0.154 diastolic blood pressure (mm Hg) mn SD 85±12 81±12 0.263 systolic blood pressure (mm Hg) mn SD 143±9 127±18 0.005 comorbidity number md IQR 3(0-4) 3(1-4) 0.705 cerebrovascular accidents number 7 9 0.353 diabetes mellitus type 2 number 6 2 0.183 % residents using antidiabetics 21 1 0.099 chronic heart failure number 8 1 0.028 % residents using antihypertensives 83 55 0.037 digoxin % 8 5 0.671 NT-proBNP pg/mL median IQR 446 (260-967) 251 (93-664) 0.024 eGFR mL/min/1.73 m2 mn SD 58±22 70±27 0.098 hemoglobin mg/dL mn SD 13±2 13±2 0.871 glucose mg/dL mn SD 118±38 101±20 0.061 HbA1c mmol/mol mn SD 43±10 43±6 0.844 triglycerides mg/dL mn SD 122±56 131±66 0.633 HDL-cholesterol male mg/dL mn SD 40±6 51±16 0.139 HDL-cholesterol female mg/dL mn SD 51±11 54±22 0.690 attended sessions% 60±37 58±34 0.840 fall without injury number 1 0 0.852 IQR, interquartile range; NT-proBNP, amino-terminal proBNP; eGFR, estimated glomerular fil-tration rate.

170

Table 4. Changes (Dif) in the exercise and control groups from baseline to the study end,

together with between-group differences of these changes.

Variable Difference

Exercise (Ex)

Within group

Control (C)

Within group

Ex vs. C Between

group adjusted

Dif Mn (95%CI) P Mn (95%CI)

P Mn (95%CI)

P

Dif waist male cm -1 (-4 – 1) 0.135 -3 (-1 – 2) 0.155 -2 (-6 – 2) 0.381

Dif waist female cm -8 (-15- -2) 0.018 -2 (-6 – 3) 0.405 -4 (-10 – 1) 0.076

Dif triglycerides mg/dL -6 (-19 – 8) 0.394 -8 (-17- 0) 0.033 -6 (-11- -1) 0.271

Dif HDL-C male mg/dL 0 (-3 – 3) 0.951 -6 (-18– 6) 0.274 -3 (-11- 5) 0.773

Dif HDL-C female mg/dL 1 (-3 – 5) 0.577 2 (0 – 5) 0.086 2 (-1- 4) 0.948

Dif DBP mm Hg -4 (-7 – 0) 0.029 -3 (-10– 4) 0.378 -4 (-8 – 1) 0.935

Dif SBP mm Hg -1 (-14 – 11) 0.830 2 (-7 – 11) 0.681 -4 (-7- 6) 0.059

Dif glucose mg/dL -13 (-27–2) 0.078 7 (-5 – 20) 0.228 -7(-26– 13) 0.127

Dif HbA1c mmol/mol 3 (0 – 6) 0.086 1 (-1 – 3) 0.329 0 (0 – 0) 0.579

Dif fatigue n 0 (-1 – 0) 0.137 0 (-1 – 0) 0.261 0 (-1- 0) 0.930

Dif dyspnea n 0 (-1 – 0) 0.328 0 (0 – 0) 0.162 0 (-1- 0) 0.696

Dif NT-proBNP

pg/mL*(md IQR)

334(-117– 785) 0.139 67(7–126) 0.030 22(-9– 140) 0.948

mn, means; SD, standard deviation; * md, median with IQR, interquartile range; MMSE30,

mini-mental state examination 30; comorbidity, Charlson Comorbidity Index (sum of current

chronic diseases out of 17 chronic diseases); CVA , cerebrovascular accidents; DBP, diastolic

blood pressure; SBP, systolic blood pressure; HDL-C, HDL–cholesterol; n, number; NT-

proBNP, amino-terminal proBNP, this variable is tested non-parametrically; eGFR, estimated

glomerular filtration rate; attendance, percentage attended sessions.

171

Supplemental Table 4a. Subgroup of attendance ≥16 sessions: changes (Dif) in exercise

and control groups from baseline to the study end, together with between-group differences

of these changes.

Variable Differences

Exercise Within-group

Control Withingroup

Ex vs. C Between-

group adjusted

(Dif) Mn(95%CI) P Mn(95%CI) P Mn(95%CI) P Dif waist m cm -4(-10-3) 0.270 -1(-4-1) 0.135 2(-6-10) 0.601 Dif waist f cm -2(-7-3) 0.408 -11(-22-0) 0.049 -9(-22-5) 0.190 Dif triglycerides -11(-19--1) 0.032 -5(-12-2) 0.134 5(-6-17) 0.563 Dif HDL-C m -7(-20-7) 0.277 0(-3-3) 1.00 8(-10-25) 0.351 Dif HDL-C f 25(-29-78) 0.321 2(-1-4) 0.2100 0(-6-6) 0.986 Dif SBP 2(-10-13) 0.762 0(-14(-14) 0.993 -2(-19-16) 0.540 Dif HbA1c % 0.1(-0.1-0.2) 0.332 0.1(0.1-1.0) 0.281 0.1(-0.2-0.3) 0.736 Dif fatigue n 0(0-0) 0.543 0(0-0) 0.136 0(-1-0) 0.543 Dif dyspnea n 0(0-0) 0.332 0(0-0) 0.229 0(-1-0) 0.670 Dif NT-proBNP 81(8-155) 0.033 367(-245-981) 0.222 280(-354-914) 0.852 Data are means ±SD; *median with interquartile range (md, iqr); m, male; f, female; DBP, dia-

stolic blood pressure; SBP, systolic blood pressure; HDL-C, HDL–cholesterol; n, number; NT-

proBNP, amino-terminal proBNP this variable is tested non-parametrically.

172

173

Chapter 6

SUMMARY, DISCUSSION, RECOMMENDATIONS AND FUTURE

PERSPECTIVES

174

Chapter 6.1

SUMMARY

A drawback of the improved survival from cardiovascular diseases (CVD) (1) is that cardiovas-

cular morbidities have grown. As we have emphasized in the introduction of this thesis, the

number of people living with heart failure (HF) has increased impressively in conjunction with a

tremendously improved life expectancy of the Dutch population (www.cbs.nl;

www.nationaalkompas.nl), during the last century. Since HF is mainly a problem of advanced

age, subjects diagnosed with chronic HF (CHF) often concern elderly. However, there is a wide

lack of knowledge with regard to data on HF when age progresses and elderly become frail and

more care dependent. The most care dependent elderly are committed to a care or nursing

home. In this thesis, those elderly persons are indicated as residential elderly. The shortage of

(research) data is mainly caused by excluding residential elderly from studies because of high

age and multimorbidities, such as CVD. Since there is a deficiency of data, no specific guidelines are available to diagnose and to treat

HF in residential elderly. From this perspective, there is debate on the applicability of existing

HF guidelines developed for CHF patients in general, to residential elderly with CHF. However,

guidelines for residential elderly with CHF are sorely missing because diagnosing CHF is notori-

ously difficult: elderly tend to subscribe complaints to aging instead of CHF or other diseases.

Subsequently, CHF treatments are often incorrectly applied, or not applied at all. The conse-

quences of these caveats may be loss of quality of life and increase of costs.

Therefore, we sought to study CHF in the residential elderly.

In Chapter 2.1 we have explored the epidemiological scale of CHF, by determining the preva-

lence of CHF in residential elderly. Based on the recent literature, when general guidelines on

CHF (2;3;4) were applied to residential elderly, diagnosing CHF was less accurate due to poor

predictive values of signs and symptoms, medical history, and electrocardiogram (ECG).

175

From medical files of residential elderly in a single nursing home in Groningen, we inventoried

signs, symptoms, medical history and registered the number of subjects with CHF. Then we

performed a physical examination and an ECG of all residents. Next we estimated the accuracy

of both, former CHF diagnostic (natriuretic peptides (NPs) not used) and current CHF diagnos-

tic procedures (including NPs), since, in contrast to the broad recognition of NPs as a screening

test for CHF, their use as biomarkers in residential elderly has remained limited. The restricted

use in elderly is caused by lack of validation of cut-off values for NPs in this group of subjects.

Finally, we studied ECGs of all residents.

A panel of cardiologists decided on the presence or absence of CHF by assessment of NPs and

the result of echocardiography, successively. Echocardiographic investigations were feasible in

98%. As a result of the present study, CHF was established in 24/103 (23%) residential eld-

erly. Fifteen (15/24) residents were not previously detected with CHF. Before the study, 22

were identified with CHF. Out of the 22 residents with CHF before the study, in only 9 residents

CHF was confirmed and of 13 residents with CHF the diagnosis was rejected.

The diagnostic accuracy of NT-proBNP at 450 pg/mL was 0.71 sensitivity, 0.67 specificity,

0.42 positive predictive value (PPV) and 0.91 negative predictive value (NPV). The diagnostic

accuracy of BNP at 100 pg/mL was: 0.71 sensitivity, 0.70 specificity, 0.41 PPV and 0.88

NPV.

