Relationship Between Ventilatory Thresholds and ...

5 downloads 0 Views 517KB Size Report
Jun 29, 2010 - cardio-respiratory interactions. ○▷ hyperpnea. ○▷ blood pressure variability. ○▷ time frequency analysis. ○▷ short time Fourier transform.
604 Physiology & Biochemistry

Relationship Between Ventilatory Thresholds and Systolic Blood Pressure Variability

Authors

F. Cottin1, P. Le Moing1, C. Filliau1, V. Martin1, Y. Papelier2

Affiliations

1

Key words ▶ heavy exercise ● ▶ cardio-respiratory ● interactions ▶ hyperpnea ● ▶ blood pressure variability ● ▶ time frequency analysis ● ▶ short time Fourier transform ●

Abstract ▼

Unité de Biologie Intégrative des Adaptations á I’Exercice (UBIAE), Institut National de la Santé et de la Recherche Médicale (INSERM) 902/EA 3872, Genopole, Evry, France 2 Laboratory of Physiology, Medicine Faculty, University of Paris-Sud, E.F.R., Hôpital Antoine Béclére, Clamart, France

During exercise, an increase in respiratory rate amplifies the blood pressure oscillations. This phenomenon is usually intensified when exercise rate exceeds the ventilatory thresholds (VTs). The present study examined whether VTs assessment was possible from systolic blood pressure variability (SBPV) analysis to give blood pressure ventilatory thresholds (BPVTs). Blood pressure, ECG, and Ventilatory equivalents (VE/VO2, VE/VCO2) were collected from 15 welltrained subjects during an incremental exhaustive test performed on a cycloergometer. The “Short-Time Fourier Transform” was applied to SBP series to compute the instantaneous high frequency SBPV power (HF-SBPV). BPVTs were

accepted after revision April 30, 2010

Introduction ▼

Bibliography DOI http://dx.doi.org/ 10.1055/s-0030-1255064 Published online: June 29, 2010 Int J Sports Med 2010; 31: 604–609 © Georg Thieme Verlag KG Stuttgart · New York ISSN 0172-4622

It is widely recognized that respiratory pump favours venous return during inspiration [19]. Therefore, the right end-diastolic volume increases during inspiration and, the intra-pericardium volume remaining nearly constant, a decrease in the left end-diastolic volume is subsequently observed. As a consequence of the Frank-Starling law, this decline in left end-diastolic volume induces a decrease in the left stroke volume resulting in a decrease in blood pressure (BP) during the inspiratory phase. The reverse effect is observed during expiration with an increase in BP. Thus, as successive decreases and increases in BP respectively occur during inspiratory and expiratory phases of breathing cycles, breathing is partly responsible for the variability of BP. During exercise, the increase in respiratory flow (VE) is associated with an increase in the breathing oscillations amplitude of BP. These oscillation amplitudes can be quantified by spectral or time-frequency analysis which provides

Correspondence Dr. Francois Cottin Unité de Biologie Intégrative des Adaptations á I’Exercice (UBIAE), Institut National de la Santé et de la Recherche Médicale (INSERM) 902/EA 3872, Genopole Boulevard F. Mitterrand 91025 Evry cedex France Tel.: + 33/169/644 881 Fax: + 33/169/644 895 [email protected]

Cottin F et al. Blood Pressure and Ventilatory Thresholds … Int J Sports Med 2010; 31: 604–609

determined in all but 3 subjects. For the 12 remaining subjects, visual examination of ventilatory equivalents and HF-SBPV power revealed 2 thresholds for both methods. There was no difference between the first (VT1 235 ± 60 vs. BPVT1 226 ± 55 W, p = 0.063) and second (VT2 293 ± 67 vs. BPVT2 301 ± 66 W, p = 0.063) thresholds. However, BPVT1 was slightly underestimated compared to VT1 (9.9 ± 15.4 W) given lower limit of agreement (LOA) at − 19.9 W and higher at 40.4 W. BPVT2 was over-estimated compared to VT2 ( − 8.8 ± 11.2 W) given lower LOA at − 30.9 W and higher at 13.4 W. Thus, BPVTs determination appears useful in conditioning programs with sedentary or pathological subjects but probably not with trained subjects.

