Michalsen (2003) Thermal hydrotherapy improves quality of life and hemodynamic function in patients with chronic heart failure.

Thermal hydrotherapy improves quality of life and
hemodynamic function in patients with chronic
heart failure
Andreas Michalsen, MD,aRainer Lu dtke, PhD,cMalte Bu hring, MD,bGu nther Spahn, MD,aJost Langhorst, MD,a
and Gustav J. Dobos, MDaEssen and Berlin, Germany
BackgroundChronic heart failure is characterized by increased peripheral vascular resistance and reduced pe-
ripheral perfusion due to adrenergic and renin angiotensin activation and impaired endothelial function. Recent studies
have shown that nonpharmacological peripheral vasodilation with thermal therapy by means of warm-water baths and
sauna has benecial effects in chronic heart failure. European hydrotherapy (according to Kneipp) additionally uses short
cold water stimuli, which lead to prolonged vasodilation and adaptive responses. Studies on the efcacy of hydrotherapy
in chronic heart failure are lacking.
MethodsWe studied 15 patients (5 men, 10 women, mean (SD) age 64.31.8 years) with mild chronic heart
failure (NYHA functional class II to III, ejection fraction 30% 40%). Patients were randomly assigned to 6 weeks of inten-
sive home-based hydrotherapy or 6 weeks restriction in a crossover intervention trial. Quality of life and heart-failurere-
lated symptoms were assessed by means of a validated questionnaire (PLC). Graded bicycle exercise test with incremental
workloads (0, 50, 75, 100 watts) was performed at the end of each treatment period. The hydrotherapeutic program
consisted of a structured combination of daily home-based external warm- and cold-water applications.
ResultsBaseline characteristics were balanced between the groups. With hydrotherapy, a signicant (P.05) im-
provement in 3 of 6 dimensions of quality of life (mood, physical capacity, enjoyment) and a signicant reduction in
heart-failurerelated symptoms was found. Heart rates at rest and at 50-Watt workload were signicantly reduced by hy-
drotherapy; blood pressure decreased nonsignicantly at rest and during exercise. The hydrotherapeutic treatment was
well accepted and no relevant adverse effects were observed.
ConclusionsA home-based hydrotherapeutic thermal treatment program improves quality of life, heart-failurere-
lated symptoms and heart rate response to exercise in patients with mild chronic heart failure. The results of this investiga-
tion suggest a benecial adaptive response to repeated brief cold stimuli in addition to enhanced peripheral perfusion
due to thermal hydrotherapy in patients with chronic heart failure. (Am Heart J 2003;146:e11.)
The systemic response in chronic heart failure (CHF)
is characterized by generalized neurohumoral excita-
tion in order to compensate for reduced cardiac out-
put.
1Increased peripheral vascular resistance and re-
duced peripheral perfusion due to sympathetic and
renin angiotensin activation and endothelial dysfunc-
tion contribute mainly to leading clinical symptoms
such as acrocyanosis, muscle fatigue, cold feet, heavi-ness in the limbs, and constipation.
2Afterload reduc-
tion by means of pharmacological vasodilators, such as
angiotensin-converting enzyme inhibitors, improves
hemodynamic function, symptoms, and prognosis in
CHF.
3,4However, vasodilation is also inducible non-
pharmacologically by warm environment and warm
thermal therapies.
5Accordingly, improved cardiovascu-
lar hemodynamics with reduction of cardiac afterload
and increased subjective well-being were found in pa-
tients with chronic CHF after sauna and warm-water
baths.
6Moreover, an improvement in vascular endo-
thelial function has been demonstrated with repeated
warm thermal treatments in healthy volunteers and
patients with CHF.
7,8
Thermally induced peripheral vasodilation can also
be achieved by cold water stimuli with subsequent
vasodilation.