In brief, the most striking results of the present study were that more than half the CHF diagno-

ses were missed (i.e. 15/24) or incorrectly made (i.e. 13/22). Hence we concluded that the

overall accuracy of identifying CHF is limited. However, given the high NPVs, the use of natri-

uretic peptides as additional diagnostic instrument seems promising, even in nursing home resi-

dents but requires further evaluation. The CHF prevalence in residential elderly was estimated

at 23%. In Chapter 2.2 we studied whether the outcomes of chapter 2.1 were consistent with the CHF

prevalence in Aruban nursing homes. Aruban nursing homes were chosen since we were inter-

ested in a comparable population with Dutch nationality living in a different environment and

176

having another lifestyle. The same design as in Chapter 2.1 was utilized to study the prevalence

of CHF in Aruban residential elderly. The other aim was to validate whether employment of na-

triuretic peptides could improve identification of CHF in residential elderly in another part of the

Dutch Kingdom. The main difference with the chapter 2.1 study was that we were unable to

obtain echocardiograms of all residential elderly, due to logistic reasons. Therefore, we could

not meet the diagnostic standards for CHF (2-4).

As a result, 51 out of 235 elderly Aruban residents were included with a mean age of 78±8

years. According to the medical files 7/51 residents were acknowledged with CHF. However,

two out of the 7 residents did not have CHF. Furthermore, CHF was established in 16 out of 51

(31%) residents. Out of the 16 residents with CHF, 11 were not previously diagnosed with CHF

and of 5/16 residents identified with CHF, the diagnosis was confirmed.

In brief, the general guidelines for HF are applied to Aruban residential elderly, infrequently.

When compared to residential elderly in Groningen, the prevalence of CHF may be at least as

high in Aruban residential elderly. In conclusion, recognition of CHF appears to be severely un-

derestimated in Aruban residential elderly. The detection of CHF will be improved by implement-

ing an appropriate guideline, including the determination of BNP.

In Chapter 2.3 we addressed the problem of the large intra-individual variations [individual co-

efficient of variation (CVi)] of NT-proBNP in plasma, since large CVis limit the applicability of

NT-proBNP for among others, guided therapy optimization in individual patients with CHF (5).

In search of more reliable CVis we compared concentrations of urine NT-proBNP (NT-proBNPu)

to concentrations of plasma NT-proBNP (NT-proBNPp), in HF patients living in Curaçao.

Urine and blood samples were taken on a single day (“within-day”): six blood samples every 2

hours and spontaneously voided urines during 24-hours. On five consecutive days (“day-to-

day”): five blood samples, five enforced urine samples, and five full 24-h urines were taken. On

the same day of six consecutive weeks (“week-to-week”): one blood sample, one enforced

177

urine sample and one full 24-h urine was collected. Out of these blood and urine samples the

total CVs (CVts), CVis and reference change values (RCVs) were calculated.

In this study 25 CHF patients were included with a mean age of 61 (range 36-80) years, 60%

was male and the average left ventricular ejection fraction was 36±15%. Median CV is for NT-

proBNPp were 9% (within-day), 18% (day-to-day) and 30% (week-to-week). For NT-proBNPu

the CVis were 34%, 21% and 28%, respectively. The reference change value (RCV) of NT-

proBNPu was higher than, or equal to, the RCV for NT-proBNPp, even after correction for

creatinine and the time period (in hours) in which the urine accumulated in the bladder.

Our data of day-to-day and week-to-week samples suggest that (immunoreactive) concentra-

tions of NT-proBNP in urine do not unequivocally correlate with (immunoreactive) NT-proBNP

concentrations in plasma, up to an NT-proBNPp threshold of about 310 pg/mL (Figure 1). In

addition, we found that beyond this threshold the higher NT-proBNP concentrations, as ex-

pressed in Figure 1, are not explained by an increase in blood pressure, during the day (results

not shown). We hypothesise that filtration of natriuretic peptides may be considered as the

dominant factor beyond the threshold and that at lower NT-proBNP levels filtration and reab-

sorption are more dynamically interacting.

In conclusion, measurements of NT-proBNP in urine have no advantages in stable patients with

chronic heart failure.

178

Figure 1. Relation between NT-proBNP in plasma and NT-proBNP in the enforced urine

voidings collected in the day-to-day and week-to-week protocols.

Data derive from 22 patients and in total 182 comparisons of NT-proBNP in plasma and urine.

Day-to-day and week-to-week samples were used. For urine we have used the enforced void-

ings.

In Chapter 3 we have investigated the prognostic value of natriuretic peptides on one-year mor-

tality in the same cohort of residential elderly as studied in chapter 2.1. The clinical relevance is

that if natriuretic peptides are related to prognosis of residential elderly with CHF, these bio-

markers may be exploited for advance care planning. Advance care planning is related to qual-

ity of life and becomes important when life expectancy is restricted (www.Verenso.nl).

To investigate one-year mortality, survival and death of residential elderly was tracked, during

one year. Since ten residential elderly could not be followed up, because they moved to un-

known destinations, we studied the remaining 93 residential elderly.

179

Eighteen out of 93 residents (mean age 81 ± 3 years, 66% female) died within one year (non-

survivors).

A mutually adjusted Cox proportional hazard regression analysis was performed. Adjustments

were made for six predefined chronic diseases, immobilization, age, sex, NT-proBNP and BNP.

The results found were that both natriuretic peptides (NT-proBNP and BNP) significantly pre-

dicted one-year mortality (HR 1.02 and p=0.001, and HR 1.16 and p=0.003, respectively). In

addition, the one-year mortality risk increases exponentially in concert with higher BNP and NT-

proBNP levels. For instance, an NT-proBNP increase up to 2,000 pg/mL was associated with

a 60% higher risk of dying, within one year.

In conclusion, both BNP and NT-proBNP are independent predictors of one-year mortality for

residential elderly with CHF. In addition, the mortality risk increases at natriuretic peptide con-

centrations elevated far beyond the diagnostic cut-off values of the guidelines (2;3;4).

In Chapter 4 we studied the relation between CHF and the support that is needed for residential

elderly with CHF during their activities of daily life (ADL). To this end, the same cohort of resi-

dential elderly was examined as in chapter 2.1. The rationale for this study question was to

draw attention of care teams to the need of support among those residents with CHF.

For this study the same group of residential elderly was assessed as in chapter 2.1. To collect

data on ADL the Minimal Data Set (MDS)-items of the Resident Assistant Instrument was used.

The MDS consists of systematic observations on dependency and delivered support and aims to

quantify dependence and aid (6). The MDS instrument consists of a software program for en-

tering the observations of the individual residential elderly, during predefined activities of daily

living (ADL). An example of such an activity is walking. The care taker observes and scores

(quantifies) whether the individual is able to walk predefined distances independently and needs

help during walking. Of 103 residents with and without CHF, the ADL-dependence and ADL-

help scores were compared and adjusted for differences between the two groups by means of

180

logistic multivariate regression. CHF was associated with ADL-help (OR 4.68 and p=0.015)

however, CHF was not related to ADL-dependence.

In conclusion, CHF is associated with a necessary increase of ADL-help for nursing home resi-

dents. In this regard, CHF is of importance for both residents (loss of autonomy and life satis-

faction) and organization (workforce planning).

In Chapter 5 we assessed a physical exercise program in residential elderly with CHF. The ra-

tionale for this study was the discrepancy between the sedentary lifestyle adhered to by resi-

dential elderly and the recommendations for physical activity from HF guidelines (2;3;4). Also

family members are stimulating their parents in care homes to become physically active. How-

ever, for a variety of reasons, residential elderly prefer to remain sedentary. Furthermore, be-

cause of fear of falling, also caretakers prefer residential elderly to maintain their sedentary life-

style. On the other hand, a sedentary life style is related to an increased risk on (progression

of) atherosclerosis and CV-metabolic diseases, such as CHF and metabolic syndrome (7).

To investigate whether regular exercise is safe and has beneficial effects on CV-metabolic indi-

ces for residential elderly, we evaluated effects of exercise on CHF symptoms, NT-proBNP and

metabolic syndrome.

A pilot single-blinded, randomized controlled feasibility trial in four care homes was performed.

In collaboration with other investigators (8), we studied whether a standardized exercise pro-

gram affects CHF symptoms, NT-proBNP and the metabolic syndrome. The other investigators

(8) studied whether the same standardized exercise program affects functional endpoints in

comparable groups.

The exercise intervention included a guided moderate-to-high intensity training of one hour

twice weekly, during 16 consecutive weeks. Exercise consisted of progressive resistance train-

ing of upper and lower extremities, and of the trunk, static and dynamic balance training, and

functional training. The control group intervention contained a guided non-physical social pro-

gram of similar frequency, duration, length and location. The study endpoints consisted of

181

symptoms of CHF (fatigue, dyspnea), NT-proBNP, and body measurements and laboratory

results, together known as parameters of the “metabolic syndrome”. The syndrome is defined as

the presence of three out of five measurements and laboratory results. These measurements

include waist, blood pressure, fasting glucose, triglycerides and high-density lipoprotein (HDL)-

cholesterol. We substituted glucose for glycated hemoglobin (HbA1c), since fasting blood col-

lection was not feasible for logistic reasons.

As a result, out of 434 residential elderly 163 were eligible. Unfortunately, only 52 (12%) were

willing to participate in the study. Of these 52, six dropped out during the study for reasons not

related to the exercise. The remaining 46 participants were 85±6 years, 65% females and all

had been randomized to exercise (n=24) or control groups (n=22). The intention-to-treat

analysis showed no significant change of any of a variable in the exercise groups compared to

controls. The as-treated group completed half or more of the 32 training sessions. The females

of the latter group had a significant reduction in waist size [-8 (range -15 - -2) cm]. The con-

trols showed a significant triglycerides decrease [-8 (-17-0) mg/dL] and an NT-proBNP in-

crease [67 (7-126) pg/mL]. The attendance of the residents was inversely related to the non-

fasting triglycerides in the exercise group (p=0.024).

