Journal of the American College of Cardiology Vol. 53, No.

7, 2009
© 2009 by the American College of Cardiology Foundation ISSN 0735-1097/09/$36.00
Published by Elsevier Inc. doi:10.1016/j.jacc.2008.08.080

Heart Failure

Increased Central Venous

Pressure Is Associated With Impaired
Renal Function and Mortality in a Broad
Spectrum of Patients With Cardiovascular Disease
Kevin Damman, MD,* Vincent M. van Deursen,* Gerjan Navis, MD, PHD,†
Adriaan A. Voors, MD, PHD,* Dirk J. van Veldhuisen, MD, PHD, FACC,*
Hans L. Hillege, MD, PHD*‡
Groningen, the Netherlands

Objectives We sought to investigate the relationship between increased central venous pressure (CVP), renal function, and
mortality in a broad spectrum of cardiovascular patients.

Background The pathophysiology of impaired renal function in cardiovascular disease is multifactorial. The relative impor-
tance of increased CVP has not been addressed previously.

Methods A total of 2,557 patients who underwent right heart catheterization in the University Medical Center Groningen,
the Netherlands, between January 1, 1989, and December 31, 2006, were identified, and their data were ex-
tracted from electronic databases. Estimated glomerular filtration rate (eGFR) was assessed with the simplified
modification of diet in renal disease formula.

Results Mean age was 59 ⫾ 15 years, and 57% were men. Mean eGFR was 65 ⫾ 24 ml/min/1.73 m2, with a cardiac
index of 2.9 ⫾ 0.8 l/min/m2 and CVP of 5.9 ⫾ 4.3 mm Hg. We found that CVP was associated with cardiac in-
dex (r ⫽ ⫺0.259, p ⬍ 0.0001) and eGFR (r ⫽ ⫺0.147, p ⬍ 0.0001). Also, cardiac index was associated with
eGFR (r ⫽ 0.123, p ⬍ 0.0001). In multivariate analysis CVP remained associated with eGFR (r ⫽ ⫺0.108,
p ⬍ 0.0001). In a median follow-up time of 10.7 years, 741 (29%) patients died. We found that CVP was an in-
dependent predictor of reduced survival (hazard ratio: 1.03 per mm Hg increase, 95% confidence interval: 1.01
to 1.05, p ⫽ 0.0032).

Conclusions Increased CVP is associated with impaired renal function and independently related to all-cause mortality in a
broad spectrum of patients with cardiovascular disease. (J Am Coll Cardiol 2009;53:582–8) © 2009 by the
American College of Cardiology Foundation

Renal dysfunction is a strong and independent predictor of hypertension, not only was renal perfusion strongly associ-
prognosis in the general population but also in patients with ated with renal function impairment but also with venous
diabetes, hypertension, coronary artery disease, and heart congestion (11). However, it is unclear whether this obser-
failure (1–7). The pathophysiology is multifactorial and vation is limited to those patients with reduced cardiac
associated with decreased renal perfusion, atherosclerosis function and pulmonary hypertension or whether it also
and inflammation, endothelial dysfunction, and neurohor- may be present in patients with a mixture of cardiovascular
monal activation (8 –10). We recently showed that in diseases with varying etiologies and treatments. In addition,
patients with cardiac dysfunction secondary to pulmonary there are only limited data on the relationship between

See page 597

From the Departments of *Cardiology, †Nephrology, and ‡Epidemiology, University

Medical Center Groningen, University of Groningen, Groningen, the Netherlands.
venous congestion, as estimated by central venous pressure
Dr. Damman is supported by the Netherlands Heart Foundation (grant 2006B157). (CVP) and the impact on prognosis, even in patients with
Dr. Voors and Dr. van Veldhuisen are Clinical Established Investigators of the and without heart failure. The studies that have been
Netherlands Heart Foundation (grants 2006T37 and D97-017, respectively).
Manuscript received February 6, 2008; revised manuscript received August 12, conducted are either small or include only noninvasive
2008, accepted August 18, 2008. assessment of increased venous congestion, such as jugular
JACC Vol. 53, No. 7, 2009 Damman et al. 583
February 17, 2009:582–8 Central Venous Pressure and Renal Function