the respiratory component of systolic blood pressure variability (SBPV): this is the spectral power of blood pressure in the HF range (HF-SBPV). It has been previously shown [3, 12, 17] that during moderate exercise while the respiratory component of heart rate variability decreases, HF-SBPV increases as the exercise load progressively increases. With a further increasing exercise load, the increase in VE becomes steeper once the first ventilatory threshold (VT1) is exceeded [1, 8] and increases again when the intensity exceeds the second ventilatory threshold (VT2). It has therefore been shown that during an exhaustive incremental cycling test, the time course of VE exhibits 2 deflection points respectively corresponding to the 2 ventilatory thresholds (VTs) [1].This phenomenon should entail an abrupt increase in venous return characterized by a marked increase in HF-SBPV corresponding to a detectable deflection point on the HF-SBPV time course. However, a recent study [9] has shown that, despite a significant difference in HF-SBPV between exercise intensities just below VT2 (pre-VT2) and just

Physiology & Biochemistry

above VT2 (post-VT2), no difference was reported between preVT1 and post-VT1 intensities. Thus, HF-SBPV may barely increase when the exercise intensity exceeds VT1 whereas it may increase more obviously when the intensity exceeds VT2. Previous studies [4, 7, 10, 11] have shown that it is possible to detect VTs from heart rate variability (HRV). However, the processing method is rather difficult. A time frequency analysis of RR series must be computed in order to first extract the time course of HF power and then, to detect the HF peak of HRV from a time frequency processing. Although the first processing is rather trivial, the second is somewhat complicated and sometimes impossible. Conversely, VTs assessment from SBPV appears easier because it can be done from the HF-SBPV time course alone. There is no need to detect the HF peak of SBPV. Indeed, as shown by Taylor and Eckberg [21], the RR interval changes appeared to follow arterial pressure changes in the upright position. In addition, a recent study [9] has shown that HF-SBP increased more than HF-HRV during heavy exercise. Thus, the effect of exercise-induced hyperpnea is more marked on SBPV than on HRV. Consequently, the assessment of VTs from SBPV appears easier than from HRV. Furthermore, for patients who have difficulty breathing with a gas exchange mask, for example, during chronic obstructive pulmonary disease (COPD), this method could be used as an alternative to determine VT1 for a training program prescription. In addition, ECG and continuous blood pressure being monitored during the exercise test, the physicians could then prevent potential patient cardiac distress during exercise. The aim of the present work was to examine whether, during a graded maximal cycling test, the time course of HF-SBPV would yield 2 clearly identifiable deflection points. In addition, the matching of these 2 points with ventilatory thresholds assessed from ventilatory equivalents was expected. It was therefore hypothesized that exceeding VT2 would correspond to a concomitant deflection point on the time course of HF-SBPV allowing VT2 determination from HF-SBPV (BPVT2). However, the effect of hyperpnea on SBPV induced by VT1 being less substantial than by VT2 [9], VT1 detection from HF-SBPV (BPVT1) should be more difficult than BPVT2.

protocol. All subjects gave their written voluntary informed consent in accordance with the guidelines of the University of Evry and the study was approved by the local ethical committee. Lastly, the study has been performed in accordance with the ethical standards of the IJSM [14].

Experimental design All subjects performed an exhaustive incremental exercise test in the upright position on an electronically braked cycle ergometer (ERGOLINE 900, Hellige, Marquette, USA) in an air-conditioned room. After a 2 min warm-up at 60 W for females or 75 W for males, each subject performed a 0.3 W.kg − 1.min − 1 power increment. Seat and handlebar heights were set for each subject and kept constant for all tests.

Data collection procedures ECG measures were digitized and recorded with a PowerLab device (ADInstruments Ltd, AUS) with a sampling frequency of 1 000 Hz. Beat to beat RR intervals were extracted from ECG using Chart5 soft (Chart5, v5.5, ADInstruments, AUS). A Finometer device (TNO, BMI, Netherlands) was used to record blood pressure from a cuff placed on a middle finger. Each subject was instructed to keep his/her finger relaxed on a special handlebar that supported the elbows and forearms during exercise. The Finometer device was equipped with a height corrector unit that compensates for level positioning of the finger cuff. The height correction between hand and heart level was done automatically thanks to the 2 sensors during the signal recording. The first sensor was placed on the finger tip while the second ▶ Fig. 1). BP signal yielded by sensor was placed at heart level (● the Finometer device has been previously experimentally validated during laboratory tests [15, 16]. The Finometer was connected to the PowerLab that digitized and sampled BP signal at 1 000 Hz. The series of systolic blood pressure (SBP) were Sensor to heart level Finometer unit Sensor to finger level

