Int J Sports Med 2011; 32(6): 438-445
DOI: 10.1055/s-0031-1271788
Training & Testing

© Georg Thieme Verlag KG Stuttgart · New York

Cardiovascular Responses During Hypoventilation at Exercise

X. Woorons1 , 2 , N. Bourdillon1 , C. Lamberto1 , 3 , H. Vandewalle3 , J.-P. Richalet1 , 3 , P. Mollard1 , A. Pichon1
  • 1Université Paris 13, Laboratoire ‘Réponses cellulaires et fonctionnelles à l'hypoxie’, EA 2363, UFR-SMBH, Bobigny, France
  • 2Association pour la Recherche et la Promotion de l'Entraînement en Hypoventilation (ARPEH), Lille, France
  • 3AP-HP, Hôpital Avicenne, Bobigny, France
Further Information
#

Correspondence

Dr. Aurelien Pichon

Laboratoire ‘Réponses

cellulaires et fonctonnelles à

l'hypoxie’

UFR SMBH

Université Paris 13

74 rue Marcel Cachin

93017 Bobigny

France

Phone: +33/14/8387/632

Fax: +33/14/8388/924

Email: aurelien.pichon@orange.fr

Publication History

accepted after revision January 25, 2011

Publication Date:
11 May 2011 (eFirst)

Table of Contents #

Abstract

This study aimed to determine the cardiovascular responses during a prolonged exercise with voluntary hypoventilation (VH). 7 men performed 3 series of 5-min exercise at 65% of normoxic maximal O2 uptake under 3 conditions: (1) normal breathing (NB) in normoxia (NB0.21), (2) VH in normoxia (VH0.21), (3) NB in hypoxia (NB0.157, inspired oxygen fraction=0.157). In both VH0.21 and NB0.157, there was a similar drop in arterial oxygen saturation and arterial O2 content (CaO2) which were lower than in NB0.21. Heart rate (HR), stroke volume, and cardiac output (–) were higher in VH0.21 than in NB0.21 during most parts of exercise whereas there was no difference between NB0.157 and VH0.21 or NB0.21. HR variability analysis suggested an increased sympathetic modulation in VH0.21 only. O2 transport and oxygen uptake were generally not different between interventions. Mixed venous O2 content (C–O2) was lower in NB0.157 than in both VH0.21 and NB0.21 and not different between the latter. CaO2–C–O2 was not different between NB0.157 and NB0.21 but lower in VH0.21. This study shows that a prolonged exercise with VH leads to a greater cardiac activity, independent from the hypoxic effect. The greater – in VH compared to normal breathing seems to be the main factor for compensating the drop of arterial oxygen content.

#

Introduction

Recent studies have reported that it could be interesting for athletes to exercise with voluntary hypoventilation (VH) during their training. This kind of exercise when performed at low pulmonary volumes induces a severe arterial desaturation [38] [39] as well as muscular deoxygenation [40]. Therefore, it may provide a good alternative for athletes who would like to perform an intermittent hypoxic training (IHT) but cannot afford to go to altitude or to use expensive devices to simulate hypoxia [17]. However, VH exercise is different from a classical exercise in hypoxia since it also induces hypercapnia. The hypercapnic response is important because it leads to respiratory acidosis and induces a right shift of the oxygen dissociation curve. Therefore, VH-induced hypercapnia fully participates in the arterial desaturation [38] [40]. Moreover, the respiratory acidosis likely impacts on the acid-base balance at the muscular level. This phenomenon may be at the origin of adaptations leading to a delayed acidosis, as reported in the only study that investigated the effects of training with VH [39]. Thus, this kind of training could be interesting for short duration and intense activities. Besides, following VH training, a tendency to an increased velocity at maximal exhaustion (0.4 km/h on average) has been found during an incremental test on a treadmill [39]. Furthermore, exercise with VH increases blood lactate concentration which could help improve or at least maintain the power of glycolytic metabolism [40].

The effects of VH exercise on respiratory and blood parameters have been well investigated [10] [15] [16] [29] [31] [38] [41] [42]. Recently, a study also focused on the effects of VH on muscular oxygenation and lactate concentration [40]. On the other hand, little is known about the cardiovascular adjustments during this type of exercise. So far, no study has ever focused on the effects of VH exercise on cardiac output (– ) or stroke volume (SV). Some studies have investigated the effects on heart rate (HR) but most of them did not find a greater increase during exercise with VH than with normal breathing [10] [15] [31] [41]. However, in these studies, a significant arterial desaturation did not occur during exercise or if so the duration of VH was probably too short (≤5 min) to significantly increase HR.