We concluded that over a short period of time guided physical exercise, performed by residen-

tial elderly, had neither beneficial effects on symptoms and a biomarker of CHF, nor on cardio-

vascular-metabolic indices. We have explained the neutral results with the high age (85 years

on average) accompanied by loss of muscle endurance, and with a training offer that is not

enough personalized.

182

In brief, the outcomes of this thesis are: the prevalences of CHF in residential elderly are 23% in Groningen, which is consistent

with literature (9), and >30% in Aruba (chapters 2.1 and 2.2).

in Dutch residential elderly, undetected diagnoses of CHF often occur. The use of natri-

uretic peptides, as an additional test for CHF, needs to be further studied. In residential

elderly in Groningen, the incorrect diagnoses of CHF can be reduced by applying natri-

uretic peptides (chapter 2.1 and 2.2).

determination of NT-proBNP concentrations in the urine of CHF patients is not more fa-

vourable compared to NT-proBNP plasma levels. This finding results from the large in-

traindividual variabilities of urine NT-proBNP which is comparable to that in plasma NT-

proBNP (chapter 2.3).

in residential elderly, natriuretic peptide testing has prognostic values [HR of NT-proBNP

1.02 and BNP 1.16 (vs. HR (10) of BNP is 2.2)], of which the outcome of BNP is con-

sistent with the literature (10) (chapter 3).

residential elderly with CHF need more ADL-help, when compared to those without CHF

(chapter 4).

a short-term period of guided exercise performed by residential elderly has neither effect

on CHF nor on cardiovascular indices (i.e. the metabolic syndrome). There are no stud-

ies to compare with (chapter 5).

183

184

185

Chapter 6.2

DISCUSSION

186

We assessed our hypothesis (Introduction) according to which general HF guidelines (2-4) are

not applicable to residential elderly with regard to diagnostics. Therefore, we applied natriuretic

peptides (NPs) threshold values, derived from HF guidelines (2-4) to the residential elderly

cohort in Groningen (chapter 2.1). It should be noticed that both, the ESC and Multidisciplinary

HF guidelines (2;3) advise to use the same cut-off values (NT-proBNP 125 and BNP 35

pg/mL), which values are lower compared to the NICE HF guideline (i.e. NT-proBNP 400 and

BNP 100 pg/mL) (4). Using these cut-off values from the guidelines, we calculated the pre-

dictive values for the CHF diagnosis in residential elderly of a single nursing home in Groningen (Table 1). We also calculated the predictive values at higher thresholds such as at NT-pro BNP

900 pg/mL and BNP 100 pg/mL. Years after our study, Mason et al. (11) published a paper

on predictive values of NPs for CHF. The cut-off values of their study in British residential eld-

erly were slightly different from those chosen in our study. In accordance with the NICE HF guideline (4) Mason et al. validated their findings for NT-proBNP and BNP cut-off points (Table

1).

The Mason group (11) and we not only evaluated isolated natriuretic peptides but also combina-

tions of medical history, symptoms and signs, ECG, and natriuretic peptides to predict the pres-

ence of CHF. We both found that any combination of studied parameters did not improve the predictive values, as determined by natriuretic peptides either (Table 1). The outcomes of Ma-

son’s study, specifically with regard to PPV, correspond well with our findings.

187

Table 1. Predictive values of NT-proBNP and BNP for the heart failure diagnosis in two

populations of residential elderly (RE): the Groningen (NL) (Chapter 2.1) and the British

population (11).

Cut-off value

pg/mL sensitivity specificity PPV NPV

Groningen (NL) RE

NT-proBNP1 125 1.00 0.28 0.29 1.00

NT-proBNP 450 0.71 0.67 0.42 0.91

NT-proBNP 900 0.67 0.85 0.56 0.89

BNP1 35 0.88 0.39 0.30 0.91

BNP 50 0.88 0.63 0.31 0.89

BNP2 100 0.71 0.70 0.41 0.88

British RE (11)

NT-proBNP2 400 0.56 0.69 0.35 0.84

NT-proBNP 760 0.62 0.75 0.42 0.87

BNP 115 0.67 0.68 0.38 0.88

Residential elderly, RE; PPV, positive predictive value; NPV, negative predictive value; NT-

proBNP, N-terminal amino pro B-type natriuretic peptide; B-type natriuretic peptide.

1 ESC HF guideline (2) and multidisciplinary HF guideline (3)

2 NICE HF guideline (4).

In addition, both groups report that when applying HF guidelines (2-4) to Dutch and British

residential elderly (11), more than half of the initial CHF diagnoses are missed (too many false

negatives). Also lowering the threshold values (2;3), when compared to the NICE (4) HF

guideline, does not provide any substantial improvements. We and Mason et al. further estab-

188

lished that in residential elderly, the calculated predictive values are not reliable for ruling-in but

only for ruling out CHF.

In this thesis, we thus adapt the algorithm of the Multidisciplinary HF guideline (3) to residential elderly (Figure 1). These modifications include the use of NPs for excluding CHF only, the use

of echocardiography for ruling in CHF, the use of NICE cut-off values (4), and utilization of

echocardiography in primary care laboratories (www.Certe.nl).

However, Oudejans et al. (12) demonstrated in geriatric outpatients with a prevalence of >40%

CHF, that natriuretic peptide testing in combination with other diagnostic tests, are very accurate

for both, ruling out and ruling in CHF. Therefore, the PPV of CHF in our group might have been

lower due to a lower a priori chance on having CHF based on the lower CHF prevalence (23%)

when compared to the a priori chance of CHF in outpatients (CHF prevalence >40%). In con-

trast to the results of the Oudejans group, but in concert with those of Mason and our study

group [Chapter 1.2 and reference (11)], we have revised our recommendations for residential

elderly.

We had to reject our hypothesis in part, since CHF cannot be diagnosed in residential elderly by

making use of HF guidelines. In addition, our hypothesis was fair in part because CHF can be

ruled out using natriuretic peptides according to the HF guidelines.

Based on our new insights, we advise to make an echocardiogram for diagnosing CHF in resi-

dential elderly. Theoretically, NPs are recommended to rule out CHF although we realize that

excluding CHF does not comply with daily practice. Further, We processed these recommenda-tions into an algorithm for CHF diagnostics (i.e. non-acute HF) for residential elderly (Figure

1).

189

190

191

Chapter 6.3

RECOMMENDATIONS AND FUTURE PERSPECTIVES

192

In the discussion, we have concluded that general HF guidelines cannot be used for diagnosing

CHF in residential elderly. Furthermore, echocardiography is warranted to establish the diagno-

sis of CHF in the individual residential elderly.

To that end, we have customized the Dutch diagnostic algorithm for CHF (3) to that particular group (Figure 1). In the future, it is crucial to assess this adjusted CHF diagnostic algorithm on

accuracy and costs. We therefore recommend to validate the customized CHF diagnostic algo-

rithm in new populations of residential elderly.

In the introduction, the paragraph on HF diagnostics, we emphasized that access to echocardi-

ography was limited. Since recently, echocardiography has become available for primary care

patients in The Netherlands, that barrier to accurately diagnose CHF seems to be removed. We

subsequently argue for the option of on-site echocardiography for those (residential) elderly

who are immobile.

To editors of the Multidisciplinary HF guideline (3), we give in consideration to adapt the algo-

rithm (3) to residential elderly. The adaptations concern non-acute HF exclusively. They include

the use of echocardiography for ruling in CHF, the use of NPs only for excluding CHF, and

referral for echocardiography also in primary care laboratories (www.Certe.nl).

In Chapter 3, we showed the prognostic values of NPs for one-year mortality of residential eld-

erly with an established diagnosis of CHF. We therefore advise to add natriuretic peptide testing

to “advance care planning” for those (residential) elderly facing a limited life expectancy. In Chapter 4, we demonstrated the relation between CHF and the needs for more ADL-help,

compared to residential elderly without CHF. This relation made us to advise care teams to rec-

ognise higher requirements for ADL-help, in residential elderly with CHF. In Chapter 5, we observed that residential elderly who performed exercise during four months,

changed neither their CHF nor their CV-metabolic indices. We have explained the absence of

improvements to the high age and subsequent loss of muscular endurance. In addition, the

193

training frequency and duration were too modest contributing to the absence of improvements

as well. Based on these findings, we recommend in future to study residential elderly with CHF

during and after a personalized exercise program of three times a week in the long-run. An ex-

ample of a personalized program is the option for the aged to select his type of training himself.

Such programs may be more effective compared to programs imposed (13). Besides, we em-

phasize to pay attention regularly to the motivation of residential elderly to exercise, since we

found low participation rates in our study.

As a consequence of our results partly processed in the diagnostic CHF algorithm specific for residential elderly (Chapter 6.2), we recommend the following:

To editors of continuing medical education (e.g. www.CME-scholing.nl), we suggest to record

our new insights into CHF in education programs for physicians elderly medicine (SO). The

insights include the found CHF prevalence of at least 23% in residential elderly. Those also

include that CHF can be better diagnosed according to the diagnostic CHF algorithm that we customized to residential elderly (Figure 1). Furthermore, that advance care for residential eld-

erly, can be improved by the determination of natriuretic peptides also. Importantly, neither CHF

symptoms and a marker nor CV-metabolic indices may not improve in residential elderly, after

exercise during a four months period. However, in the long-run, beneficial effects of exercise in

residential elderly may not be impossible, when using personalized conditions. This includes

that the resident chooses himself whether he starts walking, home-training, swimming or visits

the fitness club.

194

elderly medicine in primary care

HF outpatients’ clinic

Figure 1. Diagnostic algorithm for non-acute heart failure in residential elderly.