Methods ▼ 15 competitive cyclists or triathletes (4 females, 11 males) participated in the study. They all held a regular sport license. Although they trained 3 or 4 times a week, the subjects were not high-performing athletes. All subjects were free from cardiac or pulmonary disease. The anthropometric and physiological char▶ Table 1. Prior to acteristics of the subjects are summarized in ● participating, each subject was familiarized with the experimental procedure and informed of the risks associated with the

Special handlebar

Ergoline

Table 1 Subjects characteristics.

number age (year) height (cm) weight (kg) relative VO2MAX (ml · min − 1 · kg − 1) maximal power (W)

Males

Females

9 25 ± 13 178 ± 4 71 ± 8 62 ± 10 359 ± 47

3 24 ± 14 162 ± 2 61 ± 11 51 ± 5 233 ± 26

Data are presented as average ± standard deviation

Fig. 1 The picture displays an example of the adequate arms position of a subject during the exercise blood pressure recordings with the Finometer. A special handlebar supported the elbows and forearms allowing the subject to keep his/her finger relaxed. In addition, the Finometer device was provided with a height corrector unit that compensates for level positioning of the finger cuff. The height correction between hand and heart level was done automatically thanks to the 2 sensors during the signal recording. The first sensor was placed on the finger tip while the second sensor was placed at heart level.

Cottin F et al. Blood Pressure and Ventilatory Thresholds … Int J Sports Med 2010; 31: 604–609

605

606 Physiology & Biochemistry

assessed using a detection technique provided with the Chart5.5 software. Beat-to-beat SBP values were then extracted with a 1 mm Hg accuracy. Each SBP series was examined and eventually corrected by interpolation when necessary. This procedure was thoroughly performed in order to eliminate all artifacts. Gas exchanges (VO2, VCO2, and VE) were measured with a Quark device (Quark PFT, COSMED, Rome, Italy), which was calibrated before each test.



VT1 300W

Power VE/VO2 VE/VCO2

STFT providing 1 spectrogram every 3 s, it was then possible to ▶ Fig. 2). get the instantaneous HF power every 3 s (●

Ventilatory thresholds assessment Ventilatory thresholds were assessed from gas exchange components. The ventilatory equivalent method was used to determine VT1 and VT2 [18]. Therefore, VE/VO2 and VE/VCO2 were plotted ▶ Fig. 2). VT1 vs. work rate during the incremental exercise test (● corresponded to the first deflection point increase in the VE/VO2 graph whereas the VE/VCO2 slope remained constant [18]. In addition, VT2 related to the deflection point increase in the VE/ VCO2 graph concomitant to the second strong increase in the VE/VO2 exercise intensity relationship [18].

VT2 400W

45

350

35

300 30

250 200

25

150 100

Ventilatory equivalents

40

400

Power (W)

PSD ⋅ f

f =0.15

The Short Time Fourier Transform (STFT) was used to compute SBPV (Matlab software, 6.5.1, Mathworks Inc.). This method has been already previously described in details [4, 5, 11]. The main principle of STFT consists in choosing for analysis a small enough window in which the signal can be considered stationary [13]. Standard FFT can therefore be performed on this windowed signal. The same analysis window is applied to the next signal block with a chosen overlap, and so on until the end of the processed series. STFT is constituted by the iterative FFT performed on the successive signal blocks that are determined by regular translation of the chosen window. Therefore STFT yields a 3D figure made of all the spectra vs. time called spectrogram, which is a time-frequency representation [11]. Prior to the STFT processing, all SBP series were resampled at 4 Hz using a cubic spline function (Matlab software, 6.5.1, Math-

450

f max

HF =

Signal processing

500

works Inc.). Each spectrogram window was made of 256 successive RR resampled periods corresponding to 64 s. The successive spectrogram windows were spaced by 12 successive SBP values corresponding to a duration of 3 s. The spectral power was computed in HF range (spectral components) by integrating the power spectral density (PSD) for each spectrum of the spectrogram as follows [20]:

20

50 0 0

120

240

360

480

600

720

840

960

15 1080

Time (s)

450

Power (W)

200

Power HF (SBP)

180

400

160

350

140

300

120

250

100

200

80

150

60

100

40

50

20

0 0

120

240

360

480

600

720

840

960

0 1080

Time (s)

Cottin F et al. Blood Pressure and Ventilatory Thresholds … Int J Sports Med 2010; 31: 604–609

HF SBP (mm2 Hg)

500

Fig. 2 Example of the relationship between ventilatory thresholds and the respiratory component of systolic blood pressure variability. Top. Ventilatory threshold detection from ventilatory equivalents; x-axis: time (s); left y-axis: load (W); right y-axis: ventilatory equivalents. Solid line stands for exercise load, empty circles for VE/VO2 and full circles for VE/VCO2. The ventilatory thresholds (VT1 and VT2) are symbolized by vertical dashed lines and the corresponding loads are indicated close to these lines. Bottom. Instantaneous HF component of systolic blood pressure vs. time; x-axis: time (s); left y-axis: load (W), right y-axis: HF-SBPV (mm2 Hg). Solid line stands for HF-SBPV and exercise load. As for VTs, SBPV thresholds (BPVT1 and BPVT2) are symbolized by vertical dashed lines.

Physiology & Biochemistry

Table 2 Difference between ventilatory thresholds computed from ventilator equivalents or systolic blood pressure variability. Ventilatory

Blood

equivalents

Pressure Variability

first threshold (W) second threshold (W) first threshold ( % maximal power) second threshold ( % maximal power)

300

Significance

235 ± 60 293 ± 67 70 ± 8

226 ± 55 301 ± 66 67 ± 6

NS (p = 0.063) NS (p = 0.063) NS (p = 0.063)

88 ± 6

90 ± 5

NS (p = 0.063)

BPVT1

12 subjects

350

250

200

150

R = 0.97 p < 0.001

100 100

150

200

250

300

350

VT1

Data are presented as average ± standard deviation. NS is taken as non significant 50

Statistical analysis The Kolmogorov-Smirnoff normality test revealed that the difference between data (VT1 – BPVT1 and VT2 – BPVT2) were not normally distributed. Therefore, the Wilcoxon signed rank test was used to compare the respective ventilatory and BPV thresholds (VT1 vs. BPVT1 and VT2 vs. BPVT2). Linear regression and the Pearson Product Moment Correlation were used to test the correlation between VT1 vs. BPVT1, VT2 vs. BPVT2. Bland-Altman [6] plots were conducted to illustrate the relationship between ventilatory thresholds (VT1 and VT2) and their respective BPV thresholds (BPVT1 and BPVT2). From the Bland-Altman analysis the bias, the limits of agreement (LOA) and the random error components were computed [2]. All statistical analyses were conducted with the Sigmastat statistical software (SIGMASTAT 2.03, Jandel Scientifics, San Rafael, CA, USA, 1997). The significance level was set at p < 0.05.

Results ▼ VTs detection The 2 ventilatory thresholds were detected for all the subjects from the ventilatory equivalent method.

Blood pressure thresholds (BPTs) assessment BPTs could not be assessed in 3 subjects out of 15. For the 12 remaining subjects, the comparison between VTs and BPTs did not reveal any difference in both first (VT1 235 ± 60 vs. BPVT1 ▶ Table 2) and second threshold (VT2 226 ± 55 W, NS: p = 0.063, ● ▶ Table 2). In 293 ± 67 vs. BPVT2 301 ± 66 watts, NS: p = 0.063, ● addition, a strong correlation was observed between the first ▶ Fig. 3) as well as the second (VT1 vs. BPVT1, R = 0.97, p < 0.001, ● ▶ Fig. 4). The thresholds (VT2 vs. BPVT2, R = 0.98, p < 0.001, ● Bland Altman analysis revealed that, since the mean difference between VT1 minus BPVT1 (bias) was positive (9.9 ± 15.4

Average + 2SD

40 VT1 minus BPVT1 (W)

In addition, VTs were also assessed from the time course of HFSBPV power. As it has been specified in the introduction, 2 non linear increases in HF-SBPV vs. time, synchronous with VTs, were expected for the first (BPVT1) and the second (BPVT2) SBPV thresholds. Based on the above criteria, 2 experienced researchers independently assessed the ventilatory and SBPV thresholds for all detection methods. If there was a disagreement between them, a third experienced investigator’s opinion was required. When the latter agreed with one investigator, the corresponding threshold was retained. When all the investigators found different thresholds the trial was not taken into account.