The measure of – turns out to be essential because it could increase if submaximal exercise with VH was prolonged and led to severe hypoxemia. First, for a same absolute power, the relative intensity is higher during VH exercise since maximal O2 uptake (V˙O2max) and the associated power output (PO) decrease with arterial desaturation [6] [18] [37]. A higher relative intensity should elevate HR and maybe – . Furthermore, an increase in – may compensate, at least partially, the drop of arterial oxygen saturation (SaO2) and oxygen content (CaO2). Second, sympathetic activity may increase during exercise with VH under both the hypoxic and hypercapnic effect. Indeed, hypoxia induces greater plasma catecholamine concentrations at rest and at exercise [13] [27]. In addition, both hypercapnia and the subsequent acidosis increase the activation of the adrenergic system which induces an elevated catecholamine concentration as well [12] [20]. Finally, hypercapnia and the respiratory acidosis also lead to hyperventilation during the recovery periods following VH. This may have an effect on the cardiac activity through a greater work of respiratory muscles.

The goal of the present study was to determine the cardiovascular responses, and especially the changes in HR, SV and – , during a prolonged exercise (up to 18 min) with VH. In order to determine the role of the hypoxic effect, we also aimed at comparing these responses with an exercise carried out in intermittent hypoxia inducing the same arterial desaturation kinetics as during exercise with VH. We hypothesized that both VH and hypoxic exercise should induce a higher HR, and subsequently – , than during exercise in normal conditions.

#

Methods

#

Subjects

7 men, with no sign or history of cardiovascular or respiratory disease, participated in this study. All men were physically active practising a recreational sport for 3–4 h a week. Subjects characteristics (mean±SD) were: age 29.8±6.5 years, height 177.6±6.5 cm, weight 74.9±9.8 kg and V˙O2max 53.5±3.9 mL.kg−1 min−1. They were informed about the nature, the conditions and the risks of the experiment and gave their written informed consent. All the procedures were approved by the ethical committee Ile de France II, Paris, France and the study has been performed in accordance with the Ethical Standards in Sport and Exercise Science Research [14].

#

Protocol

Before the start of the experiment, clinical interviews and examinations were carried out. The subjects then performed a maximal exercise test on an electrically braked cycle ergometer (Jaeger ER 900, Wuerzburg, Germany) to assess V˙O2max and maximal PO. After a 3-min rest, the test began at an intensity of 30 W. The workload was then increased by 30 W every 2 min until exhaustion. Subjects were verbally encouraged to continue the exercise for as long as possible. After a resting period of about 1 h following the maximal test, the subjects performed a test to become familiarized with the hypoventilation technique. During this test, they had to cycle at 50% V˙O2max with VH carried out at or near the functional residual capacity (FRC). This respiratory technique has been well described in a recent study [40]. Briefly, it consisted of inspiring every 4 s instead of about 1 s in normal conditions. We used the pedalling frequency which was established at 75 rpm to precisely control the respiratory technique. After each inspiration (/ or 1 rev), the subjects had to make a normal expiration (1 or 1.5 rev) in order to target FRC and then had to hold their breath (3.5–4 rev) until the next inspiration. During the familiarization test, the transcutaneous arterial O2 saturation (SpO2) was measured by an ear pulse oximeter (Ohmeda Biox 3 740, Louisville, Colorado, USA).

One week after the preliminary procedures, each subject had to perform 3 cycle exercises (on the same ergometer as previously) in a randomized order and under the following conditions: (1) normal breathing (NB) in normoxia (NB0.21), (2) VH in normoxia (VH0.21), (3) NB in hypoxia (NB0.157). The degree of hypoxia was individually determined on the basis of the SpO2 reached during the familiarization test with VH. Thus, the parameters in these 2 interventions were compared at a same or quite similar SpO2. The average FIO2 (mean±SD) was 0.157±0.01.

In each condition, the exercise consisted in performing three 5-min series (S1, S2, S3) at 65% of normoxic V˙O2max. Each series was divided into 5 periods of 1 min including 15 s in NB0.21 followed by 45 s in one of the previous conditions. Between series, transition periods in NB0.21 (T1, T2, T3) were included. During these transitions, subjects kept on cycling for 1 min at the same exercise intensity.

It is important to intersperse periods with normal breathing during exercise with VH to decrease the level of arterial carbon dioxide pressure (PaCO2). Indeed, hypercapnia generates discomfort and especially cerebral vasodilatation, inducing side effects such as headaches. Furthermore, athletes who would like to carry out VH training should organise their training sessions in a similar way [39].

The total duration of exercise was then 18 min. We consider that when using VH, this duration constitutes a prolonged exercise regarding the difficulty to carry out this respiratory technique. While 18 min represents a rather short duration for light or moderate exercise, this is not so for higher intensities or harder exercise like VH. Furthermore, so far most studies dealing with VH have used an exercise duration not exceeding 5 min.

Before the start of series, the subjects remained seated for 5 min at rest on the cycle ergometer and then warmed up for 3 min at an exercise intensity corresponding to 40% of V˙O2max. The first 2-min of the warming up was performed in NB0.21 and the 3rd minute in the same condition as in the 3 series to be carried out. We did it this way especially because in VH0.21 the subjects had to prepare themselves and adjust their breathing to the pedalling frequency before beginning the series. The 3 tests were carried out in 2 different sessions. Both sessions were separated by 72 h and there was a 4-h rest between the 2 tests performed on the same day. During VH0.21, the subjects were coached all the time in order to carry out the hypoventilation technique as adequately as possible.

