Cardiac output during exercise is related to plasma atrial natriuretic peptide but not to central venous pressure in humans

Abstract

New Findings

• What is the central question of this study?

  Is cardiac output during exercise dependent on central venous pressure?

• What is the main finding and its importance?

  The increase in cardiac output during both rowing and running is related to preload to the heart, as indicated by plasma atrial natriuretic peptide, but unrelated to central venous pressure. The results indicate that in upright humans, central venous pressure reflects the gravitational influence on central venous blood rather than preload to the heart.

We evaluated the increase in cardiac output (CO) during exercise in relationship to central venous pressure (CVP) and plasma arterial natriuretic peptide (ANP) as expressions of preload to the heart. Seven healthy subjects (four men; mean ± SD: age 26 ± 3 years, height 181± 8 cm and weight 76 ± 11 kg;) rested in sitting and standing positions (in randomized order) and then rowed and ran at submaximal workloads. The CVP was recorded, CO (Modelflow) calculated and arterial plasma ANP determined by radioimmunoassay. While sitting, (mean ± SD) CO was 6.2 ± 1.6 l min⁻¹, plasma ANP 70 ± 10 pg ml⁻¹ and CVP 1.8 ± 1.1 mmHg, and when standing decreased to 5.9 ± 1.0 l min⁻¹, 63 ± 10 pg ml⁻¹ and −3.8 ± 1.2 mmHg, respectively (P < 0.05). Ergometer rowing elicited an increase in CO to 22.5 ± 5.5 l min⁻¹ as plasma ANP increased to 156 ± 11 pg ml⁻¹ and CVP to 3.8 ± 0.9 mmHg (P < 0.05). Likewise, CO increased to 23.5 ± 6.0 l min⁻¹ during running, albeit with a smaller (P < 0.05) increase in plasma ANP, but with little change in CVP (−0.9 ± 0.4 mmHg). The increase in CO in response to exercise is related to preload to the heart, as indicated by plasma ANP, but unrelated to CVP. The results indicate that in upright humans, CVP reflects the gravitational influence on central venous blood rather than preload to the heart.

1 INTRODUCTION

When humans stand up, gravity pools blood to dependent regions of the body (Matzen, Perko, Groth, Friedman, & Secher, 1991). Thus, the ‘indifference point’ (where pressure does not change during head-up or head-down tilt) for venous pressure is at the level of the diaphragm (Gauer & Thron, 1965), whereas for volume (Jarvis & Pawelczyk, 2010; Perko et al., 1997) and diameter of the inferior caval vein the indifference point is below the liver (Petersen, Carlsen, Nielsen, Damgaard, & Secher, 2014). Consequently, in an upright posture, central venous pressure (CVP) becomes negative, in parallel with reduction in the central blood volume and, therefore, preload to the heart and cardiac output (CO) decrease (Harms, van Lieshout, Jenstrup, Pott, & Secher, 2003; Matzen et al., 1991). The central blood volume is often taken as the blood within the thoracic cavity as monitored by electrical impedance by Perko et al. (1997) and Jarvis and Pawelczyk (2010). Nonetheless, it is the diastolic filling of the heart that is important for the ability to increase CO, but diastolic filling of the heart cannot be determined during whole-body exercise, such as rowing. However, it is straightforward to determine plasma atrial natriuretic peptide (ANP), which is released by distension of the atria and not related to CVP. Furthermore, plasma ANP decreases when central blood volume is reduced in response to head-up tilt (Matzen, Knigge, Schütten, & Secher, 1990), whole-body heating (Vogelsang et al., 2012) and pressure breathing (Schütten et al., 1987), whereas CVP is reduced in the first two conditions and elevated in the latter. This disparity in CVP and ANP responses to central blood volume changes suggests that CVP might be less sensitive compared with plasma ANP in tracking changes in central blood volume. Thus, the influence of CVP on preload to the heart during whole-body exercise can be evaluated in relationship to concomitant changes in plasma ANP.