Residential elderly with complaints and symptoms suspected for non-acute heart failure (HF) in a long-term care facility or primary care medicine

Physical examination, ecg, lab. consider comorbidity or HF de-cline consider further steps

NT-proBNP<400 pg/mL or BNP <100 pg/mL, ecg undisturbed

HF is unlikely. Consider other causes

Patient not diagnosed with HF

Physical examination, ecg, NT-proBNP or BNP (Hb, eGFR, TSH, K+, glucose)

Patient diagnosed with HF

Echocardiography

HFrEF and or HFpEF is likely.

Cause analysis/ treatable HF cause.

Cause analysis /no treatable cause for HF.

Consultation with SO /GP Treatment

Referral back to SO/GP with treatment advice.

Further diagnostics by the SO/GP

Yess

No

Neg.

Pos.

195

BNP, B-type natriuretic peptide; NT-proBNP, nitrogen-terminal-proBNP; SO, specialist ouder-

engeneeskunde; GP, general practitioner. References

1 Leening MJ, Siregar S, Vaartjes I, Bots ML, Versteegh MI, van Geuns RJ, et al. Heart

disease in The Netherlands: a quantitative update. Neth Heart J 2014;22(1):3-10.

2 McMurray JJ, Adamopoulos S, Anker SD, Auricchio A, Bohm M, Dickstein K, et al.

ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure

2012: The Task Force for the Diagnosis and Treatment of Acute and Chronic Heart

Failure 2012 of the European Society of Cardiology. Developed in collaboration with the

Heart Failure Association (HFA) of the ESC. Eur Heart J 2012;33(14):1787-1847.

3 Voors AA, Walma EP, Twickler TB, Rutten FH, Hoes AW. [Multidisciplinary guideline

'Heart failure 2010']. Ned Tijdschr Geneeskd 2011;155:A2957.

4 Mant J, Al-Mohammad A, Swain S, Laramee P. Management of chronic heart failure in

adults: synopsis of the National Institute For Health and clinical excellence guideline.

Ann Intern Med 2011;155(4):252-259.

5 Bruins S, Fokkema MR, Römer JW, DeJongste MJ, van der Dijs FP, van Ouweland

JM, et al. High intraindividual variation of B-type natriuretic peptide (BNP) and amino-

terminal proBNP in patients with stable chronic heart failure. Clin Chem

2004;50(11):2052-8.

196

6 Grebe C, Brandenburg H. [Resident assessment instrument. Application options and

relevance for Germany]. Z Gerontol Geriatr 2015;48(2):105-13.

7 Goff DC Jr, Lloyd-Jones DM, Bennett G, Coady S, D'Agostino RB, Sr, Gibbons, et al.

2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the

American College of Cardiology/American Heart Association Task Force on Practice

Guidelines. J Am Coll Cardiol 2014 ;63(25 Pt B):2935-2959.

8 Weening-Dijksterhuis E, de Greef MH, Krijnen W, van der Schans CP. Group exercise

has little effect on ADL, physical fitness, and care dependency in frail institutionalised

elderly people: a randomized controlled trial. Thesis 2014; www.Hanze.nl.

9 Hancock C, Close H, Mason JM, Murphy JJ, Fuat A, Singh R, et al. High prevalence

of undetected heart failure in long-term care residents: findings from the Heart Failure

in Care Homes (HFinCH) study. Eur J Heart Fail 2012;15(2):158-65.

10 Bibbins-Domingo K, Gupta R, Na B, Wu AH, Schiller NB, Whooley MA. N-terminal

fragment of the prohormone brain-type natriuretic peptide (NT-proBNP), cardiovas

cular events, and mortality in patients with stable coronary heart disease. JAMA

2007;297:169–176.

11 Mason JM, Hancock HC, Close H, Murphy JJ, Fuat A, de Belder M, et al. Utility of

biomarkers in the differential diagnosis of heart failure in older people: findings from the

heart failure in care homes (HFinCH) diagnostic accuracy study. PLoS.One 2013;8(1);

e53560.

197

12 Oudejans I, Mosterd A, Bloemen JA, Valk MJ, van Velzen E, Wielders JP, et al. Clini-

cal evaluation of geriatric outpatients with suspected heart failure: value of symptoms,

signs, and additional tests. Eur J Heart Fail 2011;13(5):518-27.

13 Beckers PJ, Denollet J, Possemiers NM, Wuyts K, Vrints CJ, Conraads VM. Maintaining

physical fitness of patients with chronic heart failure: a randomized controlled trial. Eur J

Cardiovasc Prev Rehabil 2010;17(6):660-7.

198

199

Chapter 7

SAMENVATTING, DISCUSSIE, AANBEVELINGEN EN TOEKOMST-

PERSPECTIEVEN

200

Chapter 7.1

SAMENVATTING Een nadelig gevolg van de verbeterde overleving van cardiovasculaire ziektes (CVD) (1) is dat

de cardiovasculaire morbiditeit is toegenomen. Zoals in de introductie van dit proefschrift wordt

benadrukt, is het aantal mensen met hartfalen (HF) aanzienlijk toegenomen, terwijl tegelijkertijd

de levensverwachting in de Nederlandse populatie sterk is verbeterd gedurende de laatste eeuw

(www.cbs.nl; www.nationaalkompas.nl). Omdat HF zich veelal op hogere leeftijd manifesteert,

behoren patiënten met chronisch hartfalen (CHF) meestal tot de categorie ouderen.

Er bestaat echter een groot tekort aan kennis over HF bij toenemende leeftijd en bij kwetsbare

ouderen die afhankelijker worden van zorg. Ouderen die het meest afhankelijk zijn van zorg,

worden opgenomen in een verzorgings- of verpleeghuis. Deze ouderen worden in deze thesis

aangeduid als residentiële ouderen. Het gebrek aan onderzoeksgegevens en resultaten is voor-

namelijk ontstaan doordat residentiële ouderen geweerd worden uit studies vanwege hun hoge

leeftijd en multimorbiditeit, zoals CVD. Vanwege een gebrek aan gegevens zijn er geen richtlij-

nen beschikbaar om HF bij residentiële ouderen te diagnostiseren en te behandelen. Vanuit dit

gezichtspunt bestaat er discussie over de toepasbaarheid van bestaande HF-richtlijnen bij resi-

dentiële ouderen die ontwikkeld zijn voor CHF patiënten in het algemeen. Richtlijnen voor resi-

dentiële ouderen met CHF worden node gemist omdat de CHF diagnose lastig te stellen is: ou-

deren neigen ertoe om klachten toe te schrijven aan het ouder worden in plaats van CHF of

andere ziektes. Daardoor wordt de behandeling van CHF vaak ten onrechte of helemaal niet

uitgevoerd. Als gevolg van deze ongewenste situatie gaat er kwaliteit van leven verloren en

nemen de kosten toe. Deze bevindingen vormen de aanleiding om CHF bij residentiële ouderen

aan nader onderzoek te onderwerpen. In Hoofdstuk 2.1 wordt de schaalgrootte van CHF verkend door de prevalentie van CHF onder

residentiële ouderen te bepalen. Wanneer de algemene HF richtlijnen (2-4) worden toegepast,

blijkt uit recente literatuur de HF diagnostiek minder accuraat te zijn ten gevolge van de laag-

voorspellende waarden van klachten en symptomen, medische voorgeschiedenis en elektrocar-

201

diogram (ECG). Uit de medische gegevens van residentiële ouderen van één verpleeghuis in

Groningen werden de klachten, de symptomen en de medische voorgeschiedenis verzameld en

werd het aantal ouderen met CHF geïnventariseerd. Vervolgens werden alle ouderen onderwor-

pen aan een lichamelijk onderzoek en maakten we een ECG van elk van hen. Daarna bepaal-

den we hoe accuraat de eerdere CHF diagnostiek [zonder gebruik van natriuretische peptiden

(NP’s)] en de huidige CHF diagnostiek (met NP’s) waren, omdat in tegenstelling tot de brede

erkenning van NP’s als screening test van CHF, deze NP’s nog steeds weinig gebruikt worden

als biomarker bij residentiële ouderen. Het beperkte gebruik bij ouderen wordt veroorzaakt

doordat de afkapwaarden van NP’s in deze groep onvoldoende gevalideerd zijn. Tot slot werden

de ECG’s van alle residentiële ouderen beoordeeld.

Een cardiologenpanel stelde de diagnose CHF vast of sloot deze uit door, achtereenvolgens de

uitslagen van de NP’s en resultaten van echocardiografie te beoordelen. Echocardiografisch

onderzoek bleek mogelijk in 98% van de ouderen. Het resultaat van de huidige studie was dat

CHF werd bevestigd in 24/103 (23%) van de residentiële ouderen. Voorafgaand aan de stu-

die, werden 22 ouderen geïdentificeerd met CHF. Bij deze 22 ouderen met CHF, vastgesteld

voorafgaand aan de studie, werd CHF bevestigd bij 9 ouderen en bij 13 ouderen werd de dia-

gnose CHF verworpen. De diagnostische accuratesse van NT-proBNP bij 450 pg/mL bestond

uit een sensitiviteit van 0,71, specificiteit van 0,67, positief voorspellende waarde (PPV) van

0,42 en negatief voorspellende waarde (NPV) van 0,91. De diagnostische accuratesse van

BNP bij 100 pg/mL toonde een 0,71 sensitiviteit, 0,70 specificiteit, 0,41 PPV en 0,88 NPV.