5,9Thermoregulatory mechanisms induce a
reactive prolonged peripheral vasodilation after cold
From theaKliniken Essen Mitte, Department of Internal Medicine V, Essen,bFree Uni-
versity of Berlin, Klinikum Benjamin Franklin, Immanuel Hospital, Department of Internal
Medicine, Berlin, and the
cKarl and Veronica Carstens Foundation, Essen, Germany.
Submitted August 7, 2002; accepted January 28, 2003.
Reprint requests: Gustav Dobos, MD, Kliniken Essen Mitte, Department of Internal and
Integrative Medicine; Am Deimelsberg 34 a; 45276 Essen, Germany.
E-mail: gustav.dobos@uni-essen.de
2003, Mosby, Inc. All rights reserved.
0002-87032003$30.000
doi:10.1016S0002-8703(03)00314-4

water immersions, cold water pourings, and show-
ers.10,11After serial cold water applications, a physio-
logic cold adaptation with consecutive attenuated cate-
cholamine response was observed.
12,13
In central-European physical therapy, warm-water
baths and sauna are commonly supplemented by re-
peated cold water stimuli with peripheral cold water
immersions and cold water pourings (hydrotherapy
according to Kneipp).
14Serial hydrotherapeutic cold-
and warm-water applications are also used as a sup-
portive treatment for patients with coronary artery
disease and patients with CHF in cardiological rehabili-
tative facilities. The benecial effects of hydrotherapy
in patients with CHF, hypertension, and coronary ar-
tery disease have been described in some empirical
and observational reports.
14-16However, its efcacy
has never been tested in controlled trials. The current
study was designed to test whether a specic and in-
tensive home-based hydrotherapeutic treatment pro-
gram, consisting of a structured combination of warm
and cold thermal applications, can induce improve-
ments in exercise performance, quality of life (QOL),
and heart-failurerelated symptoms in patients with
mild CHF.
Methods and patients
Participation in the study was offered by advertise-
ment in cooperating cardiological outpatient facilities.
Of 30 patients screened, 17 patients (6 men, 11
women, mean (SEM) age 66.21.6 years) met the
inclusion criteria and consented to participate in the
study after complete protocol information. Two pa-
tients (one in each group) dropped out before the end
of the rst study period and were not included in the
further data analysis. The remaining study subjects
were 15 patients (5 men, 10 women, mean (SEM)
age 64.31.8 years) with a history of mild CHF. Anal-
ysis was performed regardless of adherence with the
study treatment (intention to treat). Etiology of heart
failure was classied as idiopathic dilated cardiomyopa-
thy in 3 patients, ischemic heart disease in 8 patients,
and chronic hypertension in 4 patients. The criteria for
heart failure were clinical and included dyspnea at ex-
ertion, fatigue, or uid retention, with objective conr-
mation by an echocardiographic or ventriculographic
ejection fraction40%, documented within the last 6
months before study entry. To achieve a homogeneous
study group, we recruited only patients in sinus
rhythm with New York Heart Association (NYHA)
functional class II to III and an ejection fraction be-
tween 30% to 40%. No patient had evidence of exer-
cise-induced arrhythmias or symptomatic myocardial
ischemia at baseline.
Exclusion criteria were myocardial infarction or un-
stable angina within the previous 6 months. Other ex-clusion criteria were chronic obstructive pulmonary
disease, peripheral vascular disease, hemodynamically
signicant valvular heart disease, orthopedic limitation,
and neurologic disease.
Informed consent was obtained from all patients be-
fore they entered the study, which was approved by
the local ethics committee. No patient had a change in
his or her medical regimen for the duration of the
study.
The study was designed as a random-order crossover
comparison of 6 weeks of intensive hydrotherapy and
6 weeks of restricted therapy (control phase). Patients
were not told to expect hydrotherapy to be better for
them, merely that we were trying to nd out objec-
tively the effects of hydrotherapy in heart failure. Data
analysis was performed for all patients with at least
one follow-up examination, independent of treatment
adherence (intention to treat).