Conditions of normobaric hypoxia were created using a gas mixing device (AltiTrainer200 ®, S.M. TEC, Geneva, Switzerland) which was connected to a N2 gas bottle. Throughout the test, N2 gas was manually added to ambient air with a 3 way gauge and with a little dead space. Thus, the transitions in gas concentration took only a few seconds to obtain the desired FIO2. Both gases were added to ambient air to obtain the desired FIO2. The gas mixture was stocked in a buffer tank (30 L) before being inhaled by the subjects. Inspired O2 pressure was continuously monitored throughout the tests using an oxygen probe, located in the buffer tank (electrochemical O2 probe MOX3, City Technology, Portsmouth, UK). According to the manufacturer, the maximal difference between the PO2 measured by the AltiTrainer200 ® O2 probe and the PO2 calculated from the O2 fraction measured by an external probe (Servomex 720A, Geneva, Switzerland) is less than 1 mmHg over the whole range of PO2 (150–69 mmHg). The device is reliable for ventilation less than 200 L/min.

#

Measurements

#

Respiratory variables

The ventilatory and gas exchange variables were continuously measured using a breath-by-breath automated exercise metabolic system. The subjects breathed through a rigid mouthpiece connected to a “Y” system fixation with a double valve which ensures anti return (Jaeger, Germany). An inspiratory valve, connected to the AltiTrainer200 ® allowed the subject to inhale the hypoxic mixture. Expired gases were collected into a metabograph (Oxycon, Jeager, Germany) to measure oxygen consumption (V˙O2), expired ventilation (V˙E), end-tidal carbon dioxide pressure (PETCO2), and to calculate end-tidal O2 pressure (PETO2).

#

Arterial oxygen saturation

SpO2 was measured with the same ear pulse oximeter as during the familiarization test. This device is resistant to the effects of subject motion [5] and its accuracy has been demonstrated for SaO2 values above 75% [33]. At each test, before attaching the ear clip, the earlobe was massaged vigorously and pre-warmed with vasodilating capsaicin cream to increase perfusion.

#

Blood pressure

The arterial pressure was measured with Dinamap 1 846 SX (Critikon, Tampa, USA), at rest and at the fourth minute of each series.

#

Cardiac output

HR was measured continuously using electrocardiography. SV and – were continuously measured using a non invasive impedance cardiograph device, the PhysioFlow PF-05 (Manatec Biomedical, Paris, France). This bioimpedance method uses changes in thoracic impedance during cardiac ejection to calculate SV. The PhysioFlow concept and methodology have been validated at rest and at exercise [8] and during a maximal progressive exercise [28] [30] [35]. 6 electrodes were placed, 2 for an electrocardiography measurement and 4 “impedance” electrodes placed at the base of the neck and on the processus xiphoideus. Before placing these electrodes, the skin was slightly scraped with an abrasive sponge and cleaned with alcohol. After placing the electrodes and carefully fixing the thread with an adhesive band in order to prevent any movement, the subjects took their sitting position on the ergocycle and kept still and silent during all the calibration process. Since arterial pressure is necessary to start the calibration process, it was measured at rest just before the beginning of exercise (Dinamap 1 846 SX, Critikon, Tampa, USA). We watched over the signal quality and stability during the entire test. A new calibration procedure at rest was performed before each test.

The breath-by-breath measurements, SpO2 and the cardiac parameters were averaged over 15-s intervals and data analysed at the end of each series (S1, S2, S3) and each transition period (T1, T2, T3).

#

Autonomic nervous system activity

RR intervals were recorded using a heart rate monitor with a sampling rate of 1 000 Hz (Suunto t6, Suunto Oy, Vantaa, Finland). Data were analysed during the whole S3 and transferred to a PC computer via a USB link. The software enabled the visualization of heart rate and the extraction of a cardiac period (RR interval) file in ASCII format for heart rate variability (HRV) analysis. RR intervals were examined and artefactual data were replaced using interpolation from the 5 previous normal-to-normal (NN) intervals. Suitable NN periods were then selected for analysis. According to a previous study [25], series of 256 consecutive NN intervals were extracted and resampled at 4 Hz. Because of the effect of ventilation on HRV frequency-domain analysis during high intensity exercise [24] [26], we chose to perform nonlinear analyses (HRV analysis software, Kuopio, Finland). The nonlinear properties of HRV were analysed using measures of Poincaré plot [34] and detrended fluctuation analysis (DFA) [23]. For the Poincaré plot analysis, we calculated SD1 and SD2 which describe short-term (parasympathetic modulation) and long-term (parasympathetic and sympathetic modulations) variability, respectively [43]. Typically in DFA, the correlations are divided into short-term (α1) and long-term fluctuations (α2). The short-term fluctuations are characterized by the slope α1 obtained from the (log n, log F(n)) graph within range 4≤n≤16. Correspondingly, the slope α2 obtained from the range 16≤n≤64 characterizes long-term fluctuations.