In het kort waren de meest opvallende resultaten van de huidige studie dat meer dan de helft

van de CHF diagnoses werd gemist (namelijk 15/24) of onjuist werd gesteld (namelijk 13/22).

Daarom kan worden geconcludeerd dat het identificeren van CHF inaccuraat is. Maar uitgaande

van de hoge NPV, lijkt het gebruik van NP’s als additionele diagnostische test veelbelovend,

zelfs bij residentiële ouderen. Het laatste moet echter nog nader worden geëvalueerd. De CHF-

prevalentie onder residentiële ouderen werd vastgesteld op 23%.

202

In Hoofdstuk 2.2 is onderzocht of de uitkomsten van hoofdstuk 2.1 consistent zijn met de CHF-

prevalentie in Arubaanse verpleeghuizen. Er werd voor Arubaanse verpleeghuizen gekozen

vanwege onze interesse in een vergelijkbare onderzoekspopulatie met die in Nederland, die

echter woont in een andere omgeving en er een andere levensstijl op na houdt. Er werd gebruik

gemaakt van hetzelfde studieontwerp zoals gehanteerd in hoofdstuk 2.1 om de prevalentie van

CHF bij Arubaanse residentiële ouderen te berekenen.

Een ander doel was het maken van een inschatting of de toepassing van NP’s voor het opspo-

ren van CHF bij residentiële ouderen in een ander deel van het Nederlands Koninkrijk verbeterd

kan worden. Het belangrijkste verschil met de studie uit hoofdstuk 2.1 was dat er geen echocar-

diogram van alle residentiële ouderen kon worden verkregen vanwege logistieke belemmerin-

gen. Daarom kon niet worden voldaan aan de diagnostische standaard voor CHF (2-4). Als

resultaat werden 51 van de 235 Arubaanse ouderen met een gemiddelde leeftijd van 78±8 jaar

in de studie opgenomen. Afgaand op de medische gegevens, werd CHF vastgesteld bij 7/51

ouderen. Echter twee van de zeven ouderen hadden geen CHF. Verder werd CHF bevestigd bij

16 van de 51 (31%) ouderen. Van de 16 ouderen met CHF, waren 11 niet eerder gediagnosti-

seerd met CHF en van 5/16 ouderen bij wie CHF eerder was vastgesteld, werd de diagnose

bevestigd.

Kort gezegd, de algemene richtlijnen voor HF worden laagfrequent toegepast bij Arubaanse

residentiële ouderen. Vergeleken met residentiële ouderen uit Groningen, is de prevalentie van

CHF waarschijnlijk minstens zo hoog onder de Arubaanse residentiële ouderen. Als conclusie

blijkt de herkenning van CHF ernstig onderschat te worden onder Arubaanse residentiële oude-

ren. De opsporing van CHF kan worden verbeterd door een geschikte richtlijn te implementeren,

inclusief de bepaling van BNP.

Hoofdstuk 2.3 behandelt het probleem van de grote intra-individuele variaties [individuele va-

riatie coëfficiënt (CVi)] van NT-proBNP in plasma, aangezien grote CVis de toepasbaarheid van

NT-proBNP beperken voor o.a. de NT-proBNP-geleide therapieoptimalisatie bij individuele pati-

203

enten met CHF (5). Op zoek naar meer betrouwbare CVis werden de concentraties van urine-

NT-proBNP (NT-proBNPu) vergeleken met die van plasma-NT-proBNP (NT-proBNPp) bij pati-

enten met CHF woonachtig op Curaçao. Op één enkele dag ("op-1-dag") werden urine en

bloed monsters afgenomen: zes bloedmonsters elke 2 uur en de spontaan uitgeplaste urines

gedurende 24 uur. Op vijf opeenvolgende dagen ("dag-tot-dag") werden verzameld: vijf

bloedmonsters, vijf geforceerd uitgeplaste urine monsters en vijf volledige 24-uurs urines. Op

dezelfde dag van zes opeenvolgende weken ("week-tot-week") werden verzameld: één

bloedmonster, één geforceerd uitgeplast urinemonster en een volledige 24-uurs urine.

Uit deze bloed- en urinemonsters werden de totale CVs (CVts), de CVis en de coëfficiënten van

het percentage verandering van de referentiewaarden (RCV) berekend. In de studie werden 25

patiënten met CHF opgenomen met een gemiddelde leeftijd van 61 jaar (spreiding 36-80),

60% was man en de gemiddelde linker ventrikel ejectiefractie (LVEF) was 36±15%. De media-

ne CVis voor NT-proBNPp waren 9% (“op-1-dag”), 18% (“dag-tot-dag”) en 30% (“week-tot-

week”).

De CVis van NT-proBNPu waren respectievelijk 34%, 21% en 28%. De RCV van NT-proBNPu

was hoger dan of gelijk aan de RCV van NT-proBNPp, zelfs na correctie voor creatinine en de

periode (in uren) waarin de urine zich in de blaas had verzameld. De gegevens van de “dag-

tot-dag” en “week-tot-week” monsters doen vermoeden dat (immunoreactive) concentraties van

NT-proBNPu in de urine niet ondubbelzinnig correleren met de (immunoreactive) NT-proBNP concentraties in plasma, tot aan de NT-proBNPp drempel van ongeveer 310 pg/mL (Figuur 1).

Daarnaast werd opgemerkt dat boven deze drempel de hogere NT-proBNP concentraties, zoals

weergegeven in Figuur 1, niet verklaard kunnen worden door een verhoging van de bloeddruk

gedurende de dag (resultaten niet weergegeven).

Veronderstelt kan worden dat filtratie van natriuretische peptiden kan worden beschouwd als

dominante factor voorbij de drempelwaarde en dat bij lagere NT-proBNP-waarden, filtratie en

terugresorptie meer dynamisch interacteren. Geconcludeerd kan worden dat het bepalen van

NT-proBNP in urine geen voordelen biedt bij patiënten met stabiel chronisch hartfalen.

204

Figuur 1. De relatie tussen NT-proBNP in plasma and NT-proBNP in geforceerd uitgeplaste

urines verzameld volgens de “dag-tot-dag” en “week-tot-week” protocollen.

De data zijn afkomstig van 22 patiënten en in totaal 182 corresponderende plasma en

urine monsters. De “dag-tot-dag”- en week-tot-week”-bloedmonsters werden gebruikt. Voor de

urine werden de geforceerd uitgeplaste urinemonsters gebruikt.

In hoofdstuk 3 werd de prognostische waarde onderzocht van NP’s op de sterfte in één jaar bij

hetzelfde cohort residentiële ouderen dat werd bestudeerd in hoofdstuk 2.1. De klinische rele-

vantie hiervan is dat als NP’s gerelateerd worden aan de prognose van residentiële ouderen

met CHF, deze biomarkers kunnen worden benut voor “advance care planning”. Deze vorm van

205

zorgplanning is gerelateerd aan de kwaliteit van leven en wordt belangrijk wanneer de levens-

verwachting beperkt is (www.Verenso.nl). Om de sterfte in één jaar te onderzoeken, werden

gedurende één jaar de overleving en het overlijden van residentiële ouderen bijgehouden. Om-

dat tien residentiële ouderen niet konden worden vervolgd doordat zij naar onbekende bestem-

mingen verhuisden onderzochten we de resterende 93 residentiële ouderen.

Achttien uit 93 bewoners (gemiddelde leeftijd 81 ± 3 jaar, 66% vrouw) stierven binnen één

jaar(niet-overlevenden).

Een wederzijds aangepaste Cox proportionele risico regressieanalyse werd uitgevoerd. Er von-

den correcties plaats voor zes vooraf gedefinieerde chronische ziekten, immobilisatie, leeftijd,

geslacht, NT-proBNP en BNP.

De gevonden resultaten lieten zien dat beide NP’s (NT-proBNP en BNP) de één-jaars sterfte

significant voorspelden (respectievelijk, HR 1,02 en p = 0,001 en HR 1,16 en p = 0,003). Bo-

vendien nam het risico op overlijden in één jaar exponentieel toe parallel aan de stijging van

BNP en NT-proBNP in de hogere concentraties. Bijvoorbeeld, een toename van NT-proBNP tot

2.000 pg/mL was geassocieerd met een 60% hoger risico op overlijden binnen één jaar.

Concluderend zijn zowel BNP als NT-proBNP onafhankelijke voorspellers van de één-jaars

mortaliteit onder residentiële ouderen met CHF. Daarbij neemt het sterfterisico toe bij NP con-

centraties die verhoogd zijn tot ver boven de diagnostische afkapwaarden van de richtlijnen (2-

4). In hoofdstuk 4 is de relatie bestudeerd tussen CHF en de steun die nodig is voor residentiële

ouderen met CHF tijdens hun activiteiten van het dagelijks leven (ADL). Daarvoor werd hetzelf-

de cohort residentiële ouderen onderzocht als in hoofdstuk 2.1. Het doel van deze studievraag

was aandacht te trekken van zorgteams voor de behoefte aan steun onder residentiële ouderen

met CHF.