The exercise test was carried out on an upright bicy-
cle ergometer with 2-minute incremental-stage work
loads of 0, 50, 75, and 100 watts (w). All tests were
done at the same time in the morning after an over-
night fast. Heart rate was recorded from the 12-lead
electrocardiogram and blood pressure measured by
cuff sphygmomanometer every minute. Rate-pressure
product (double product) was derived as systolic
blood pressure times heart rate.
QOL was assessed by means of a validated inventory:
the quality of life prole for chronic diseases (PLC).17
The PLC is based on 40 items, which are grouped to 6
functional or symptom dimensions: physical capacity
(8 items), positive mood (5 items), negative mood
depression (8 items), ability to enjoy and relax (8
items), social well-being (5 items) and social function-
ingcontact ability (6 items). Each item is a question
that has to be answered on a 5-pointrating scale (0-4).
High internal consistency (0.75) and retest-reliabil-
ity have been demonstrated for the scales in different
populations.
18,19A relevant increase in QOL is dened
as a procedure-related signicant increase in at least 2
of the 6 dimensions. Heart-failurerelated symptoms
were measured by an additional 5-point Likert-scaled
symptom questionnaire with 17 items.
Generally, results are expressed as meanSEM. Ac-
cording to the recommendations for the analysis of
crossover trials,
20we tted split-plot ANOVA models
to the data. Group (carryover effect) was modeled
as a between-subject factor; time point and treatment
as within-subject factors. Tests for treatment effects
are valid if no carryover effects can be assumed. Ef-
fects were assumed to be statistically signicant withP
values.05.
Hydrotherapeutic program
At the start of the hydrotherapy period, patients
were instructed, over 90 minutes, in the common
American Heart Journal
October 20032Michalsen et al

techniques of thermal therapy and hydrotherapy by an
experienced physician. Patients were advised to prac-
tice warm and cold applications at least 3 times a day
to a total maximum of 30 minutes daily. Warm thermal
applications consisted of peripheral warm water baths
(arm baths, foot baths) with incremental temperature
(maximum 40C) and warm sheet packs. For cold ap-
plications, short-term arm or foot baths and peripheral
water pourings with a water temperature below 18C
were taught. Patients were instructed to apply the hy-
drotherapeutic applications long enough to induce a
postprocedural reactive feeling of warmth with respec-
tive mild redness of the treated skin area, but no
longer than 15 minutes for baths and 5 minutes for
cold pourings. The instructions were completed by a
practical demonstration. For each patient, an individual
home-based combination program was developed at
the end of the instruction.
Results
The baseline characteristics of the patients are given
in Table I. Drug treatment did not change during the
3-month study period. Both groups were statistically
similar with regards to ejection fraction, age, body
mass index, and pharmacological treatments. Of 17
randomized patients, 2 (one in each group) dropped
out before the end of therst study period and were
not included in the presented data analysis. Causes for
study termination were not related to the intervention
(acute coronary bypass operation in one patient; acute
surgery of the prostate because of newly diagnosed
cancer in the other). All other subjects completed the
trial. The hydrotherapeutic program was well toler-
ated. Adherence, assessed by interview, was excellentin 14 of 15 study subjects. One patient did not prac-
tice the hydrotherapeutic applications because of lack
of motivation, but was included in the data analysis.
QOL and symptoms
Baseline values for QOL and heart-failurerelated
symptom scores were balanced between the hydro-
therapy-rst and restriction-rst group; in the dimen-
siondepression,the restriction-rst patients had a
nonsignicant lower initial level of depression. With
hydrotherapy, the values in all dimensions of QOL in-
creased in both groups. With restriction, levels de-
creased in all dimensions. Baseline values were not
reached in 4 dimensions; however, no signicant car-
ryover effects were detected. Complete data for QOL
in both groups are given in Table II. With pooling of
the data according to hydrotherapyrestriction levels of
all dimensions, QOL was increased with hydrotherapy
(Table III). Signicant increases were observed in the
dimensions of positive mood, physical capacity, and
enjoymentrelaxation. Additionally, heart-failurerelated
symptoms were signicantly reduced after 6 weeks of
hydrotherapy.