#

Arterialized blood measurements

An arterialized blood sample (95 μL) was taken from the left earlobe of the subjects in the last 15 s of S1, T2 and S3 to obtain arterial oxygen pressure (PaO2), PaCO2, hemoglobin concentration ([Hb]), pH and P50. The earlobe was prewarmed with a vasodilating capsaicin cream and the arterialized measurements were made by a well-trained technician in order to improve their accuracy. The earlobe samples were collected into a capillary tube and immediately analysed (Radiometer ABL 700, Copenhagen, Denmark). The accuracy of this technique and its validity when compared to arterial blood samples has been recently established [19].

#

Calculated parameters

The following parameters were calculated at the end of each series and each transition period from some of the previous measurements:

  • CaO2 (mL/L)=[Hb]×1.34×10×SpO2 /100 with [Hb] in g/dL and SpO2 in %. We used [Hb] measured at S1 for the calculation of CaO2 at S1 and T1, [Hb] at T2 for CaO2 at S2 and T2 and [Hb] at S3 for CaO2 at S3 and T3. The content of dissolved O2 was neglected.

  • O2 transport (L/min): Q˙aO2=Q˙ ×CaO2 /1 000.

  • Difference in arterial-venous O2 content (mL/L): CaO2-CVO2=(V˙O2 /Q˙)×1 000 with V˙O2 and Q˙ in L/min.

  • Mixed venous O2 content (mL/L): CVO2=CaO2 – (CaO2-CVO2)

  • Tissue O2 extraction (%): O2ER=V˙O2 /Q˙aO2.

#

Statistics

To determine the differences between exercise interventions at the end of each series and recovery period, we performed 2-way analysis of variance (ANOVA) separating S1, S2 and S3 from T1, T2 and T3. For each intervention, we also performed a 1-way ANOVA for repeated measures to assess whether there was a difference between series and transition periods. In both statistical tests, when a significant main effect was found, the Bonferroni post hoc test was used. We expressed the results as mean±SD in tables and mean±SE in figures for more clarity. The level of significance was set at P<0.05.

#

Results

#

PO

During the maximal test, the mean maximal PO of the subjects was 285.9±11.6 watts. The mean PO during exercise at 65% V˙O2max was 185.9±7.5 watts.

#

Respiratory variables

The results are presented in [Fig. 1]. V˙O2 was not different between interventions at the end of each series. On the other hand, at the end of each transition period, V˙O2 was higher in both VH0.21 and NB0.157 than in NB0.21 and was not different between VH0.21 and NB0.157. At S1, S2 and S3, PETO2 in VH0.21 was lower than in NB0.21 and higher than in NB0.157. At T1, T2 and T3, PETO2 was higher in VH0.21 than in both NB0.21 and NB0.157. V˙E was lower at the end of each series and higher at the end of each transition period in VH0.21 than in both NB0.157 and NB0.21. PETCO2 in VH0.21 was higher at the end of each series and not different at the end of each transition period than PETCO2 in both NB0.21 and NB0.157.

Zoom Image

Fig. 1 Respiratory variables during exercise with normal breathing in normoxia (NB0.21, filled circle), voluntary hypoventilation in normoxia (VH0.21, filled square), normal breathing in hypoxia (NB0.157, open square). V˙O2, oxygen consumption; V˙E expired ventilation; PETO2, end-tidal O2 pressure, PETCO2, end-tidal carbon dioxide pressure. S1, S2, S3, series 1, 2 and 3; T1, T2, T3, transition 1, 2 and 3. Significantly different from *NB0.21, †NB0.157, (a) previous series. Values are mean±SE; P<0.05.

In VH0.21, V˙O2, V˙E and PETO2 were higher and PETCO2 lower at the end of each transition period than in the previous series. In NB0.157, V˙O2, PETO2 and PETCO2 were higher and V˙E lower at T1, T2 and T3 than at the previous series. There was no difference in V˙O2, PETO2, PETCO2 or V˙E between series and transition periods in NB0.21.

#

Cardiac variables

The results are presented in [Fig. 2]. HR, SV and – were higher in VH0.21 than in NB0.21 during most part of the exercise. At the end of the transition periods, HR (T2 and T3) and – (T1 and T2) were higher in VH0.21 than in NB0.157 whereas there was no difference at the end of each series. SV was not different between VH0.21 and NB0.157 throughout exercise. The results of ANOVA on HR for S1, S2 and S3 revealed a main effect of NB0.157 as compared to NB0.21. However, the post hoc test did not show any difference between both interventions whatever the series.