During exercise, leg muscles pump blood towards the heart (Beecher, Field, & Krogh, 1936; Rowell, 1993), as illustrated by increased plasma ANP (Vogelsang et al., 2006). We hypothesized that even though orthostasis reduces CVP, nevertheless, the enhanced cardiac preload, as reflected by plasma ANP, supports the increase in CO. A comparison was made between running and rowing because the seated position of rowing was considered to be associated with only a small gravitational reduction in CVP, thereby allowing for evaluation of CVP on the increase in CO during exercise.

2 METHODS

2.1 Ethical approval

Seven healthy subjects (four men; mean ± SD: age 26 ± 3 years, height 181 ± 8 cm and weight 76 ± 11 kg), after informed oral and written consent, volunteered to participate in the study as approved by the Ethical Committee of Copenhagen (KF 01-186/2) and conforming to the Declaration of Helsinki, except for registration in a database. All subjects were recruited from a local club and, as part of their training, were familiarized with ergometer rowing over several years.

2.2 Experimental design and procedures

After instrumentation, the subjects rested supine for ∼15 min and then performed, in random order, rowing on an ergometer (Concept II, Morrisville, VT, USA) and walking followed by running on a treadmill (Runrace, Technogym, Gambettola, Italy). The two types of exercise were at three intensities aiming at a heart rate (HR) of 120, 140 and 160 beats min⁻¹ and lasted ∼7 min each.

2.3 Oxygen uptake and pulmonary ventilation

Pulmonary ventilation () and oxygen uptake () were determined by an Oxyscreen metabolic cart (CPX/D; Medical Graphics, St Paul, MN, USA), with values reported as the average over 15 s for the last 1 min of observation.

2.4 Cardiovascular variables

The CVP was measured with a catheter advanced to the right atrium after cannulation of an arm vein, and values were averaged over several heartbeats. Mean arterial pressure (MAP) was determined from the radial artery of the non-dominant arm. Both catheters were connected to transducers (Baxter Healthcare Corporation, Irvine, CA, USA) positioned on the back at the level of the mammary papilla, with infusion of isotonic saline (3 ml h⁻¹) to avoid clot formation.

A modified Modelflow method estimated CO from the radial arterial pressure (Finapres Medical Systems, Amsterdam, The Netherlands) to yield stroke volume, which when multiplied by HR provides an estimate of CO (Wesselling, De Wit, Weber, & Smith, 1983), and values were adjusted to CO values calculated by the Fick principle. Heart rate was recorded from a three-lead ECG, with electrodes (Medicotest Q-10-A, Copenhagen, Denmark) placed on the sternum and the cervical vertebrae to minimize noise from muscles. All data were collected in the last 1 min of exercise and, likewise, in the last 1 min of rest.

2.5 Blood samples

Blood was sampled from the radial artery and the central venous catheter in tubes prepared with heparin and analysed for pH, oxygen saturation () and tension () (ABL, Radiometer, Copenhagen, Denmark). Also, arterial samples (in EDTA tubes) were centrifuged and plasma kept at −80°C until analysed for ANP by radioimmunoassay (Schütten et al., 1987). The ANP analysis has a sensitivity of 3.1 pg ml⁻¹, and the intra- and interassay coefficients of variation were 4 and 5%, respectively.

2.6 Statistical analysis

Data are presented as means ± SD, and comparisons across exercise mode and intensity were evaluated by two-way ANOVA with Newman–Keuls post hoc validation. Relationships were evaluated separately for each type of exercise by linear and logarithmic regressions, and a value of P < 0.05 was considered statistically significant.

3 RESULTS

From the sitting to the standing position, HR increased, while MAP, CO, , and did not change significantly (Table 1). In contrast, CVP, plasma ANP and venous decreased according to the postural change (P < 0.05).