Hemodynamic data
Exercise tests were completed in all 15 patients. All
patients reached a 75-W load as maximum exercise; 8
patients reached 100-W load. Heart rate at rest and the
rate-pressure product at rest were signicantly lower
after the hydrotherapeutic program. Blood pressure at
rest was nonsignicantly reduced after hydrotherapy
(Table IV).
Table I.Baseline characteristics by treatment order
Hydrotherapy
rst (n8)Restriction
rst (n7)
Age (y) 67.32.3 61.02.2
Sex (MF) 35 25
Weight (kg) 74.44.5 79.63.9
Body mass index (kgm
2)26.41.3 28.31.3
No in NYHA IIIII 71 61
Ejection fraction (%) 35.41.4 35.30.8
Resting heart rate (beatsmin) 78.46.1 81.04.2
Heart failure etiology
Hypertension (n) 2 2
Ischemic heart disease (n) 4 4
Dilative cardiomyopathy (n) 2 1
ACE inhibitors (n) 6 5
Diuretics (n) 3 4
Digitalis (n) 4 4
-Blockers (n) 3 5
Results are given as meanSEM or in absolute numbers.NYHA,New York
Heart Association Class;ACE,angiotensin-converting enzyme inhibitors.
Figure 1
Heart rate at rest and in response to graded bicycle exercise work-
loads (0, 50, 75, 100 W) after 6 weeks of hydrotherapy (solid
line) or restriction (dashed line); pooled data for n15. *P.05
for between group difference (P.003 at 50 W; 0.016 at rest).
American Heart Journal
Volume 146, Number 4Michalsen et al3

Figure 1 shows the effect of hydrotherapy compared
to restriction on heart rate with incremental work-
loads. With hydrotherapy, heart rate at the 50-W load
was signicantly reduced (P.003); there was a
trend towards a lower heart rate at 75 W (P.10)
and 100 W (P.09). There was a modest but nonsig-
nicant reduction in blood pressure during exercise
with hydrotherapy compared to restriction. Rate-pres-
sure product was nonsignicantly reduced at both the
50-W and 75-W workload and signicantly reduced at
the 100-W workload. In interviews at the end of the
study, 14 of 15 patients rated hydrotherapy as gener-
ally benecial for wellbeing.
Discussion
Hydrotherapy has a long tradition in central Europe
and is established in some rehabilitative cardiological
clinics for the adjunctive nonpharmacological treat-
ment of patients with cardiovascular disease and heart
failure.
16The current study, for therst time, investi-
gated the clinical and hemodynamic effects of serialhydrotherapy in CHF by means of a randomized, con-
trolled trial. Our data indicate that an intensive 6-week
hydrotherapy program according to Kneipp improves
QOL, as well as heart-failurerelated symptoms, and
reduces resting heart rate and heart rate exercise re-
sponse to a clinically relevant extent.
There are several potential mechanisms that could
be responsible for the observed benecial effects of
intensive hydrotherapy on clinical symptoms and he-
modynamic function in CHF. Hydrotherapy consists of
serial thermal warm- and cold-water applications.
Whereas warm-water applications and immersions di-
rectly lead to peripheral vasodilation,
5cold immersions
and pourings provoke immediate vasoconstriction with
reactive prolonged vasodilation and subsequent cold
adaptation with serial treatment.
11
Thermal therapy, by means of sauna and warm-water
baths, has been shown to reduce cardiac afterload and
preload and to increase cardiac output in CHF.
6More-
over, repeated sauna treatments improved peripheral
vascular endothelial function and decreased brain natri-
uretic peptide concentrations in patients with chronic
mild heart failure.