Zoom Image

Fig. 2 Cardiovascular variables during exercise with normal breathing in normoxia (NB0.21, filled circle), voluntary hypoventilation in normoxia (VH0.21, filled square), normal breathing in hypoxia (NB0.157, open square). HR, heart rate; SV, stroke volume; –, cardiac output; S1, S2, S3, series 1, 2 and 3; T1, T2, T3, transition 1, 2 and 3. Significantly different from *NB0.21, †NB0.157, (a) previous series. Values are mean±SE; P<0.05.

In VH0.21, HR was higher at T1, T2 and T3 and – higher at T1 than at the previous series whereas there was no difference in SV. In NB0.21 and NB0.157, HR, SV and – were not different at the end of each transition period compared to the previous series.

#

Autonomic nervous system activity

There was no difference in SD1 and α1 between interventions ([Table 1] ). On the other hand, SD2 and α2 where higher in VH0.21 than in both NB0.21 and NB0.157 but were not different between the latter.

Table 1 Indices of heart rate variability (HRV) at S3.

NB0.21

VH0.21

NB0.157

SD1 (ms)

2.3±0.6

3.0±0.8

2.5±0.7

SD2 (ms)

6.0±0.9

9.7±3.3*

7.1±2.5

α1

0.63±0.19

0.63±0.20

0.50±0.13

α2

1.26±0.14

1.6±0.08*

1.18±0.32

Values are mean±SD. Indices were calculated from non linear analyses during the third series (S3) of exercise with normal breathing in normoxia (NB0.21), voluntary hypoventilation in normoxia (VH0.21) and normal breathing in hypoxia (NB0.157). SD1: short-term RR Poincaré plot variability; SD2, long-term RR Poincaré plot variability; α1: short term RR variability from detrend fluctuation analysis; α2: long-term RR fluctuations from detrend fluctuation analysis. *, significantly different from NB0.21, , significantly different from NB0.157, P<0.05

#

Arterialized blood measurements

The results are presented in [Table 2] . At T2, there was no difference in all parameters between interventions except for P50 which was higher in VH0.21 than in NB0.21.

Table 2 Arterialized blood measurements.

S1

T2

S3

PaO2 (mmHg)

NB0.21

84.4±5.2

80.7±5.9

80.1±6.6

VH0.21

63.3±5.5*

82.1±3.0

65.5±8.4*

NB0.157

61.0±4.6*

82.4±4.6

55.7±7.1*

PaCO2 (mmHg)

NB0.21

40.7±2.5

39.0±4.0

37.9±3.8

VH0.21

49.9±2.6*

40.2±3.9

43.7±5.5*

NB0.157

37.2±3.0*

37.6±4.1

34.4±4.2*

[Hb] (g/dl)

NB0.21

16.1±0.8

15.9±0.8

15.8±0.6

VH0.21

15.8±0.8

16.0±0.9

16.1±0.7

NB0.157

15.6±1.1

15.7±1.0

15.8±0.6

pH

NB0.21

7.37±0.02

7.37±0.03

7.37±0.02

VH0.21

7.29±0.03*

7.32±0.03*

7.27±0.04*

NB0.157

7.38±0.03

7.34±0.03*

7.36±0.04

P50 (mmHg)

NB0.21

26.7±1.1

26.5±0.9

26.4±0.9

VH0.21

30.9 ±1.1*

29.4±1.7* *

31.4±2.1*

NB0.157

27.1±2.5

27.6±1.8

27.3±1.2

Values are mean±SD. S1, S3, series 1 and 3; T2, second transition; PaO2, arterial oxygen pressure; PaCO2, arterial carbon dioxide pressure; [Hb], haemoglobin concentration; P50, PO2 for SaO2=50%; NB0.21, normal breathing in normoxia; VH0.21, voluntary hypoventilation in normoxia; NB0.157, normal breathing in hypoxia; different from *NB0.21, †NB0.157 (p<0.05)

At S1 and S3, PaCO2 and P50 were higher and pH lower in VH0.21 than in both NB0.21 and NB0.157.

PaO2 was lower in VH0.21 than in NB0.21 at S1 and S3 and higher in VH0.21 than in NB0.157 at S3.

There was no difference in [Hb] between interventions at any moment of exercise.

#

Oxygen transport

The results are presented in [Fig. 3] and can be summarized as follows:

Zoom Image

Fig. 3 Variables of oxygen transport during exercise with normal breathing in normoxia (NB0.21, filled circle), voluntary hypoventilation in normoxia (VH0.21, filled square), normal breathing in hypoxia (NB0.157, open square). SpO2, arterial oxygen saturation; CaO2, arterial oxygen content; –aO2, oxygen transport; C–O2, mixed venous oxygen content; CaO2 – C–O2, difference in arterial-venous O2 content; O2ER, tissue oxygen extraction. S1, S2, S3, series 1, 2 and 3; T1, T2, T3, transition 1, 2 and 3. Significantly different from *NB0.21, †NB0.157, (a) previous series. Values are mean±SE; P<0.05.