Voor deze studie werd dezelfde groep residentiële ouderen onderzocht als in hoofdstuk 2.1. Om

gegevens over ADL te verzamelen, werden de “Minimal Data Set (MDS)” items van het “Resi-

dent-Assistant Instrument” gebruikt. De MDS bestaat uit systematische observaties van de af-

206

hankelijkheid en van de geboden ondersteuning en heeft als doel om de afhankelijkheid en de

hulp te kwantificeren (6). Het MDS instrument is in feite een softwareprogramma voor de invoer

van observaties van individuele residentiële ouderen tijdens vast omschreven ADL. Een voor-

beeld van een dergelijke activiteit is lopen. De verzorgende observeert en scoort (kwantificeert)

of het individu in staat is om bepaalde afstanden onafhankelijk te lopen en of hij hulp nodig

heeft bij het lopen. Van de 103 residentiële ouderen met en zonder CHF, werden de scores van

ADL-afhankelijkheid en ADL-hulp vergeleken en gecorrigeerd voor de verschillen tussen beide

groepen door middel van multivariabele logistische regressieanalyse. CHF was geassocieerd

met ADL-hulp (OR 4,68 en p=0,015) echter, CHF was niet gerelateerd aan de ADL-

afhankelijkheid.

Geconcludeerd kan worden dat CHF bij residentiële ouderen geassocieerd is met de behoefte

aan meer ADL-hulp. In dit verband is CHF van belang voor zowel de residentiële ouderen (ver-

lies van autonomie en levenssatisfactie) als ook voor de organisatie (personele inzet).

In hoofdstuk 5 wordt een fysiek oefenprogramma onderzocht onder residentiële ouderen met

CHF. De reden voor deze studie is inzicht te verkrijgen in de gevonden discrepantie tussen de

sedentaire levensstijl van residentiële ouderen en de gedane aanbevelingen voor fysieke activi-

teit van de HF-richtlijnen (2-4). Ook stimuleren familieleden hun ouders in verzorgingshuizen

om lichamelijk actief te worden. Echter, om diverse redenen blijven residentiële ouderen liever

sedentair. Bovendien, uit angst voor vallen, geven ook verzorgers er de voorkeur aan dat resi-

dentiële ouderen zoveel mogelijk sedimentair zijn.

Een sedentaire levensstijl is gerelateerd aan een verhoogd risico op toename van atherosclero-

se en CV-metabole ziekten, zoals CHF en het metabool syndroom (7).Onderzocht werd of li-

chamelijke training een gunstige invloed had op symptomen van CHF, NT-proBNP, het meta-

bool syndroom en CV-metabole indices bij residentiële ouderen. Een pilot, enkelvoudig-

207

geblindeerde, gerandomiseerde en gecontroleerde haalbaarheidstudie werd uitgevoerd in vier

verzorgingstehuizen. In samenwerking met andere onderzoekers (9) werd bestudeerd of een

gestandaardiseerd trainingsprogramma invloed heeft op symptomen van CHF, NT-proBNP en

het metabool syndroom. De andere onderzoekers (8) onderzochten of hetzelfde gestandaardi-

seerde oefenprogramma de, door hen gekozen, meer functionele eindpunten beïnvloedt in ver-

gelijkbare groepen. De trainingsinterventie omvatte een begeleide training van matig-tot-hoge

intensiteit van één uur, twee keer per week gedurende 16 opeenvolgende weken. De training

bestond uit progressieve weerstandstraining van bovenste en onderste extremiteiten en van de

romp, uit statische en dynamische evenwichtstraining en uit functionele training. De interventie

van de controle groep bestond uit een begeleid, niet-fysiek, sociaal programma van dezelfde

frequentie, duur, lengte en locatie. De studie eindpunten bestonden uit symptomen van CHF

(vermoeidheid, kortademigheid), en NT-proBNP, uit lichaamsmetingen en uit laboratoriumresul-

taten, die als combinatie bekend zijn als parameters van het "metabool syndroom". Het syn-

droom is gedefinieerd als aanwezigheid van drie uit vijf lichaamsmetingen en laboratoriumresul-

taten. Deze metingen omvatten de taille, de bloeddruk, de nuchtere glucose, de triglyceriden en

het high-density lipoproteïne (HDL)-cholesterol. Glucose wordt vervangen door geglyceerd he-

moglobine (HbA1c), aangezien een nuchtere bloedafname logistiek niet haalbaar bleek te zijn.

Als resultaat, kwamen van de 434 residentiële ouderen, 163 in aanmerking voor deelname.

Helaas, waren slechts 52 (12%) bereid om deel te nemen aan de studie. Van deze 52, vielen

er zes af tijdens de studie vanwege redenen die niet waren gerelateerd aan de training. De res-

terende 46 deelnemers waren 85±6 jaar, 65% was vrouw en allen werden gerandomiseerd

voor trainingsgroep (n=24) of controle groep (n=22). De “intention-to-treat analyse” toonde

geen significante verandering van welke variabele dan ook bij de trainingsgroep vergeleken met

de controles. De groep die geanalyseerd werd “as-treated” voltooide de helft of meer van de 32

trainingssessies. De vrouwen van de laatstgenoemde groep bleken een afname van taille om-

vang te hebben [-8 (spreiding -15--2) cm]. De controles toonden een significante verlaging

van triglyceriden [-8 (-17-0) mg/dL] en een stijging van NT-proBNP [67 (7-126) pg/mL].

208

Het aantal bijgewoonde sessies was omgekeerd gerelateerd aan de niet-nuchtere triglyceriden

bij de trainingsgroep (p = 0,024). Wij concludeerden dat fysieke training onder begeleiding,

uitgevoerd door residentiële ouderen gedurende een korte periode, geen effect had op sympto-

men en een biomarker van CHF, noch op CV-metabole indexen. Een mogelijke verklaring voor

de neutrale resultaten is de hoge leeftijd van de residentiële ouderen (gemiddeld 85 jaar), die

gepaard gaat met een verlies van spieruithoudingsvermogen en door een trainingsaanbod dat

onvoldoende is afgestemd op het individu.

209

Kort samengevat zijn de resultaten van deze thesis: de prevalenties van CHF bij residentiële ouderen zijn 23% in Groningen hetgeen strookt

met de literatuur (9), en >30% op Aruba (hoofdstukken 2.1 en 2.2).

bij Nederlandse residentiële ouderen komt CHF vaak voor. Het gebruik van natriureti-

sche peptiden, als aanvullende test voor CHF, moet verder bestudeerd worden. Bij resi-

dentiële ouderen in Groningen, kunnen onjuiste diagnoses van CHF worden terugge-

bracht door diagnostisch natriuretische peptiden toe te passen (hoofdstuk 2.1 en 2.2).

bepaling van NT-proBNP concentraties in de urine van patiënten met CHF heeft geen

voordelen vergeleken met plasma-NT-proBNP concentraties. Deze bevinding komt voort

uit de grote intra-individuele variabiliteit van urine-NT-proBNP die vergelijkbaar is met

die in plasma-NT-proBNP (hoofdstuk 2.3).

bij residentiële ouderen, zijn bepalingen van natriuretische peptiden van prognostische

betekenis (HR van NT-proBNP 1,02 en BNP 1,16 (HR van BNP is 2.2 (10)), waarbij

het resultaat van BNP overeenkomt met dat in de literatuur (10) (hoofdstuk 3).

CHF bij residentiële ouderen is gerelateerd aan de behoefte aan meer ADL-hulp verge-

leken met ouderen zonder CHF (hoofdstuk 4).

een korte periode van training uitgevoerd door residentiële ouderen heeft geen effect op

CHF noch op CV indices (zoals b.v. het metabool syndroom). Er zijn geen studies voor-

handen om dit te kunnen vergelijken (hoofdstuk 5).

210

211

Hoofdstuk 7.2

DISCUSSIE

212

We toetsten onze hypothese (Introductie) die stelt dat de diagnostiek van niet-acuut HF, zoals

die aanbevolen wordt in de algemene HF richtlijnen (2-4), niet toegepast kan worden op resi-

dentiële ouderen. Daartoe, gebruikten we de afkapwaarden van de natriuretische peptiden

(NP’s) zoals aanbevolen door de HF richtlijnen (2-4) in het cohort Groninger residentiële oude-

ren (hoofdstuk 2.1). Het is van belang te vermelden dat zowel de ESC als de Multidisciplinaire

HF richtlijn (2;3) dezelfde afkapwaarden adviseren (NT-proBNP 125 and BNP 35 pg/mL) en

dat de laatste waarden lager zijn vergeleken met die van de NICE HF richtlijn (namelijk NT-

proBNP 400 and BNP 100 pg/mL) (4). Gebruikmakend van deze afkapwaardes uit de richtlij-

nen, berekenden we de voorspellende waarden voor de diagnose CHF, in residentiële ouderen van één Gronings verpleeghuis (Tabel 1). Tevens bepaalden we de voorspellende waardes ook

bij hogere afkapwaarden zoals bij NT-pro BNP 900 pg/mL en BNP 100 pg/mL. Jaren na onze

studie publiceerde Mason et al. (11) een artikel over voorspellende waarden van NP’s voor CHF

bij Engelse residentiële ouderen. De afkapwaarden van de studie van Mason et al. verschilden

weinig van die uit onze studie. De laatstgenoemde studie groep valideerde haar resultaten met de NT-proBNP en BNP afkapwaarden volgens de NICE guideline (4)(Tabel 1).

213

Tabel 1. Voorspellende waarden van NT-proBNP en BNP voor de diagnose hartfalen

in twee populaties residentiële ouderen: de Groninger (NL) (hoofdstuk 2.1) en de Britse

populatie (11).

afkapwaarde

pg/mL sensitiviteit specificiteit PPV NPV

Groninger RE (NL)

NT-proBNP1 125 1.00 0.28 0.29 1.00

NT-proBNP 450 0.71 0.67 0.42 0.91

NT-proBNP 900 0.67 0.85 0.56 0.89

BNP1 35 0.88 0.39 0.30 0.91

BNP 50 0.88 0.63 0.31 0.89

BNP2 100 0.71 0.70 0.41 0.88

British RE (11)

NT-proBNP2 400 0.56 0.69 0.35 0.84

NT-proBNP 760 0.62 0.75 0.42 0.87

BNP 115 0.67 0.68 0.38 0.88

Residentiële ouderen (residential elderly), RE; Nederland, NL; PPV, positief voorspellende

waarde; NPV, negatief voorspellende waarde; NT-proBNP, N-terminal amino pro B-type natri-

uretic peptide; B-type natriuretic peptide.