7Experimental studies have demon-
strated that CHF impairs endothelial-dependent vasodi-
lation in response to acetylcholine,
21and in humans,
peripheral resistance vessels are largely impaired by
CHF.
22This may occur because of decreased shear
stress due to reduced peripheral perfusion.23Because
shear stress is an important stimulus for the peripheral
vascular production of endothelium-derived nitric ox-
ide (NO),
24treatments that improve peripheral perfu-
sion, like thermal therapy and hydrotherapy, could
induce NO production by the vessels.
Besides, serial immersions and pourings with cold
water induce physiologic adaptive mechanisms. During
cold stimulation, adrenaline and noradrenaline in-
creases are provoked; with serial treatments, an attenu-
Table II.Descriptive statistics of quality of life and heart-failurerelated symptoms during study period by treatment order
Dimension Group Baseline 6 Weeks 12 Weeks
Positive mood Hydrotherapy rst 1.90.2 2.30.2 2.10.2
Restriction rst 1.90.2 1.70.3 2.10.3
Depression Hydrotherapy rst 2.10.3 2.50.3 2.40.2
Restriction rst 2.80.1 2.50.3 2.90.1
Physical capacity Hydrotherapy rst 2.10.3 2.20.2 1.80.1
Restriction rst 2.00.3 1.70.3 2.20.2
Enjoymentrelaxation Hydrotherapy rst 2.20.2 2.50.2 2.30.1
Restriction rst 2.30.2 2.10.3 2.40.2
Social wellbeing Hydrotherapy rst 2.50.3 3.10.2 3.00.2
Restriction rst 2.40.4 2.40.3 2.70.3
Social functioning Hydrotherapy rst 2.60.2 2.70.2 2.60.1
Restriction rst 2.60.3 2.10.4 2.50.3
Heart failure-related symptoms Hydrotherapy rst 18.02.7 13.62.5 17.12.8
Restriction rst 18.15.1 19.64.0 13.43.7
Table III.Quality of life and heart-failurerelated symptoms
Dimension Hydrotherapy RestrictionP
(hydrotherapy
vs
restriction)
Positive mood 2.20.2 1.90.2 .050
Depression 2.70.2 2.50.2 .122(NS)
Physical capacity 2.20.1 1.80.1 .033
Enjoymentrelaxation 2.50.2 2.20.1 .022
Social wellbeing 2.90.2 2.70.2 .053(NS)
Social functioning 2.60.2 2.30.2 .073(NS)
Heart-failure
related symptoms13.52.1 18.32.3 .028
Pooled data after 6 weeks of hydrotherapy versus 6 weeks of restriction.American Heart Journal
October 2003
4Michalsen et al

ated catecholamine response as well as diminished
catecholamine resting levels were found.11,12,25Further
adaptations relate to cytokine production, circulating
cortisol concentrations, and increased antioxidative
capacity.
26,27In summary, these adaptive mechanisms
can be interpreted as an increased tolerance to envi-
ronmental stress and show similarities to some effects
of physical exercise in CHF.
28
Overall, the treatment program was well accepted
and no patient had worsened clinical symptoms during
hydrotherapy. Compliance, a major problem in CHF
therapy,
29was excellent, presumably because of the
short treatment period and because hydrotherapy was
perceived as generally enjoyable and refreshing.
The current data must be interpreted in the context
of some important limitations of the study. We applied
hydrotherapy only to a limited number of patients
from outpatient facilities with mainly NYHA functional
class II and a modestly impaired ejection fraction. Only
half of the population received diuretics, and we can-
not rule out the possibility that the observed symptom-
atic improvement is partly due to nonspecic treat-
ment effects. Being overweight was common in the
studied patients, but no weight changes were induced
by the treatment. Due to the recruitment strategy, the
ratio of women to men in the study population was
unusual for heart failure trials, which reduces the ex-
ternal validity of our results.