At the end of each series, SpO2 and CaO2 in both VH0.21 and NB0.157 were not different whereas they were lower than in NB0.21. – aO2 was higher in VH0.21 than in NB0.157 at T1, S2, T2 and T3, and higher in VH0.21 than in NB0.21 at the end of each transition period.

CaO2 – C–O2 was lower in VH0.21 than in both NB0.21 and NB0.157 at the end of each series and higher in NB0.157 than in both VH0.21 and NB0.21 at T1 only. C––O2 was lower in NB0.157 than in both VH0.21 and NB0.21 during most part of the exercise but was not different between the latter. O2ER was higher in NB0.157 than in VH0.21 during most part of the exercise and higher in NB0.157 than in NB0.21 at T1 and T2.

Finally, SpO2, CaO2, CaO2 – C–O2 and – aO2 were higher at the end of each transition period than at the previous series in both VH0.21 and NB0.157 whereas there was no difference in NB0.21.

#

Arterial pressure

The results are presented in [Table 3] . ANOVA revealed a main effect of VH0.21 for diastolic pressure and mean arterial pressure (MAP) but not for systolic pressure. At rest, there was no difference between interventions for diastolic pressure and MAP. Diastolic pressure was higher in VH0.21 than in NB0.157 at S1, S2 and S3 and higher in VH0.21 than in NB0.21 at S3. MAP was greater in VH0.21 than in NB0.157 at the end of each series and greater in VH0.21 than in NB0.21 at S2 and S3.

Table 3 Arterial blood pressure.

Rest

S1

S2

S3

Syst (mmHg)

NB0.21

117.1±3.9

183.6±5.7

185.0±7.7

185.7±6.8

VH0.21

119.7±3.9

187.9±13.0

194.4±8.6

193.6±7.7

NB0.157

119.3±6.1

185.0±7.8

187.9±7.0

189.3±7.9

Dias (mmHg)

NB0.21

78.6±4.2

72.1±5.3

66.4±5.0

65.7±4.9

VH0.21

76.0±3.1

80.0±7.9

77.1±6.8

77.9±6.9*

NB0.157

76.4±3.6

66.4±6.7

56.4±4.3

56.4±5.8*

MAP (mmHg)

NB0.21

91.3±3.7

108.9±4.6

105.6±5.6

105.3±5.5

VH0.21

90.4±2.8

115.6±9.3

118.2±7.1*

116.0±6.7*

NB0.157

90.6±4.2

105.6±6.5

99.8±3.7

100.3±5.7

Values are mean±SD. S1, S2, S3, series 1, 2 and 3; Syst, systolic pressure; Dias, diastolic pressure; MAP, mean arterial pressure; NB0.21, normal breathing in normoxia; VH0.21, voluntary hypoventilation in normoxia; NB0.157, normal breathing in hypoxia; different from *NB0.21, †NB0.157 (p<0.05)

#

Discussion

The main finding of this study is that VH exercise led to a greater increase in – compared to exercise in normal breathing, whereas V˙O2 was not different between interventions. The greater – was the result of both a higher HR and SV. This phenomenon did not occur in NB0.157 while the drop in SpO2 was the same as in VH0.21.

The greater – and SV in VH than in normal exercise constitute an original finding since no study so far has ever focussed on these variables. Even though several studies have investigated the effects of VH exercise on HR, most of them did not find any increase compared to exercise in normal breathing [10] [15] [31] [41]. The short exercise duration (≤5 min) used in these studies may be one reason explaining the lack of increase in HR. In the present study, it is noticeable that HR was not higher in VH0.21 at the end of the first 5-min series but only from the first transition period. Thus, the exercise duration should probably be longer than 5 min to elevate HR, maybe because of the time required to obtain a significant increase in sympathetic activity. On the other hand, the effects on SV seem to be faster since SV increased already in the first 5-min series during VH0.21 with a subsequent increase in – . In a recent study, an increased HR was reported during a 5-min exercise with VH but the respiratory technique consisted in performing a prolonged expiration down to the residual volume [38]. In these conditions, there was a stronger solicitation of the respiratory muscles and a higher metabolic demand which was probably at the origin of the greater HR.

We hypothesized that the hypoxic effect of VH exercise resulting from the severe arterial desaturation would induce an increase in HR. This would have been another reason explaining why HR remained unchanged in most previous studies where hypoxemia did not occur. However, it is noticeable that in the present study, NB0.157 did not lead to a significantly higher HR, or a higher SV and – Q˙, than in NB0.21. This may appear surprising because some studies using both a same or close FIO2 and exercise intensity as in the present study reported a higher HR or – during submaximal exercise in hypoxia than in normoxia, whereas SV remained unchanged [2] [21] [22]. However, unlike a classical exercise in hypoxia, the arterial desaturation was not continuous during NB0.157 and lasted only 45 s with a recovery of 15 s after each hypoxic series. Thus, the duration of hypoxemia was probably not continuous enough to induce significant effects on the cardiac activity. It is therefore unlikely that a hypoxic effect plays a role in the increase in – , or if so, only at a minor level.