1 ESC HF richtlijn (2) en multidisciplinaire HF richtlijn (3)

2 NICE HF richtlijn (4).

De groep van Mason (11) en wij onderzochten niet uitsluitend NP’s maar ook combinaties van

medische voorgeschiedenis, klachten en symptomen, ECG, en NP’s met het doel om CHF aan

te tonen. Wij beiden vonden dat geen van de mogelijke combinaties van parameters, de voor-

214

spellende waarden verbeterde vergeleken met die berekend met behulp van NP’s alleen. De

resultaten van Mason’s studiegroep, vooral wat betreft de PPV, komen goed overeen met onze

bevindingen.

Beide onderzoeksgroepen rapporteren ook dat toepassing van de HF richtlijnen (2-4) op Ne-

derlandse en Engelse residentiële ouderen, daarin resulteert dat meer dan de helft van de CHF

diagnoses gemist worden (te veel fout negatieven). Ook verlaging van de afkapwaarden (2;3)

ten opzichte van de NICE richtlijn (4) levert geen enkele verbetering op. Samen met de Mason

groep, komen we tot de conclusie dat de gevonden voorspellende waarden ongeschikt zijn om

CHF aan te tonen maar alleen geschikt zijn om CHF uit te sluiten.

Daarom besloten we om in dit proefschrift het algoritme van de Multidisciplinaire HF richtlijn (3)

aan te passen voor residentiële ouderen. Deze aanpassingen houden in dat NP’s alleen ge-

bruikt worden om CHF uit te sluiten, dat echocardiografie noodzakelijk is om CHF aan te tonen,

dat de afkapwaarden uit de NICE richtlijn (4) gebruikt worden, en dat echocardiografie ook door

laboratoria voor de eerstelijnsgezondheidszorg uitgevoerd worden (www.Certe.nl).

Toch toonde Oudejans et al. (12) aan bij geriatrische poliklinische patiënten van wie >40%

CHF had, dat de bepaling van NP’s in combinatie met andere diagnostische tests wel degelijk

accuraat is om zowel CHF aan te tonen als om CHF uit te sluiten. Waarschijnlijk was de PPV

voor CHF bij ons onderzoekscohort lager omdat de vooraf kans op CHF lager was door de la-

gere CHF prevalentie 23%, vergeleken met de vooraf kans bij de poliklinische patiënten (CHF

prevalentie >40%). In tegenstelling tot de resultaten van de Oudejans groep (12) maar in

overeenstemming met de resultaten van Mason et al. en onze studie groep [Chapter 1.2 en

(11)], hebben we onze aanbevelingen voor residentiële ouderen bijgesteld. Onze hypothese is

deels verworpen omdat de algemene HF richtlijnen ongeschikt zijn om CHF aan te tonen bij

residentiële ouderen. Daarnaast werd onze hypothese deels aanvaard omdat CHF wel uitge-

sloten kan worden met behulp van NP’s volgens de HF richtlijnen.

215

Uitgaande van onze pas verworven inzichten wordt geadviseerd om een echocardiogram te la-

ten maken om CHF aan te tonen. Theoretisch worden NP’s aanbevolen om CHF uit te sluiten

alhoewel we ons realiseren dat uitsluiten van CHF niet volstaat in de dagelijkse praktijk. Deze

adviezen hebben we opgenomen in een diagnostisch algoritme voor CHF (d.w.z. niet-acuut HF) voor residentiële ouderen (Figuur 1).

216

217

Hoofdstuk 7.3

AANBEVELINGEN EN TOEKOMST PERSPECTIEVEN

218

In de discussie stelden we vast dat algemene HF richtlijnen ongeschikt zijn om CHF te diagnos-

tiseren bij residentiële ouderen. Ook is bij de individuele residentiële oudere, een echocardio-

gram essentieel om de diagnose CHF te stellen. Daartoe hebben we het Nederlandse diagnos-tische algoritme voor alle CHF patiënten (3), aangepast aan die specifieke groep (Figuur 1). In

de toekomst is het van essentieel belang om dit aangepaste diagnostische CHF algoritme te

toetsen op accuratesse en kosten. Daarom raden we aan om het aangepaste diagnostische

CHF algoritme te valideren in nieuwe populaties residentiële ouderen.

In de introductie, het hoofdstuk over HF diagnostiek benadrukten we de beperkte mogelijkheden

om een echocardiogram aan te vragen. Met de recente beschikbaarheid van echocardiografie

voor eerstelijns patiënten in Nederland, lijkt die barrière, om CHF accuraat te diagnostiseren

geslecht te zijn. Aansluitend pleiten we ervoor om een echocardiogram thuis of in de instelling

te kunnen maken voor die residentiële oudere die immobiel zijn.

We geven redacteuren van de Multidisciplinaire HF richtlijn (3) in overweging om de richtlijn

aan te passen aan residentiële ouderen. De aanpassingen betreffen alleen niet-acuut HF. De

wijzigingen bestaan daar uit dat CHF alleen kan worden aangetoond met een echocardiogram,

dat CHF uitgesloten kan worden met NP’s en dat ook verwezen kan worden naar eerstelijns

laboratoria voor echocardiografie (www.Certe.nl).

In Hoofdstuk 3 presenteerden we de prognostische betekenis van NP’s voor één-jaars overle-

ving van residentiële ouderen met bewezen CHF. Op grond daarvan wordt geadviseerd om een

NP bepaling toe te voegen aan de palliatieve zorgplanning voor (residentiële) ouderen met een

beperkte levensverwachting. We laten in Hoofdstuk 4 de relatie zien tussen CHF en de behoefte aan meer ADL-hulp verge-

leken met residentiële ouderen zonder CHF. Vanwege deze relatie raden we zorgteams aan om

oog te hebben voor toegenomen behoeften aan ADL-hulp van residentiële ouderen met CHF.

219

Hoofdstuk 5 leert ons dat bij residentiële ouderen die trainden gedurende vier maanden, geen

verandering gezien werd van CHF noch van CV-metabole indices. Het uitblijven van verbetering

werd verklaard door de hoge leeftijd en het bijkomend verlies van spieruithoudingsvermogen.

Bovendien was de frequentie van trainen te laag en de duur te kort hetgeen ook bijdroeg aan

het uitblijven van effect. Op basis hiervan wordt aanbevolen om, in de toekomst, residentiële

ouderen te bestuderen tijdens en na een individueel trainingsprogramma van drie maal per

week gedurende een langere tijd. Een voorbeeld van een individueel programma vormt de mo-

gelijkheid voor de oudere om zelf zijn trainingsvorm te kiezen. Dergelijke programma’s lijken

effectiever te zijn dan vaststaande programma’s. Bovendien wordt sterk aangeraden om regel-

matig aandacht te besteden aan de motivatie van residentiële ouderen om te trainen omdat we

te maken hadden met een klein aantal deelnemers in onze studie (Hoofdstuk 5).

Uitgaande van onze resultaten die deels verwerkt zijn in het diagnostisch CHF algoritme speci-

fiek voor residentiële ouderen (chapter 6.2), doen we de volgende aanbevelingen:

Aan redacteuren van medische bijscholing (bv CME-scholing.nl), geven we de suggestie om

onze nieuwe inzichten op het gebied van CHF te verwerken in bijscholingen voor specialisten

ouderengeneeskunde (SO). Het gaat om de gevonden CHF prevalentie van tenminste 23% bij

residentiële ouderen. Bovendien dat CHF bij voorkeur gediagnostiseerd wordt volgens het dia-gnostische CHF algoritme dat we aangepast hebben aan residentiële ouderen (Figuur 1). Ver-

der dat palliatieve zorg voor de residentiële oudere beter gepland kan worden door gebruik te

maken van een natriuretische peptide bepaling. Ook van belang is dat noch CHF symptomen

en een CHF marker, noch CV-metabole indices bij residentiële ouderen verbeteren na een vier

maanden durende training. Daarentegen is het niet onmogelijk dat fysieke training voordelen

oplevert wanneer de training langere periode plaatsvindt en op individuele leest geschoeid is.

Hieronder wordt verstaan dat de oudere zelf kiest of hij gaat lopen, fietsen, zwemmen of de

fitnessclub bezoekt.

220

ouderengeneeskunde 1e lijn

HF polikliniek

Figuur 1. Diagnostisch algoritme voor niet-acuut hart falen bij residentiële ouderen. NT-proBNP, B-type natriuretic peptide; NT-proBNP, amino-terminaal-pro BNP; SO, specialist ouderengeneeskunde; HFrEF, hartfalen met afgenomen ejectie fractie; HFpEF, hartfalen met behouden ejectie fractie.