The current study tested the efcacy of hydrother-
apy by means of a crossover design. As no signicant
carryover effects could be detected, we assume that
the benecial effects of thermal hydrotherapy are not
maintained with restriction and, consequently, the ap-
plications need to be applied continuously by the pa-
tient for sustained benet. Therefore, the long-term
effects of hydrotherapy may be compromised by de-
creasing adherence, as is commonly seen with other
self-care interventions, such as exercise training. Thus,
assessment of long-term adherence remains critical in
future studies with hydrotherapy.
Finally, an important limitation results from the un-
blinding of the intervention; however, blinding is notfeasible in a self-care based physical therapy approach.
Besides the nonspecic effects due to unblinding, the
ritualistic component of serial hydrotherapy might fur-
ther contribute to the improvement in QOL. Assess-
ment of outcome expectations, as well as comparison
of 2 differentially scheduled treatment programs, could
be useful for further evaluation of these subjective
components.
Ourndings imply that an appropriately performed
home-based hydrotherapeutic program may provide a
practical, salutary, nonpharmacological therapy for pa-
tients with CHF without the need for expensive reha-
bilitation facilities. Furthermore, this therapy approach
may be applicable in patients who are unable to par-
ticipate in exercise training.
We conclude that a simple, home-based, hydrothera-
peutic program is feasible and effective in improving
symptoms, QOL, and hemodynamics of patients with
nonsevere CHF. Further studies are needed to investi-
gate the effects of hydrotherapy in patients with larger
populations and more severe heart failure, and to clar-
ify the mechanisms behind this nonpharmacological
therapy approach.
References
1. Benedict CR, Johnstone DE, Weiner DH, et al. Relation of
neurohumoral activation to clinical variables and degree of
ventricular dysfunction: a report from the registry of studies
on left ventricular dysfunction. J Am Coll Cardiol 1994;23:
141020.
2. Floras JS. Clinical aspects of sympathetic activation and parasyma-
phetic withdrawal in heart failure. J Am Coll Cardiol 1993;22:72
84A.
3. Pfeffer MA, Braunwald E, Moye LA, et al. Effect of captopril on
mortality and morbidity in patients with left ventricular dysfunction
after myocardial infarction: results of the survival and ventricular
enlargement trial. The SAVE Investigators. N Engl J Med 1992;
327:66977.
4. Pouleur H. Results of the treatment trial of the studies of left ventric-
ular dysfunction (SOLVD): the SOLVD investigators. Am J Cardiol
1992;70:1356C.
Table IV.Hemodynamic data at rest at baseline and after 6 weeks of restriction or hydrotherapy
Baseline Hydrotherapy RestrictionP
(hydrotherapy
vs restriction)
Blood pressure at rest
Systolic 142.34.5 137.53.7 143.35.0 .207 (NS)
Diastolic 84.71.4 83.32.1 85.72.1 .195 (NS)
Resting heart rate (beatsmin) 79.63.4 76.13.4 79.43.3 .016
RPP (beatsminmm Hg) 11406714 10491577 11369692 .024
Pooled data; n15.RPP,Rate-pressure product. American Heart Journal
Volume 146, Number 4
Michalsen et al5

5. Rowell LB, Brengelmann GL, Blackmon JR, et al. Redistribution of
bloodow during sustained high skin temperature in resting man.
J Appl Physiol 1970;28:41520.
6. Tei C, Horikiri Y, Park J-C, et al. Acute hemodynamic improvement
by thermal vasodilation in congestive heart failure. Circulation
1995;91:258290.
7. Kihara T, Biro S, Imamura M, et al. Repeated sauna treatment im-
proves vascular endothelial and cardiac function in patients with
chronic heart failure. J Am Coll Cardiol 2002;39:7549.
8. Imamura M, Biro S, Kihara T, et al. Repeated thermal therapy im-
proves impaired vascular endothelial function in patients with cor-
onary risk factors. J Am Coll Cardiol 2001;38:10838.