Since there was no significant change in HR, SV and – during NB0.157, hypercapnia likely played a role in the greater cardiac activity during VH0.21. Some studies have reported that exercise in hypercapnia led to a greater sympathetic and adrenergic stimulation with higher plasma catecholamine concentration [12] [20]. Both epinephrine and norepinephrine level concentrations seem to be greater under these conditions [12]. In the present study, HRV analysis suggests that VH0.21 induced a significant increase in sympathetic modulation to the heart. Indeed, while the indices of long term variability were higher, the indices of short term variability, representing mainly parasympathetic modulation of RR intervals, remained stable. Though exercise in hypoxia is also known to increase sympathetic activity and the level of circulating catecholamines [13] [27], we did not find any change in the autonomic activity during NB0.157. As mentioned above, this may be due to the fact that VH induces an intermittent hypoxemia and also that the drop of SaO2, although severe, might not be strong enough. Thus the increase in sympathetic activity during VH0.21 seems to be the consequence of the hypercapnic effect alone. The slightly higher diastolic pressure and MAP, which occurred during VH0.21, may also be linked to the elevated PCO2 since there was no change in these variables during NB0.157.

The respiratory pattern used during VH0.21 may also be partly responsible for the higher cardiac activity. The way VH is carried out probably induces modifications in intrathoracic pressures and consequently in the right ventricular function. During inspiration, the right ventricular filling pressure increases whereas afterload decreases. In contrast, right ventricular preload falls and afterload increases during expiration. Consequently, right ventricular filling and emptying will be enhanced during inspiration and lowered during expiration. During VH, the large and very brief inspirations probably led to an increase in SV through a “pump” effect. Furthermore, this may also have increased HR by stimulating the pulmonary stretch receptors. These mechanoreceptors present in the smooth muscle of the airways respond to excessive stretching of the lung during large inspirations [36] and lead to inhibition of the cardiac vagal motor neurones and consequently to an elevation of heart rate [3]. On the other hand, it is unlikely that the way VH is carried out led to changes in HRV. The frequency of ventilation during VH is maintained in the short term variability (0.25 Hz) and has probably not artificially increased the indices of long term variability. Moreover, frequency domain analyses of HRV have shown no modification of respiratory sinus arrhythmia (high frequency component of the spectrum).

The results show that HR and– were even greater during transitions than during exercise periods with VH. HR was also greater in VH0.21 than in NB0.157 at the end of each transition period whereas not different at the end of each series. It is likely that the hyperventilation following the periods with VH led to a higher cardiac activity. A stronger ventilation increases the work of respiratory muscles and probably elevates their O2 cost as previously reported [1] [9]. This could partly explain the higher pulmonary V˙O2, which reflects the whole-body demand, at the end of each transition period in VH0.21, and therefore the greater HR and – . The higher V˙O2 might also be the consequence of the oxygen debt undergone during periods with VH where anaerobic glycolysis seems to be greater under the hypoxic effect [40]. This hypothesis is supported by the fact that in NB0.157, V˙O2 was also greater during transitions periods. Finally, hyperventilation can lead to tachycardia since there is a central medullary coupling between cardiac and respiratory neurons [4] [11].

The main consequence of the higher Q˙– during VH0.21 was that the oxygen transport in arterial blood was maintained. While VH0.21 induced a dramatic drop of SpO2 and CaO2, Q˙–aO2 did not change compared to NB0.21. This increase in– may be considered as the main compensation for the severe arterial desaturation during VH0.21 since there was no change in C–O2 and since CaO2 -C–O2 was lower. This response was different from NB0.157 where C––O2 decreased, which maintained CaO2-C––O2. However, it is noteworthy that – aO2 remained unchanged in NB0.157 compared to NB0.21. This is surprising because during NB0.157 – did not increase whereas the drop in CaO2 was the same as in VH0.21. This may be explained by the fact that there was a slight increase in HR during NB0.157. This increase may have somewhat raised –, which was probably sufficient to prevent a significant change in – aO2 compared to NB0.21, although the mean values were a little lower. Thus, we could conclude that in NB0.157, unlike in VH0.21, the drop of CaO2 was primarily compensated by a lower C–O2 whereas a slightly higher – may have acted as a secondary compensation factor.