Table 1. Variables (means ± SD) at seated and standing rest and during rowing and running at three levels of targeted heart rate (HR)

	Rest		HR 120 beats min^−1		HR 140 beats min^−1		HR 160 beats min^−1
	Sitting	Standing	Rowing	Running	Rowing	Running	Rowing	Running
HR (beats min^−1)	79 ± 8	90 ± 7	115 ± 13‡	117 ± 13	138 ± 17‡	139 ± 18†	159 ± 15‡	158 ± 16†
MAP (mmHg)	92 ± 6	91 ± 5	97 ± 4	93 ± 4	97 ± 3	93 ± 3	96 ± 4	92 ± 4
CVP (mmHg)	1.8 ± 1.1	−3.8 ± 1.2	2.7 ± 0.9*	−5.4 ± 1.2	3.1 ± 1.1*	−5.4 ± 0.9	3.8 ± 0.9*	−4.7 ± 0.8
ANP (pg ml^−1)	70 ± 10	63 ± 10	107 ± 9*‡	76 ± 10†	135 ± 10*‡	97 ± 11†	156 ± 11*‡	104 ± 11†
CO (l min^−1)	5.0 ± 1.3	5.2 ± 0.9	18.1 ± 6.4‡	16.4 ± 3.9†	20.4 ± 6.4‡	22.6 ± 7.0†	22.5 ± 5.5‡	23.5 ± 6.0†
(l min^−1)	0.41 ± 0.1	0.47 ± 0.1	1.98 ± 0.7*	1.65 ± 0.5†	2.52 ± 0.8*	2.36 ± 0.8†	2.98 ± 0.8*	2.80 ± 0.7†
(l min^−1)	11.6 ± 3.2	12.4 ± 5.6	36.9 ± 10.0*	33.8 ± 10.1†	53.7 ± 10.8*	49.2 ± 10.2†	68.9 ± 9.8*	62.1 ± 10.2†
Hb [g (100 ml)^−1]
Arterial	13.9 ± 1.5	13.8 ± 1.6	14.1 ± 1.4	13.8 ± 1.7	14.4 ± 1.6	14.1 ± 1.4	14.7 ± 2.7	14.0 ± 1.7
Venous	13.3 ± 2.7	13.0 ± 3.8	13.8 ± 1.7	13.1 ± 1.9	14.3 ± 2.7	13.3 ± 1.1	14.4 ± 2.7	13.3 ± 23.4
pH
Arterial	7.42 ± 0.04	7.44 ± 0.02	7.41 ± 0.03	7.42 ± 0.01	7.40 ± 0.03	7.42 ± 0.02	7.40 ± 0.03	7.40 ± 0.04
Venous	7.39 ± 0.01	7.40 ± 0.01	7.38 ± 0.10	7.38 ± 0.01	7.36 ± 0.01	7.38 ± 0.01	7.32 ± 0.05*‡	7.38 ± 0.03
(mmHg)
Arterial	106 ± 10	102 ± 11	94 ± 9	97 ± 3	93 ± 10	94 ± 7	90 ± 10‡	93 ± 9
Venous	36 ± 6	32 ± 3	27 ± 7	30 ± 6	24 ± 1‡	30 ± 6	23 ± 2‡	26 ± 3†
(%)
Arterial	98.9 ± 0.6	98.7 ± 1.0	98.0 ± 1.1	98.5 ± 0.2	97.6 ± 1.5	98.1 ± 0.7	97.3 ± 1.8	97.9 ± 1.0
Venous	60.1 ± 13.3	56.4 ± 6.1	44.1 ± 13.6	49.7 ± 13.6	37.3 ± 8.8‡	48.6 ± 11.3	34.3 ± 8.3‡	39.6 ± 3.5†

• Abbreviations: ANP, atrial natriuretic peptide; CVP, central venous pressure; Hb, blood haemoglobin; MAP, mean arterial pressure; , partial pressure of oxygen; , haemoglobin oxygen saturation; , pulmonary ventilation; and , oxygen uptake.
• ^*Different from running (P < 0.05)
• ^†Different from standing (P < 0.05).
• ^‡Different from sitting rest (P < 0.05).