venous pressure (12–15). In the present study, we therefore and percentages for categorical Abbreviations
aimed to investigate the relationship between CVP, renal variables. Associations between and Acronyms
function, and mortality in patients with a mixture of baseline variables were evaluated
CI ⴝ confidence interval
cardiovascular diseases with varying etiologies and by means of 1-way analysis of
CVP ⴝ central venous
treatments. variance, the Kruskal-Wallis test, pressure
and chi-square or Fisher exact
eGFR ⴝ estimated
tests, when appropriate. Two- glomerular filtration rate
Methods
sided p values were used, taking HR ⴝ hazard ratio
Case identification. Using the patient registration system p ⬍ 0.05 to be statistically sig-
of the University Medical Center Groningen, the Nether- nificant. CVP was divided into
lands, all patients that underwent right heart catheterization tertiles to assess relationships between baseline characteris-
between January 1, 1989, and December 31, 2006, were tics and CVP. A fractional polynomial parameterization of
identified. exposure was used to explore nonlinearity between different
Data extraction. Retrospective chart review was performed predictors and renal function. In this technique, each
to analyze characteristics of all patients that were identified exposure value is expressed as a polynomial of degree ⬎1
during the electronic search. For each patient, date of birth, (e.g., quadratic, cubic, and so on), yielding an estimated
sex, race, and weight and height were collected. Comorbid model with multiple predictors (i.e., separate predictors for
conditions, including hypertension, coronary artery disease, the linear, quadratic, terms, respectively). We used a Cox
cardiac valve disease, congenital heart disease, history of proportional hazards survival model to estimate hazard
stroke, hypercholesterolemia, and diabetes, in addition to ratios (HRs) with 95% confidence intervals (CIs). At first,
medical treatment at the time of catheterization also were in multivariate analysis, CVP was fitted into a stepwise
extracted for each patient. Furthermore, the reason for multivariate Cox regression analysis on a continuous scale.
performing right heart catheterization was identified. The In secondary analysis, CVP was fitted into the model and, in
study was approved by the institutional review board of the multiple steps, the model was adjusted for other variables
University Medical Center Groningen. and parameters. The internal validity of the regression
Heart catheterization. Hemodynamic variables obtained model was assessed by the bootstrap resampling technique
during catheterization included systolic blood pressure (mm (18). For each of 100 bootstrap samples, the model was
Hg), diastolic blood pressure (mm Hg), cardiac output refitted and tested on the original sample to obtain a
(thermodilution, l/min), and right atrial pressure as indicator bias-corrected estimate of predictive accuracy. Statistical
of CVP (CVP, mm Hg). Cardiac index (l/min/m2) was analyses were performed using SPSS version 12.0 (Chicago,
determined as cardiac output divided by the body surface area, Illinois) and STATA version 9.0 (College Station, Texas).
which was calculated as: 0.007184·weight0.425·length0.725.
Measurements obtained from cardiac catheterization were Results
obtained from the patient during a resting state.
Baseline characteristics. A total of 3,757 right heart cath-
Renal function measurement. Serum creatinine at the day
eterizations were conducted between 1989 and 2006. Of
of catheterization was extracted. For the patients who did
these, 2,557 (68%) were first or only right heart catheter-
not have laboratory measurements on the day of catheter- Primary Indication for Heart Catheterization
ization, measurements obtained within 3 days before cath-
Table 1 Primary Indication for Heart Catheterization
eterization were taken as the baseline value. Of patients
included in the study, 2,282 (89%) had at least 1 serial Percentage of Patients
creatinine measurement within 3 days of catheterization. Aortic valve disorders 29

Renal function was estimated as glomerular filtration rate Aortic valve stenosis 23
Aortic valve insufficiency 6
(GFR) by using the simplified modification of diet in renal
Mitral valve disorders 15
disease equation (estimated glomerular filtration rate
Mitral valve insufficiency 14
[eGFR] [ml/min/1.73 m2] ⫽ 186.3 ⫻ [serum creati-
Mitral valve stenosis 1
nine]⫺1.154 ⫻ age⫺0.203 ⫻ [0.742 if female]) (16). Esti- Pulmonary valve disorders 1
mated GFR values ⬎200 ml/min/1.73 m2 were set equal to Pulmonary valve insufficiency 0.8
200 ml/min/1.73 m2, according to Coresh et al. (17). Pulmonary valve stenosis 0.2
Mortality data. Survival status was determined using the Heart failure 16
electronic patient registration database of the University Coronary artery disease 13
Medical Center Groningen. Follow-up started directly after Pre-transplantation (noncardiac) 11
right heart catheterization. The primary end point consisted Rhythm disorders 5

of death from any cause. Pulmonary hypertension 3

Statistical analysis. Data are given as mean ⫾ standard Congenital heart disease 2
Post-heart transplantation 2
deviation when normally distributed, as median and inter-
Other 6
quartile range when skewed distributed, and as frequencies
584 Damman et al. JACC Vol. 53, No. 7, 2009
Central Venous Pressure and Renal Function February 17, 2009:582–8