30 20

Average = 10 W

10 0 –10 –20

Average – 2SD

–30 0

50

100

150

200

250

300

350

Average power output (W)

Fig. 3 Top: Linear regression. The first blood pressure variability threshold (BPVT1) is expressed vs. the first ventilatory threshold (VT1). The significance and the R value are displayed on chart. Bottom: Bland-Altman plot. The difference between VT1 and BPVT1 is expressed vs. average of the first thresholds power output. Center line equals mean difference between ventilatory and BP threshold power output (bias). The upper and lower lines are the mean difference ± 2 times standard deviation of the difference giving the lower limit of agreement at − 19.9 W and the higher at 40.4 W.

▶ Fig. 3), when compared with ventilatory equivalent, BP W, ● detection tended to slightly underestimate the first ventilatory threshold. The lower LOA was − 19.9 W while the higher was 40.4 W. Conversely, mean difference between VT2 minus BPVT2 ▶ Fig. 4), when compared (bias) being negative ( − 8.9 ± 11.2 W, ● with ventilatory equivalent, BP detection tended to slightly overestimate the second ventilatory threshold. The lower LOA was − 30.9 W while the higher was 13.4 W. In addition, the random error component was computed for each treatment. The 95 % random error component [2] was ± 30.2 W for VT1 vs. BPVT1 and ± 22.1 W for VT2 vs. BPVT2. The mean power stage being 21 W, it was then found a random error component of 1.5 exercise stages between VT1 vs. BPVT1 and 1 stage between VT2 vs. BPVT2.

Discussion ▼ The main results of this study are that, during a graded maximal cycling test, the hyperpnea encountered when exercising above the ventilatory thresholds was detected from SBPV time frequency analysis. Indeed, hyperpnea amplified breathing oscillations of BP, which were quantified by time frequency analysis with instantaneous HF-SBPV. The hypothesis of the present study is that increasing exercise intensity above the ventilatory thresholds would generate 2 marked increases visually identifiable as 2 deflection points on the time course of HF-SBPV. The ▶ Fig. 3, ● ▶ Fig. 4, ● ▶ Table 2) support this present results (● hypothesis.

Cottin F et al. Blood Pressure and Ventilatory Thresholds … Int J Sports Med 2010; 31: 604–609

607

608 Physiology & Biochemistry

point. On the other hand, in 3 subjects out of 15, no deflection point was identified. Thus, if the statistical analysis of the above mentioned study had been done on the 12 remaining subjects who had a clear BPVT1, the difference in HF-SBPV between preVT1 and post-VT1 would have been significant.

450 400

BPVT2

350 300

Comparison with VTs assessment from HRV time-frequency analysis

250 200

R = 0.98 p < 0.001

150 150

200

250

300

350

400

450

VT2 20

Average + 2SD

VT2 minus BPVT2 (W)

10 0 Average = –9W

–10 –20 –30 Average – 2SD

–40 0

100

200

300

400

500

Average power output (W)

Fig. 4 Top: Linear regression. The second blood pressure variability threshold (BPVT2) is expressed vs. the second ventilatory threshold (VT2). The significance and the R value are displayed on chart. Bottom: Bland-Altman plot. The difference between VT2 and BPVT2 is expressed vs. average of the first thresholds power output (bias). Center line equals mean difference between ventilatory and BP threshold power output. The upper and lower lines are the mean difference ± 2 times standard deviation of the difference giving the lower limit of agreement at − 30.9 W and the higher at 13.4 W.