Residentiële ouderen met klachten en symptomen verdacht voor niet-acuut hartfa-len (HF) in een zorginstelling of eerstelijnsgezondheidszorg

Lichamelijk onderzoek, ecg, lab. Overweeg comorbiditeit, toename HF en verwijzing cardioloog

NT-proBNP<400 pg/mL of BNP<100 pg/mL, afwijkend ECG

HF is onwaarschijnlijk, overweeg andere oorzaken van klachten

Patient bij wie HF niet is vastge-steld

Lichamelijk onderzoek, ecg, NT-proBNP of BNP (Hb, eGFR, TSH, K+, glucose)

Patient met de diagnose HF

echocardiografie

HFrEF / HFpEF is waarschijnlijk

Onderzoek naar (behandelbare) oorzaak HF

Analyse van oorzaak/ geen behan-delbare oorzaak van HF

Overleg met SO/HA over gewenste behandeling. Behandeling

Verwijzing terug naar SO/HA met behandeladvies

Behandeling van HF door SO of huisarts (HA)

ja

Neeno Neg.

Pos.

221

DANKWOORD

Jaren geleden stond ik bij een verpleeghuisbewoner verdacht van hartfalen. Bij het raadplegen

van de toen vigerende richtlijnen voor hartfalen, dacht ik direct dat de diagnostiek van hartfalen

beter moest kunnen. Nu schrijf ik het dankwoord van mijn proefschrift waarin ik aantoon dat de

kwaliteit van leven van verpleeghuisbewoners met hartfalen inderdaad verbetert als je de dia-

gnose correct stelt. Veel is daaraan vooraf gegaan. Het is fantastisch dat dit gelukt is. De vele

mensen die hieraan hebben bijgedragen wil ik graag bedanken.

Mijn eerste dank gaat uit naar de verpleeg- en verzorgingshuisbewoners, hun mantelzorgers en

de patiënten. Zonder hen waren de onderzoeken niet mogelijk geweest.

Prof. dr. F.A.J. Muskiet, beste Frits, van het begin af aan was je enthousiast om (NT-pro)BNP

te gaan bepalen bij verpleeghuisbewoners in het kader van hartfalen diagnostiek. Na elke on-

derzoeksbespreking, gaf je “college” niet alleen over hartfalen, (patho)fysiologie, veroudering

en insuline resistentie, maar ook over laag-gradige ontsteking o.a. in de hersenen, vitamine D

en life-style. Ik heb gretig naar je geluisterd. Van jou heb ik geleerd op het scherpst van de

snede te discussiëren.

Prof. J.L. Hillege, beste Hans, je vroeg altijd eerst wat ik wilde bereiken met het onderzoek.

Daarna kwam je een met een praktisch voorstel voor berekeningen. Je hebt bijgedragen aan de

groei van mijn statistisch inzicht.

Dr. M.J.L. de Jongste, beste Mike, je hebt me altijd positief gesteund bij het verwezenlijken van

dit proefschrift. Bij het maken van onderzoeksdesigns, zei je dat je je tijd ver vooruit was. Dat

bleek ok zo te zijn. Ons onderzoeksdesign werd later op grote schaal toegepast. Ook heb je mij

goed begeleid bij het raadplegen van anderen op het juiste moment. Ik heb heel veel van je

geleerd. Elk overleg gaf mij hernieuwde energie en richting. Daarvoor wil ik je hartelijk bedan-

ken. Dr. I.C.C. van der Horst, beste Iwan, dank voor je opbouwende commentaar. Prof. dr. R.A.

de Boer, beste Rudolf dank voor de prettige samenwerking.

222

De leden van de beoordelingscommissie wil ik hartelijk bedanken voor de tijd die ze hebben

genomen om mijn proefschrift te beoordelen: prof. de. W.P. Achterberg (beste Wilco) en prof.

dr. A.W. Hoes. Prof. dr. A.A. Voors, beste Adriaan, fijn dat je de tijd hebt genomen om mijn

proefschrift te beoordelen en dat je de tijd gaat nemen om te opponeren.

In het verpleeghuis “het Zonnehuis-Noord”, Zuidhorn is mijn onderzoek begonnen. Samen met

jullie, Mike en Frits, hebben we ons voorstel voor hartfalen diagnostiek met behulp van (NT)-

proBNP geïntroduceerd en konden we van start. Drs. J. Koster, beste Johan dank voor jouw

inspanning om goede (98%) echografische afbeeldingen te maken. Dank aan de laboratorium

medewerkers van het Zonnehuis, jullie waren altijd bereid om bloed af te nemen en ecgs te

maken, en dank aan de administratieve medewerkers voor jullie hulp bij het regelen van afspra-

ken. Heel prettig om met je samen te werken, Judith, in de (voormalige) Westerburcht. Ik be-

waar goede herinneringen aan de tweejaarlijkse onderzoeksbesprekingen met de Zonnehuis-

groep in Soesterberg. Dank Zonnehuisgroep voor de financiële steun.

Dankzij de flexibiliteit van jullie, dr. E. Weening en dr. M.H.G. de Greef, konden we aanhaken

aan het bewegingsonderzoek in vier verzorgingshuizen in Groningen. Ik heb de samenwerking

met jou Betsie, en de gesprekken met jou, Mathieu als heel inspirerend ervaren. Het manage-

ment team van De Hoven, hebben het ons financieel mogelijk gemaakt om ecgs en NT-proBNP

bepalingen te laten doen. Mijn dank gaat uit naar de interesse voor ons onderzoek die de We-

tenschappelijke Raad van De Hoven, Prof. dr. D.L. Knook en Meindert Bolt hebben getoond.

Dank voor de leuke samenwerking met jullie, de bewegingsmedewerkers uit Appingedam.

Van het Universitair Medisch Centrum Groningen (UMCG), wil ik bedanken Eduard Heine, voor

het bewaren van onze bloedmonsters op het laboratorium. Egbert Knol, je hebt statistische be-

werkingen van de scorelijsten goed uitgelegd, bedankt. Alma Guikema, dank je voor je hulp bij

het printen en verzenden van de boekjes. Eline, ik heb het als verrijking ervaren om samen met

je te schrijven aan het onderzoek dat je op Aruba hebt uitgevoerd. Wat hebben we gestaag ge-

223

werkt aan ons stuk, Anneliene. Ik heb respect voor je nauwkeurigheid van data verwerking.

Dank je voor onze prettige samenwerking op afstand.

Beste Corina, Jos, Meindert, Evert, Luch, Hilde, Nazir, Judith en Victoria, wat zijn onze toets-

groep bijeenkomsten van belang geweest. Daarin konden we veilig moeilijkheden en mogelijk-

heden bespreken in de ouderengeneeskunde. Hartelijk dank daarvoor.

Christien, we hebben niet alleen gezocht naar goeie Engelse termen. Ik ervaar het als heel

prettig om levensvragen met je te bespreken en daarna samen te lachen. Zo’n vriendschap

geeft kracht. Ik ben zo blij dat je een van mijn paranimfen bent. Daphne, dank voor je Engelse

woorden! Ria, dank voor de vele gesprekken over hoe onze doelen te bereiken. Feilloos wist je

direct knelpunten te noemen en deze te relativeren met een lach. Laten we daarmee doorgaan.

Lex, dank voor je bijdrage aan het Nederlands. Ik mag je niet bedanken Engbert, maar noem

hier wel je bijdrage aan tekst en opmaak.

Als laatste maar zo belangrijk, kom ik toe aan mijn familie. Lieve Els en Hanneke, jullie zijn als

zussen een inspiratie bron voor mij hoe te gaan voor wat je wilt. Altijd heb ik steun gevoeld bij

je in dit traject. Ik ben er trots op Els, dat je mijn paranimf bent.

Lieve Philip, Erik, Wendelien en Sunna, er is in de wereld geen groter geluk dan jullie. Dankzij

jullie, doe ik mee met de, voor mij “volgende generatie”. De warmte en humor die ik voel als

we bij elkaar zijn, maken me tot een gelukkig mens. Patijn, Rick, Froukje en Danielle, dank je

wel dat jullie ons zijn komen versterken. En Quinten, dankzij jou heb ik een nieuwe rol.

Tenslotte, Willem, jou wil ik bedanken voor alle jaren continue steun. Steeds heb je meege-

dacht over mijn toekomst, welke stappen ik zou zetten. Je hebt vele uren alleen doorgebracht

als ik achter de pc zat. Met al mijn liefde draag ik dit proefschrift aan je op. Je bent de belang-

rijkste in mijn leven die dit allemaal mogelijk heeft gemaakt. Je bent mijn basis van waaruit ik

kan groeien. Ik kijk ernaar uit dat we weer meer tijd hebben voor elkaar.

224

CURRICULUM VITAE

Maaike Barents obtained her medical degree in 1977 at the Rijks Universiteit Groningen

(RUG). She started to investigate the diagnosis of cow’s milk allergy among infants while work-

ing for the Dutch Youth Health Organization in cooperation with the UMCG, Children’s Depart-

ment (prof. dr. H.S.A. Heymans).

In 1994, she acquired her specialty degree as physician-sexuologist at the Dutch Society of

Sexuologists, NVVS.

After her residency in the Academic Rehabilitation Centre in Haren (prof. Dr. J.H. Arendzen and

drs. L.D.W. Vos) she decided to specialize in elderly care medicine at the Free University of

Amsterdam in collaboration with nursing home het Zonnehuis in Zuidhorn. In the same time, she

took the initial steps for the work as described in this thesis in collaboration with the UMCG

(Thorax centre, Department of cardiology (dr. M.J.L. de Jongste) and Department of Clinical

Chemistry (prof. F.A.J. Muskiet)). When settled as elderly care physician in Groningen and in

addition to patient care, she completed her Ph D work at the last described Department. She is

married to Willem Bohmers and they have two sons (Philip, 1978 and Erik, 1980) and two

daughters (Wendelien, 1982 and Sunna, 1984).