Baseline Characteristics According to Tertiles of CVP

Table 2 Baseline Characteristics According to Tertiles of CVP

Total Tertile 1 (0 to 3 mm Hg) Tertile 2 (4 to 6 mm Hg) Tertile 3 (>6 mm Hg) p Value for Trend
n 2,557 911 855 791
Age (yrs) 59 ⫾ 15 60 ⫾ 15 59 ⫾ 15 58 ⫾ 15 0.0032
Sex (% male) 57 59 58 54 NS
SBP (mm Hg) 133 ⫾ 29 133 ⫾ 28 134 ⫾ 27 129 ⫾ 31 0.0100
DBP (mm Hg) 68 ⫾ 13 66 ⫾ 12 68 ⫾ 12 69 ⫾ 13 0.0010
CO (l/min) 5.5 ⫾ 1.6 5.7 ⫾ 1.6 5.5 ⫾ 1.5 5.0 ⫾ 1.5 ⬍0.0001
Cardiac index (l/min/m2) 2.9 ⫾ 0.8 3.1 ⫾ 0.7 3.0 ⫾ 0.7 2.7 ⫾ 0.8 ⬍0.0001
CVP (mm Hg) 5.9 ⫾ 4.3 2⫾1 5⫾1 11 ⫾ 4 ⬍0.0001
eGFR (ml/min/1.73 m2) 65 ⫾ 24 65 ⫾ 23 67 ⫾ 24 62 ⫾ 24 0.0001
Medical history (%)
Heart failure 16 15 15 19 NS
Coronary artery disease 24 24 25 24 NS
Congenital heart disease 5 4 5 7 0.0189
Valve disease 51 50 55 49 NS
Hypercholesterolemia 6 7 5 6 NS
Diabetes mellitus 9 8 8 10 NS
Hypertension 20 21 20 18 NS
Stroke 5 4 5 6 NS
Medication (%)
Diuretics 42 37 38 53 0.0001
Beta-blocker 28 25 29 31 0.0388
ACEI or ARB 38 36 32 45 0.0001
Aldosterone antagonist 9 5 6 15 0.0001

ACEI ⫽ angiotensin-converting enzyme inhibitor; ARB ⫽ angiotensin-II receptor blocker; CO ⫽ cardiac output; CVP ⫽ central venous pressure; DBP ⫽ diastolic blood pressure; eGFR ⫽ estimated glomerular
filtration rate; NS ⫽ not significant; SBP ⫽ systolic blood pressure.

ization of unique patients and formed the study population. years, and 57% were men (Table 2). In the total study
Main reasons for right heart catheterization are shown in population, both mean cardiac index (2.9 ⫾ 0.8 l/min/m2)
Table 1. Aortic and mitral valve disorders accounted for and mean CVP (5.9 ⫾ 4.3 mm Hg) were within the normal
44% of indications, whereas in 16%, acute or chronic heart range. The distribution of CVP among the study population
failure was the predominant reason. Mean age was 59 ⫾ 15 is shown in Figure 1. Mean eGFR was moderately im-
paired: 65 ⫾ 24 ml/min/1.73 m2.
The distribution of different factors over tertiles of CVP
is shown in Table 2. Most of the characteristics were equally
distributed across tertiles of CVP, except for the highest
tertile (CVP ⬎6 mm Hg). Both cardiac output and cardiac
index were significantly lower in the highest tertile com-
pared with lower tertiles (p ⬍ 0.0001), corresponding to r ⫽
⫺0.259 (p ⬍ 0.0001) for the association between CVP and
cardiac index. Furthermore, patients in the highest tertile
were treated more frequently with angiotensin-converting
enzyme inhibitor/angiotensin-II receptor blockers, beta-
blockers, diuretics, and aldosterone antagonists. Prevalence
of heart failure showed a trend toward increasing with
higher tertiles of CVP (p ⫽ 0.0781), whereas congenital
heart disease was also more prevalent in the highest tertile.
Finally, eGFR was significantly lower in the highest tertile
Figure 1 Distribution of CVP and Curvilinear Relationship of CVP, compared with both lower tertiles (p ⬍ 0.001).
Between CVP and eGFR in the Study Population Curvilinear fitting and the relationship between CVP
Adjusted for age, sex, and cardiac index. The curvilinear model had the follow- and eGFR. Figure 1 shows the curvilinear relationship
ing individual polynomial components for the relationship between CVP and between CVP and eGFR in the total study population as
eGFR: First order: Y ⫽ ⫺25.8·(CVP ⫹ 1)/10 (Wald 28.2, p ⬍ 0.0001) and sec-
obtained by fractional polynomial modeling. Estimated
ond order: Y ⫽ 35.7·([CVP ⫹ 1]/10)0.5 (Wald 17.4, p ⬍ 0.0001). CVP ⫽ cen-
tral venous pressure, eGFR ⫽ estimated glomerular filtration rate. GFR showed a small increase when CVP increased from 1
to 6 mm Hg. However, in CVP values ⬎6 mm Hg, a steep
JACC Vol. 53, No. 7, 2009 Damman et al. 585
February 17, 2009:582–8 Central Venous Pressure and Renal Function