The fact that C–O2 was not different between VH0.21 and NB0.21 whereas it was lower in NB0.157 may appear surprising. However, the unchanged or even increased (S2) – aO2 in VH0.21, as a result of the greater – , may have prevented and made useless a decrease in C–O2. The mechanism by which this is possible remains unclear. We hypothesized in a recent study using NIRS technique that the greater muscle deoxygenation during VH may have partly been due to a lower femoral venous oxygen saturation [40]. The results of the present study seem to contradict this hypothesis. The lower CaO2 -C–O2 we found in VH0.21 could rather suggest a greater myoglobin desaturation. Myoglobin has been assumed to contribute importantly to the NIRS signal [32], even though it was also reported that its participation was only 10% [7]. However, the results and conclusions concerning C–O2 must be treated with caution. First, C––O2 was not directly measured but calculated from V˙O2 and – which was also indirectly measured by non invasive impedance cardiography. Therefore the calculated C–O2 may not be as precise as a direct measurement with a blood venous catheterism. Second, C–O2 represents the mixed, and not the femoral (CvO2) oxygen in venous blood. A slightly lower CvO2 might be statistically masked if venous oxygen content coming from another part of the body does not vary. This could be the case especially during submaximal exercise where the muscle blood flow is not high. Thus, although we found that C–O2 in VH0.21 was not different from NB0.21, a slight and lower CvO2 cannot be excluded.

The present findings could be useful for the application of VH training. The increased HR must be taken into account in the targeting of exercise intensity. Thus, as previously suggested [40], it should be wiser for athletes to exercise below ventilatory threshold to avoid a too strenuous effort and overtraining. In this study, the subjects exercised at 65% V˙O2max. This may appear low since high exercise intensities, close to the second ventilatory threshold, are required during IHT to obtain improvement in peformance [17]. However, unlike during IHT, exercise intensity is relative to V˙O2max measured at sea level during VH exercise. This is a bit misleading since at a FIO2 inducing the same arterial desaturation as VH, the decrease in V˙O2max is about 15% in trained subjects [18]. Even though hypoxemia is intermittent during VH, this decrease may impact on the relative exercise intensity which is, when referring to sea level V˙O2max, probably slightly underestimated. Therefore, exercise intensity should probably remain in a range between 65 and 75% of normoxic V˙O2max at the most during VH training.

In summary, this study showed that exercise with VH increased the cardiac work through a greater HR, SV, – , MAP and sympathetic modulation. The greater – , allowed to maintain the oxygen transport and was probably the main factor for compensating the fall in SaO2 and CaO2. These adaptations were different from exercise in hypoxia inducing a similar drop in SaO2. Thus, although not clearly demonstrated in the present study, other factors like the hypercapnic effect and the specific respiratory pattern may explain the greater cardiac activity during exercise with VH.

#

Acknowledgements

We gratefully thank all the volunteers who accepted to participate in this experiment as well as the personnel of the “Service de Physiologie – Explorations Fonctionnelles et Médecine du Sport” of Avicenne Hospital, Bobigny.

#

References

#

Correspondence

Dr. Aurelien Pichon

Laboratoire ‘Réponses

cellulaires et fonctonnelles à

l'hypoxie’

UFR SMBH

Université Paris 13

74 rue Marcel Cachin

93017 Bobigny

France

Phone: +33/14/8387/632

Fax: +33/14/8388/924

Email: aurelien.pichon@orange.fr

#

References

#

Correspondence

Dr. Aurelien Pichon

Laboratoire ‘Réponses

cellulaires et fonctonnelles à

l'hypoxie’

UFR SMBH

Université Paris 13

74 rue Marcel Cachin

93017 Bobigny

France

Phone: +33/14/8387/632

Fax: +33/14/8388/924

Email: aurelien.pichon@orange.fr

Zoom Image

Fig. 1 Respiratory variables during exercise with normal breathing in normoxia (NB0.21, filled circle), voluntary hypoventilation in normoxia (VH0.21, filled square), normal breathing in hypoxia (NB0.157, open square). V˙O2, oxygen consumption; V˙E expired ventilation; PETO2, end-tidal O2 pressure, PETCO2, end-tidal carbon dioxide pressure. S1, S2, S3, series 1, 2 and 3; T1, T2, T3, transition 1, 2 and 3. Significantly different from *NB0.21, †NB0.157, (a) previous series. Values are mean±SE; P<0.05.

Zoom Image

Fig. 2 Cardiovascular variables during exercise with normal breathing in normoxia (NB0.21, filled circle), voluntary hypoventilation in normoxia (VH0.21, filled square), normal breathing in hypoxia (NB0.157, open square). HR, heart rate; SV, stroke volume; –, cardiac output; S1, S2, S3, series 1, 2 and 3; T1, T2, T3, transition 1, 2 and 3. Significantly different from *NB0.21, †NB0.157, (a) previous series. Values are mean±SE; P<0.05.

Zoom Image

Fig. 3 Variables of oxygen transport during exercise with normal breathing in normoxia (NB0.21, filled circle), voluntary hypoventilation in normoxia (VH0.21, filled square), normal breathing in hypoxia (NB0.157, open square). SpO2, arterial oxygen saturation; CaO2, arterial oxygen content; –aO2, oxygen transport; C–O2, mixed venous oxygen content; CaO2 – C–O2, difference in arterial-venous O2 content; O2ER, tissue oxygen extraction. S1, S2, S3, series 1, 2 and 3; T1, T2, T3, transition 1, 2 and 3. Significantly different from *NB0.21, †NB0.157, (a) previous series. Values are mean±SE; P<0.05.