3.1 Running

From standing rest to running, HR, CO, and increased, while MAP was maintained, CVP remained negative and plasma ANP increased (Table 1; P < 0.05). Venous and decreased with exercise intensity (P < 0.05), with no significant change in blood pH. The increase in CO was curvilinear related to the plasma ANP (r² = 0.94, P < 0.05), while the correlation to CVP if anything was negative, and similar correlations were established when and blood gas variables were were applied to calculate CO (for plasma ANP, r² = 0.94, P < 0.05; and for CVP, r² = 0.54, P < 0.05; Figure 1).

Figure 1

Logarithmic regressions between cardiac output (CO) and plasma atrial natriuretic peptide (ANP) (top panel) and linear regressions between CO and central venous pressure (CVP) (bottom panel) at rest and during three exercise intensities for running and rowing (n = 7, P < 0.05)

3.2 Rowing

Similar to running, HR, CO, and increased with rowing intensity, but the increases were more pronounced (Table 1). The MAP did not increase significantly, but in contrast to running, both CVP and plasma ANP increased, while venous and decreased with exercise intensity (P < 0.05) and there was a small decrease in blood pH. Also, similar to running, CO was related to plasma ANP during rowing (r² = 0.95, P < 0.05) and in that situation, also to CVP (r² = 0.89, P < 0.05), as confirmed when CO was calculated based on Fick's principle (for plasma ANP, r² = 0.95, P < 0.05; and for CVP, r² = 0.88, P < 0.05; Figure 1). Taking the observations during rowing and running together, there was no correlation between CO and CVP (r² = 0.17), whereas CO was correlated with plasma ANP (r² = 0.94, P < 0.05).

4 DISCUSSION

Accepting the gravitational influence on CVP and considering that imaging of the heart during whole-body exercise is challenging, in the present study we used plasma ANP to indicate filling of the heart. We found that exercise enhances cardiac preload, as evaluated by plasma ANP, in accordance with results by Nicolai and Zuntz (1914), who carried out such evaluation during treadmill walking using radiography, and confirming the findings by Vogelsang et al. (2006).

In the standing posture, CVP decreased to −3.8 mmHg, corresponding to a column of blood in the inferior caval vein of some 5 cm, reflecting the distance to the pressure indifference point probably positioned at the level of the diaphragm (Gauer & Thron, 1965). In the abdominal cavity, venous pressure could be influenced by the abdominal muscles pressing on the organs and the caval vein. Thus, in the open ‘central’ veins, the influence of gravity on pressure does not extent all the way to the ‘first’ venous valve at end of the femoral vein. Conversely, CVP did not become negative during sitting on the ergometer, probably because the subjects were in a position that would press the legs against the abdomen and thereby apply pressure to the caval vein. In the same way, we consider it likely that during rowing the pressure applied by the legs against the abdomen would contribute to the increase in CVP (Pott et al., 1997). Also, during rowing the stroke is accomplished by stabilizing the trunk by a ‘Valsalva-like manoeuvre’, and thus, fluctuations in blood pressure are more related to the rhythm of rowing than to the function of the heart (Clifford, Hanel, & Secher, 1994). Even though, in the present study, CVP data were not synchronized with each rowing stroke, we can speculate that CVP fluctuated markedly within one rowing cycle and reached a maximal value during the stroke as influenced by the entrained respiration (Mahler, Shuhart, Brew, & Stukel, 1991; Pott et al., 1997; Steinacker, Both, & Whipp, 1993). In contrast, CVP demonstrated a small further decrease during running from the value established during standing, and we can only speculate that this decrease was a manifestation of low abdominal pressure on the inferior caval vein, but we lack data on intra-abdominal venous pressure during running and rowing.