decrease is observed with increasing CVP values. This

finding resulted in a partial correlation of r ⫽ 0.064, p ⫽
0.0218 in patients with CVP ⱕ6 mm Hg, and r ⫽ ⫺0.212,
p ⬍ 0.0001 in patients with CVP ⬎6 mm Hg (adjusted for
age, sex, and cardiac index). On a continuous scale, CVP
was also significantly associated with eGFR (r ⫽ ⫺0.110,
p ⬍ 0.0001) after transformation.
Besides CVP, age (r ⫽ ⫺0.438, p ⬍ 0.0001), sex (r ⫽
0.137, p ⬍ 0.0001), and cardiac index (r ⫽ 0.249, p ⬍
0.0001) were associated with eGFR. In addition, lower
eGFR values also were found to be related with the use of
any type of cardiovascular medication and a history of
diabetes and hypertension. There was a significant interac-
tion between CVP and cardiac index on the relationship
Kaplan-Meier Analysis of Event-Free
with eGFR. The observed biphasic relationship between Figure 3
Survival According to Tertiles of CVP
CVP and eGFR was most pronounced in patients with
HR: 1.22 (95% CI: 1.00 to 1.49), p ⫽ 0.0466 for CVP 4 to 6 mm Hg; HR:
relatively normal cardiac index (Fig. 2). 1.65 (95% CI: 1.35 to 2.01), p ⬍ 0.0001 for CVP ⬎6 mm Hg, both compared
In multivariate analysis, CVP remained associated with with CVP 0 to 3. CI ⫽ confidence interval; HR ⫽ hazard ratio; other abbrevia-
eGFR (r ⫽ ⫺0.108, p ⬍ 0.0001, adjusted for covariates), tions as in Figure 1.

which was confirmed by bootstrap analysis (Online Table

1). After adjustment for the year of catheterization, the years and, during follow-up, 741 (29%) of the patients died.
association between CVP and eGFR was numerically un- Crude mortality ranged from 24% in the lowest tertile to
changed: r ⫽ ⫺0.105, p ⬍ 0.0001. Excluding patients with 29% and 35% in the 2 highest tertiles of CVP (p ⫽ 0.0001
a history of heart transplantation, who were likely to receive for trend). On a continuous scale, greater CVP levels were
renal function compromising immunotherapy, CVP re- associated with impaired survival (HR: 1.05 per mm Hg
mained associated with eGFR (r ⫽ ⫺0.108, p ⬍ 0.0001) in increase, 95% CI: 1.04 to 1.07, p ⬍ 0.0001). Kaplan-Meier
multivariate analysis. Including only patients without heart survival curves for tertiles of CVP are shown in Figure 3,
failure, similar associations were present (r ⫽ ⫺0.080, p ⫽ showing that patients with the greatest CVP in particular
0.0034). Excluding both heart transplant recipients and were at risk for increased mortality (HR for CVP ⬎6 mm
heart failure patients, the association between CVP and Hg vs. ⱕ6 mm Hg: 1.49, 95% CI: 1.26 to 1.76, p ⬍
eGFR remained (r ⫽ ⫺0.079, p ⫽ 0.0042). 0.0001). Baseline eGFR (HR: 1.09 per 10 ml/min/1.73 m2
CVP and mortality. Mortality data were available in all decrease, 95% CI: 1.05 to 1.13, p ⬍ 0.0001) and cardiac
patients, whereas time of death was available in 2,424 (95%) index (HR: 0.74 per l/min/m2 increase, 95% CI: 0.66 to
of patients. Median follow-up among survivors was 10.7 0.84, p ⬍ 0.0001) also were strong predictors of mortality.
Other factors associated with reduced survival are shown in
Table 3. In stepwise multivariate Cox regression analysis,
CVP remained significantly associated with reduced survival
(HR: 1.03 per mm Hg increase, 95% CI: 1.01 to 1.05,
p ⫽ 0.0032) (Table 3). Finally, we fitted a second model,
adjusting CVP for other covariates (Online Table 2). CVP
remained an independent predictor of impaired survival
(HR: 1.03 per mm Hg increase, 95% CI: 1.01 to 1.05, p ⫽
0.0144). To account for the effects of changing therapy
during the study inclusion time, we adjusted for the year of
catheterization. This secondary analysis yielded similar re-
sults (HR: 1.03 per mm Hg increase, 95% CI: 1.00 to 1.05,
p ⫽ 0.0207). Excluding patients with known heart failure
and heart transplant recipients, CVP was still associated
with mortality (HR: 1.03 per 5 mm Hg increase, 95% CI:
Curvilinear Relationship Between CVP and
1.00 to 1.06, p ⫽ 0.0369).
Figure 2
eGFR According to Different Cardiac Index Values
Discussion
p ⫽ 0.0217 for interaction between cardiac index and CVP on the relationship
with eGFR. Solid line ⫽ cardiac index ⬍2.5 l/min/m2; dashed line ⫽ cardiac
The present study shows that increased CVP is associated
index 2.5 to 3.2 l/min/m2; dotted line ⫽ cardiac index ⬎3.2 l/min/m2. Abbre-
viations as in Figure 1. with impaired renal function in a broad spectrum of
cardiovascular patients who underwent right heart catheter-
586 Damman et al. JACC Vol. 53, No. 7, 2009
Central Venous Pressure and Renal Function February 17, 2009:582–8