9. Kirsch KA, Rocker DL, von Ameln H, et al. The cardiaclling pres-
sures following exercise and thermal stress. Yale J Biol Med 1986;
59:25765.
10. Cordes J. Die thermische Hautreaktion in der Hydrotherapie. Z.
Physiother 1972;24:2415.
11. Hildebrandt G. Therapeutische Physiologie. In: Amelung G, Hilde-
brandt G, editors. Balneologie und Klimatologie. Berlin: Springer;
1985. p. 32856.
12. Huttunen P, Rintamaki H, Hirvonen J. Effect of regular winter
swimming on the activity of the sympathoadrenal system before
and after a single cold water immersion. Int J Circumpolar Health
2001;60:4006.
13. Cordes J. Aktuelle Erkenntnisse zu theoretischen Grundlagen und
zur Praxis der Hydrotherapie. Z Physiother 1984;36:41523.
14. Buhring M. Die Kneippsche Hydrotherapie in der Praxis. Thera-
peutikon 1988;2:806.
15. Gutenbrunner C, Ruppel K. Zur Frage der adaptiven Blutdrucknor-
malisierung im Verlauf von Baderkuren unter besonderer Beruck-
sichtigung von Homogenisierungseffekten und Lebensalter. Phys
Rehab Kur Med 1992;2:5864.
16. Bruggemann W. Hydrotherapie. Berlin: Springer; 1986. p. 836.
17. Siegrist J, Broer M, Junge A. PLC-Prol der Lebensqualitat chro-
nisch Kranker. Gottingen: Beltz-Test GmbH; 1996.18. Junge A, Funfstuck G, Siegrist J. Ein Fragebogen zur Erfassung
der Lebensqualitat - erste teststatistische Ergebnisse am Beispiel
von Hypertonikern. Diagnostika 1990;36:3538.
19. Goldbeck L, Schmitz TG. Comparison of three generic question-
naires measuring quality of life in adolescents and adults with cys-
ticbrosis: the 36-item short form health survey, the quality of life
prole for chronic diseases, and the questions on life satisfaction.
Qual Life Res 2001;10:236.
20. Jones B, Kenward MG. Design and analysis of cross-over trials.
London: Chapman and Hall; 1989.
21. Kaiser L, Spickard RC, Olivier NB. Heart failure depresses endo-
thelium-dependent responses in canine femoral artery. Am J
Physiol 1989;256:H9627.
22. Kubo SH, Rector TS, Bank AJ, et al. Endothelium-dependent vaso-
dilation is attenuated in patients with heart failure. Circulation
1991;84:158996.
23. Drexler H. Hypertension, heart failure, and endothelial function.
Am J Cardiol 1998;82:202S.
24. Buga GM, Gold ME, Fukuto JM, et al. Shear stress-induced re-
lease of nitric oxide from endothelial cells grown on beads. Hyper-
tension 1991;17:18793.
25. Budd GM, Brotherhood D, Thomas DW, et al. Cardiovascular and
metabolic responses to noradrenaline in men acclimatized to cold
baths. Eur J Appl Physiol 1993;67:4506.
26. Dugue B, Leppanen E, Leppanen E. Adaptation related to cyto-
kines in man: effects of swimmimng in ice-cold water. Clin Physiol
2000;20:11421.
27. Siems WG, Brenke R, Sommerburg O, et al. Improved antioxida-
tive protection in winter swimmers. QJM 1999;92:1938.
28. Kiilavuori K, Naveri H, Leinonen H, et al. The effect of physical
training on hormonal status and exertional hormonal response in
patients with chronic congestive heart failure. Eur Heart J 1999;
20:45664.
29. Michalsen A, Konig G, Thimme W. Preventable causative factors
leading to hospital admission with decompensated heart failure.
Heart 1998;80:43741.
American Heart Journal
October 20036Michalsen et al