The increase in preload to the heart, as indicated by plasma ANP and increased CO during exercise, manifested despite the fact that CVP decreased during running and increased during rowing, which indicates that CVP does not dictate filling of the heart. The relationship between CO and plasma ANP illustrates the importance of the Starling ‘law of the heart’ for cardiovascular control, but it should be considered that during exercise there is sympathetic influence on the heart, as illustrated by an attenuated CO response to exercise after β-adrenergic blockade (Pawelczyk, Hanel, Pawelczyk, Warberg, & Secher, 1992). Taken together, CVP is not what drives blood into the ventricles, but filling of the ventricles depends on the amount of blood in the atria and nearby veins that is advanced to the ventricles by ‘suction’ during the diastole as the heart is ‘untwisting’ (Nakatani, 2011; https://www.youtube.com/watch?v=N6ORMHi9rcU). Consequently, CVP cannot be used for monitoring the central blood volume, except in heart failure patients whose hearts have lost the ability to ‘twist’.

The data indicate that a larger preload is needed for a given CO during rowing than during running, although MAP did not increase (and HR was matched) and therefore, on average, afterload to the heart was the same during the two exercise interventions. Unfortunately, stroke volume was not recorded in alignment with the stroke of rowing, but we speculate that the Valsalva-like manoeuvre associated with rowing hindered the increase in CO during approximately half of the stroke. Thus, CO during rowing is likely to depend on ventricular filling in the part of the stroke where the rower did not hold his breath.

As expected when the workload for rowing and running was matched to a given HR, (and ) was larger during rowing than during running given the orthostatic increase in HR as the central blood volume decreases (e.g. Matzen et al., 1991), as indicated by the reduction in plasma ANP and the reduction in venous oxygenation. During running, the maximal HR can be calculated as 208 − 0.7 × age of the subjects (Tanaka, Monahan, & Seals, 2001), i.e. ∼190 beats min⁻¹ for the present subjects, while for rowing the maximal HR is only a little more than 180 beats min⁻¹ (e.g. Volianitis et al., 2008), in accordance with a lower HR in supine than upright positions (Stenberg, Astrand, Ekblom, Royce, & Saltin, 1967). Accordingly, a HR of 160 beats min⁻¹ would correspond to approximately a 90% effort during rowing, but only an 84% effort during running, and the fact that the relative workload was larger during rowing than during running is illustrated by the finding that only during rowing did blood pH decrease; however, a maximal effort and rating of perceived effort were not included in the protocol.

Limitations to the study include that plasma ANP indicates filling of mainly the right atrium and not diastolic filling of the (left) ventricle, but in the present study it was not possible to make an (e.g. echocardiographic) evaluation of (left) ventricular filling during the two modes of exercise. Also, we accept that determination of CO could have been made by more conventional methods, such as by thermodilution after placement of a pulmonary artery catheter or by following the concentration in blood of Indocyanine Green or technetium, which we have used in previous evaluations of CO during exercise (Pawelczyk et al., 1992; Secher, Clausen, Klausen, Noer, & Trap-Jensen, 1977). Also, the study did not address whether there is a sex difference in the plasma ANP response to exercise, although women are reported to demonstrate a level that is about twice as high as men (Clark, Elahi, & Epstein, 1990).

In conclusion, the present results suggest that the ‘muscle pump’ contributes significantly to enhance CO during exercise, as indicated by plasma ANP for evaluation of preload to the heart. In contrast, in an upright posture CVP can both increase and decrease in response to exercise, and the deviations are unrelated to the gradually increasing plasma ANP concentration with exercise intensity. We consider these findings to reflect on the evaluation of circulation during exercise. Often, total peripheral resistance during exercise is expressed by subtracting CVP from MAP, even though the inclusion of a pump (i.e. muscle pump) while expressing a resistance seems unjustified from a physics perspective. The present findings suggest that in upright humans CVP is not a ‘resistance’ to venous return but a reflection of the gravitational influence on venous blood and independent of how much blood is present in the veins, as long as they stay open, which is the definition of ‘central’ veins.

ACKNOWLEDGEMENTS

We thank Elsa Larsen for technical help in determination of plasma ANP.

AUTHOR CONTRIBUTIONS

All authors contributed to conception and design of experiments, collection and assembly of data, data analysis and interpretation and manuscript writing. All authors approved the final version of the manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed. Research governance was by N.H.S.

COMPETING INTERESTS

None declared.