Multivariate Stepwise Cox Regression Model

Table 3 Multivariate Stepwise Cox Regression Model

Univariate Multivariate

Variables Hazard Ratio (95% CI) p Value Hazard Ratio (95% CI) Wald Statistic p Value
Age (per 10 yr increase) 1.05 (1.00–1.10) 0.0880 1.21 (1.12–1.31) 25.4 ⬍0.0001
Sex 1.03 (0.88–1.20) 0.713
Cardiac index (per l/min/m2 increase) 0.74 (0.66–0.84) ⬍0.0001 0.81 (0.71–0.93) 9.1 0.0026
CVP (per mm Hg increase) 1.05 (1.04–1.07) ⬍0.0001 1.03 (1.01–1.05) 8.7 0.0032
eGFR (per 10 ml/min/1.73 m2 decrease) 1.09 (1.05–1.13) ⬍0.0001
Medication
Diuretic use 1.43 (1.22–1.67) ⬍0.0001
ACEI or ARB use 1.29 (1.10–1.52) 0.0017
Aldosterone antagonist use 1.92 (1.50–2.45) ⬍0.0001 1.50 (1.10–2.02) 6.9 0.0087
Medical history (%)
Coronary artery disease 1.26 (1.05–1.50) 0.0112
Diabetes mellitus 1.83 (1.44–2.31) ⬍0.0001 1.76 (1.34–2.31) 16.6 ⬍0.0001
Indication for catheterization
Aorta insufficiency 0.63 (0.43–0.93) 0.0203
Congenital heart disease 0.36 (0.17–0.77) 0.0078
Pre-transplantation (noncardiac) 1.50 (1.21–1.85) ⬍0.0001 2.54 (1.90–3.40) 39.2 ⬍0.0001
Heart failure 1.48 (1.21–1.85) 0.0001 1.71 (1.34–2.20) 18.1 ⬍0.0001
Rhythm disorders 0.49 (0.29–0.82) 0.0062

CI ⫽ confidence interval; other abbreviations as in Table 2.

ization. In addition, the slope between CVP and impaired with renal failure and vice versa (26 –28). Interestingly, a
eGFR was steeper with relatively preserved cardiac function. recent study showed that renal dysfunction is even more
Finally, an increased CVP was a strong and independent important in defining mortality risk in patients with pre-
determinant of all-cause mortality, which was especially served cardiac function compared with those with systolic
observed in patients with a CVP ⬎6 mm Hg. dysfunction (29,30). Our present study confirmed that
There are only limited data on the relationship between increased CVP is an important risk factor for decreased
increased CVP and renal impairment. Studies in animals renal function in patients with preserved and decreased
have shown that increasing renal venous pressure leads to a cardiac function.
reduction in glomerular filtration, which was probably Curvilinear effect of CVP with eGFR. We observed a
mediated by a decreased renal perfusion (19). Renal vein biphasic relationship between eGFR and increasing CVP.
constriction led to a decrease in GFR in rats (20), whereas In the physiologic ranges of CVP, up to 6 mm Hg, eGFR
renal function decreased when renal vein pressure was increases gradually. This subtle increase in eGFR may be a
increased in dogs, but only when cardiac output was reduced reflection of increased cardiac filling to preserve cardiac
(21). We recently showed that in patients with reduced function by Frank-Starling mechanism (pre-load), and sub-
cardiac function, secondary to pulmonary hypertension, sequent renal perfusion (31). This gradual increase in eGFR
increased CVP was strongly associated with renal impair- was observed across the full spectrum of low to high cardiac
ment, especially when renal perfusion was already impaired index. We observed a decrease in eGFR when CVP
(11). Early studies on increased renal vein pressure in heart increases above 6 mm Hg. In these patients, the equilibrium
failure patients and animals showed a marked reduction in among venous return, CO, and CVP may have shifted
renal blood flow as well as water and salt excretion toward a plateau phase or optimum, where CO is not
(22,23), but the effect on GFR was not uniform. One further increased in response to greater CVP (32).
small report showed a strong relationship between CVP Greater CVP levels will then decrease renal perfusion
and renal blood flow in advanced heart failure (24). pressure, which will further impair eGFR. However, if
Drazner et al. (13) reported that in patients with in- greater CVP levels preserve CO, and despite this mech-
creased jugular venous pressure on examination, serum anism, eGFR decreases with greater CVP, this action
creatinine was significantly greater. In patients who under- suggests that increased CVP also may exert an effect on
went elective cardiac surgery, pre-operative presence of high GFR in this group of patients, independent of renal
CVP was a strong predictor of the occurrence of acute renal perfusion.
injury, independent of the presence of low CO (25). Importantly, the relationship between CVP and GFR is
However, especially in patients with preserved cardiac bound to be bidirectional. Not only may CVP influence
function, data regarding the relationship between CVP and GFR, but even mildly impaired renal function may initiate
renal function are scarce. Diastolic dysfunction, a disease salt and water retention, resulting in increased cardiac filling
characterized by increased filling pressures, often coexists pressures (33). Because of the cross-sectional nature of our
JACC Vol. 53, No. 7, 2009 Damman et al. 587
February 17, 2009:582–8 Central Venous Pressure and Renal Function

analysis, we were unable to investigate the cause-effect selected patient populations, such as patients who under-
associations, and our present analysis must be regarded as went Fontan surgery or lung transplantation, greater CVPs
hypothesis generating. were strong predictors of outcome (37,38). The prognostic
CVP and eGFR in patients with and without cardiac importance of increased CVP in patients with normal
dysfunction. Of particular interest is the interaction be- cardiac function has not been reported previously. We show
tween cardiac index and CVP on eGFR. Patients who have that in a selected patient population, increased CVP re-
a combination of reduced perfusion (cardiac index) and mained a determinant of all-cause mortality, independent of
increased venous congestion (CVP) suffer from fluid over- cardiac function. This association with mortality was most
load and decreased organ perfusion, leading, among other prominent in patients with severely increased CVP, even
things, to renal dysfunction. We showed that patients with after adjustment for other baseline characteristics. This
high CVP levels often also have decreased cardiac index and finding was additive to the observation that greater CVP
reduced eGFR. Remarkably, CVP and cardiac index levels predispose to lower eGFR, which may influence
showed an interaction on the relationship with eGFR, with survival by different mechanisms.
an even more pronounced relationship in patients with Clinical implications. Increased CVP frequently is ob-
relatively normal cardiac index. This further strengthens the served in patients with and without reduced cardiac func-
observation of a relationship between CVP and eGFR, tion, comprising almost 20% of patients in the present
which is not exclusively due to reduced cardiac systolic
patient population. Recognition of these patients is essential
function. It also challenges the intuitive notion that fluid
because not only is renal dysfunction much more frequently
overload, although deleterious from a cardiovascular per-
observed, but the risk of mortality also increases with
spective, will invariably be beneficial from the point of view
increasing CVP levels. Treatment to selectively lower CVP
of preservation of renal function. Because our analysis does
may be favorable to reduce symptoms and signs of conges-
not allow to dissect cause and effect relationships, however,
tion, improve GFR, and improve prognosis.
it might also well be that the relatively normal cardiac index
is maintained at the expense of the increased CVP. In that Study limitations. The present study comprises a se-
case, apparently, such a renal hemodynamic profile does not lected patient population that had a specific indication for
translate into better renal function. Our present findings right heart catheterization. Patients undergoing right
seem inconsistent with our earlier findings, showing that heart catheterization are prone to have greater right-
patients with reduced renal perfusion in particular are prone sided filling pressures, and the present observations may
to a detrimental effect of CVP on GFR (11). However, we therefore not represent the general cardiovascular popu-
did not measure renal hemodynamics or renal function by lation. However, this is a large cohort study, with invasive
clearance techniques in the present study, and the popula- cardiac function and CVP measurements across the full
tion was also different. Furthermore, our previous study range. Second, this is a retrospective analysis, and no
consisted of patients with much lower cardiac indexes, all of invasive data are available on renal blood flow and true
which makes a comparison difficult. Nevertheless, this GFR in these patients. Furthermore, it should be ad-
inconsistency needs to be further addressed in future studies. dressed that our study population had very different
New therapeutic agents in the treatment of acute heart catheterization indications, medical history, and medica-
failure that are specifically targeted at improvement of tion regime, all of which could have influenced our
cardiac function and reducing venous congestion recently results. In our study, the presence of heart failure was a
have shown promising results regarding renal function. A clinical diagnosis, rather than related to reduced cardiac
substudy of the SURVIVE (Survival of Patients With Acute index on catheterization. Therefore, the prevalence of
Heart Failure in Need of Intravenous Inotropic Support) heart failure in our study is most likely underestimated.
study, in which levosimendan was compared with dobut- The relationship between increased CVP and renal per-
amine, showed that improvement of renal function was fusion has been observed in heart failure. However our
more pronounced in the group receiving levosimendan, present study is the first to show an independent effect of
despite the obvious positive inotropic effect of dobutamine CVP on renal function. Finally, the retrospective nature
(34). The specific venodilatory effect observed with levosi- of this study and the cross-sectional design limited our
mendan, with subsequent reduced CVP, may be the patho- ability to investigate the cause-effect relationship between
physiologic mechanism, supporting a direct pathophysio- renal impairment and increased CVP, which may actually
logic link between CVP and renal impairment (35). mutually influence each other.
CVP and mortality. Increased CVP and jugular venous
pressure predispose to the development of heart failure in
patients with cardiac dysfunction and have been associated Conclusion
with reduced survival in patients with heart failure
(12,13,36). Small studies have shown that invasive assess- Increased CVP is associated with impaired renal function
ment of CVP is a predictor of cardiovascular outcome in and is independently related to all-cause mortality in a broad
patients with advanced heart failure (14,15). In other spectrum of patients with cardiovascular disease.
588 Damman et al. JACC Vol. 53, No. 7, 2009
Central Venous Pressure and Renal Function February 17, 2009:582–8

19. Doty JM, Saggi BH, Sugerman HJ, et al. Effect of increased renal
Reprint requests and correspondence to: Prof. Dr. Hans L. venous pressure on renal function. J Trauma 1999;47:1000 –3.
Hillege, Hanzeplein 1, 9700 RB Groningen, the Netherlands. 20. Stolarczyk J, Carone FA. Effects of renal lymphatic occlusion and
E-mail: [email protected]. venous constriction on renal function. Am J Pathol 1975;78:285–96.
21. Priebe HJ, Heimann JC, Hedley-Whyte J. Effects of renal and hepatic
venous congestion on renal function in the presence of low and normal
cardiac output in dogs. Circ Res 1980;47:883–90.
REFERENCES 22. Maxwell MH, Breed ES, Schwartz IL. Renal venous pressure in
chronic congestive heart failure. J Clin Invest 1950;29:342– 8.
1. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic 23. Blake WD, Wegria R, Keating RP, Ward HP. Effect of increased
kidney disease and the risks of death, cardiovascular events, and renal venous pressure on renal function. Am J Physiol 1949;157:1–13.
hospitalization. N Engl J Med 2004;351:1296 –305. 24. Kos T, Pacher R, Wimmer A, et al. Relationship between kidney
2. Ruilope LM, van Veldhuisen DJ, Ritz E, Luscher TF. Renal function: function, hemodynamic variables and circulating big endothelin levels
the Cinderella of cardiovascular risk profile. J Am Coll Cardiol in patients with severe refractory heart failure. Wien Klin Wochenschr
2001;38:1782–7. 1998;110:89 –95.
3. Hillege HL, van Gilst WH, van Veldhuisen DJ, et al. Accelerated 25. Palomba H, de Castro I, Neto AL, Lage S, Yu L. Acute kidney injury
decline and prognostic impact of renal function after myocardial prediction following elective cardiac surgery: AKICS score. Kidney Int
infarction and the benefits of ACE inhibition: the CATS randomized 2007;72:624 –31.
trial. Eur Heart J 2003;24:412–20. 26. Bruch C, Rothenburger M, Gotzmann M, et al. Chronic kidney
4. Nag S, Bilous R, Kelly W, Jones S, Roper N, Connolly V. All-cause disease in patients with chronic heart failure—impact on intracardiac
and cardiovascular mortality in diabetic subjects increases significantly conduction, diastolic function and prognosis. Int J Cardiol 2007;118:
with reduced estimated glomerular filtration rate (eGFR): 10 years’ 375– 80.
data from the South Tees Diabetes Mortality study. Diabetes Med 27. Miyazato J, Horio T, Takiuchi S, et al. Left ventricular diastolic
2007;24:10 –7. dysfunction in patients with chronic renal failure: impact of diabetes
5. Rahman M, Pressel S, Davis BR, et al. Renal outcomes in high-risk mellitus. Diabetes Med 2005;22:730 – 6.
hypertensive patients treated with an angiotensin-converting enzyme 28. Ogata C, Horio T, Kamide K, Takiuchi S, Kawano Y. Association
inhibitor or a calcium channel blocker vs a diuretic: a report from the between left ventricular diastolic dysfunction and renal hemodynamic
Antihypertensive and Lipid-Lowering Treatment to Prevent Heart change in patients with treated essential hypertension. Hypertens Res
Attack Trial (ALLHAT). Arch Intern Med 2005;165:936 – 46. 2003;26:971– 8.
6. Reddan DN, Szczech LA, Tuttle RH, et al. Chronic kidney disease, 29. Ahmed A, Rich MW, Sanders PW, et al. Chronic kidney disease
mortality, and treatment strategies among patients with clinically signif- associated mortality in diastolic versus systolic heart failure: a propen-
icant coronary artery disease. J Am Soc Nephrol 2003;14:2373– 80. sity matched study. Am J Cardiol 2007;99:393– 8.
7. Hillege HL, Nitsch D, Pfeffer MA, et al. Renal function as a predictor 30. Bruch C, Reinecke H, Rothenburger M, et al. Transmitral flow
of outcome in a broad spectrum of patients with heart failure. patterns and the presence of chronic kidney disease provide indepen-
Circulation 2006;113:671– 8. dent and incremental prognostic information in patients with heart
8. Ljungman S, Laragh JH, Cody RJ. Role of the Kidney in Congestive failure and systolic dysfunction. J Am Soc Echocardiogr 2007;20:
Heart-Failure - Relationship of Cardiac Index to Kidney-Function. 989 –97.
Drugs 1990;39:10 –21. 31. Guyton AC, Jones CE. Central venous pressure: physiological signif-
9. Stam F, van Guldener C, Schalkwijk CG, ter Wee PM, Donker AJM, icance and clinical implications. Am Heart J 1973;86:431–7.
Stehouwer CDA. Impaired renal function is associated with markers 32. Magder S, Bafaqeeh F. The clinical role of central venous pressure
of endothelial dysfunction and increased inflammatory activity. Neph- measurements. J Intensive Care Med 2007;22:44 –51.
rol Dial Transplant 2003;18:892– 8. 33. Schrier RW. Role of diminished renal function in cardiovascular
10. Zoccali C, Mallamaci F, Tripepi G. Inflammation and atherosclerosis mortality: marker or pathogenetic factor? J Am Coll Cardiol 2006;
in end-stage renal disease. Blood Purif 2003;21:29 –36. 47:1– 8.
11. Damman K, Navis G, Smilde TD, et al. Decreased cardiac output, 34. Yilmaz MB, Yalta K, Yontar C, et al. Levosimendan improves renal
venous congestion and the association with renal impairment in function in patients with acute decompensated heart failure: compar-
patients with cardiac dysfunction. Eur J Heart Fail 2007;9:872– 8. ison with dobutamine. Cardiovasc Drugs Ther 2007;21:431–5.
12. Unverferth DV, Magorien RD, Moeschberger ML, Baker PB, Fetters 35. Damman K, Voors AA. Levosimendan improves renal function in
JK, Leier CV. Factors influencing the one-year mortality of dilated acute decompensated heart failure: cause and clinical application.
cardiomyopathy. Am J Cardiol 1984;54:147–52. Cardiovasc Drugs Ther 2007;21:403– 4.
13. Drazner MH, Rame JE, Stevenson LW, Dries DL. Prognostic 36. Drazner MH, Rame JE, Dries DL. Third heart sound and elevated
importance of elevated jugular venous pressure and a third heart sound jugular venous pressure as markers of the subsequent development of
in patients with heart failure. N Engl J Med 2001;345:574 – 81. heart failure in patients with asymptomatic left ventricular dysfunction.
14. Gardner RS, Henderson G, McDonagh TA. The prognostic use of Am J Med 2003;114:431–7.
right heart catheterization data in patients with advanced heart failure: 37. Khairy P, Fernandes SM, Mayer JE Jr., et al. Long-term survival,
how relevant are invasive procedures in the risk stratification of modes of death, and predictors of mortality in patients with Fontan
advanced heart failure in the era of neurohormones? J Heart Lung surgery. Circulation 2008;117:13–5.
Transplant 2005;24:303–9. 38. Pilcher DV, Scheinkestel CD, Snell GI, Davye-Quinn A, Bailey MJ,
15. Morley D, Brozena SC. Assessing risk by hemodynamic profile in Williams TJ. High central venous pressure is associated with pro-
patients awaiting cardiac transplantation. Am J Cardiol 1994;73: longed mechanical ventilation and increased mortality after lung
379 – 83. transplantation. J Thorac Cardiovasc Surg 2005;129:912– 8.
16. Smilde TD, van Veldhuisen DJ, Navis G, Voors AA, Hillege HL.
Drawbacks and prognostic value of formulas estimating renal function Key Words: renal function y central venous pressure y mortality y
in patients with chronic heart failure and systolic dysfunction. Circu- prognosis y heart catheterization.
lation 2006;114:1572– 80.
APPENDIX
17. Coresh J, Selvin E, Stevens LA, et al. Prevalence of chronic kidney
disease in the United States. JAMA 2007;298:2038 – 47.
18. Harrell FE Jr., Lee KL, Califf RM, Pryor DB, Rosati RA. Regression For supplementary tables on regression analysis for eGFR and multivariate
modelling strategies for improved prognostic prediction. Stat Med Cox regression analysis for all-cause mortality, please see the online
1984;3:143–52. version of this article.