A few studies [3, 12, 17] have investigated SBPV during moderate exercise. In the first mentioned study [3] exercise loads were set at 50, 75 and 100 % VT1. In the second study [12] exercise loads were set at 25, 50 and 75 % VO2MAX. Finally, in the third study [17] exercise loads were set at 20 and 40 % VO2MAX. Despite the discrepancies of exercise intensities, all these studies revealed that HF-SBPV increased with increasing load, which is consistent with the results of the present work. To the best of our knowledge, there is only one study [9] that has investigated non-invasive blood pressure variability during heavy exercise. As mentioned in the introduction, this study has reported a significant increase in HF-SBPV between exercise intensities just below VT2 (pre-VT2) vs. just above VT2 (postVT2). The present study confirms this result since a deflection point in the time course of HF-SBPV concomitant with exceeding VT2 has been shown. However, this recent study reported no difference between pre-VT1 vs. post-VT1 exercise intensities. To justify this result, the authors suggested that the ventilatory stimulus induced by exercise intensities above VT1 was not strong enough to provoke a concomitant significant increase in HF-SBPV. Therefore, a deflection point corresponding to exceeding VT1 on the time course of HF-SBPV was not actually expected in the present study. In spite of this, in 12 subjects out of 15 the first deflection point was detected. Plausible explanations of this inconsistency were twofold. On the one hand, the method of assessment of the BPVTs was visual. Thus, a slight increase in the slope of HF-SBPV might be visually noticed even if there was no significant difference between the data pre and post deflection

A previous study [11] determined VTs from HRV time frequency analysis in similar conditions (subjects, exercise test, methods). For the first ventilatory threshold, a 0.45 W difference for VT1 minus HRV-VT1 was found. Compared to the present study, the difference for VT1 minus BPVT1was 10 W. Similarly, for the second ventilatory threshold, a 0.91 W difference for VT2 minus HRV-VT2 was found whereas in the present study, the difference for VT2 minus BPVT2 was − 9 W. Thus, if the assessment of VTs from HRV was less easy than from SBPV, it was obviously more accurate.

Methodological limitations of the study This study has shown that VTs assessment was possible from SBPV time-frequency analysis. However, it must be used cautiously in a training perspective for the following reasons: ▶ As for the assessment of VTs from HRV analysis [4, 7, 10, 11], the deflection point detection was visual. It is not certain that a visual analysis was more accurate than a computer-driven method. The visual determination seemed more subjective. Future studies will need to implement a computer-driven method allowing a less subjective detection of the HF-SBPV deflection points. ▶ Although there was no difference between VTs and BPVTs, the probability that this result was due to chance was close to the significance limit (P = 0.063). In addition, the Bland-Altman analysis revealed a 10 W difference for VT1 minus BPVT1 and − 9W for VT2 minus BPVT2. Furthermore, the LOA seemed rather large. A random error component corresponding to 1.5 exercise stages between VT1 vs. BPVT1 and 1 stage between VT2 vs. BPVT2 was found. Therefore, from a training perspective, the BPVTs method is probably not reliable enough to detect subtle improvements of physical fitness in already trained sporty subjects. However, the method might be used in a conditioning program with sedentary or pathological subjects who have a larger progress margin.

Conclusion ▼ To conclude, this study has shown that despite a good correlation between VTs and BPVTs , the BPVTs method is probably not accurate enough to monitor the improvements of physical fitness in already trained sporty subjects. However, it could for instance be used for exercise prescription in patients with pulmonary disease such as COPD. Besides, this new method is easier to compute but less accurate than the method of VTs assessment from heart rate variability. Unfortunately, the Finometer device is quite expensive. But it is plausible that in the near future the technological progress will allow physicians to continuously monitor the BP of their patients during exercise tests at a lower cost. Lastly, additional investigations need to be conducted with the implementation of a computer-driven detection of BPVT.

Cottin F et al. Blood Pressure and Ventilatory Thresholds … Int J Sports Med 2010; 31: 604–609

Physiology & Biochemistry References 1 Ahmaidi S, Hardy JM, Varray A, Collomp K, Mercier J, Prefaut C. Respiratory gas exchange indices used to detect the blood lactate accumulation threshold during an incremental exercise test in young athletes. Eur J Appl Physiol 1993; 66: 31–36 2 Atkinson G, Nevill AM. Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Med 1998; 26: 217–238 3 Bartels MN, Jelic S, Ngai P, Gates G, Newandee D, Reisman SS, Basner RC, De Meersman RE. The effect of ventilation on spectral analysis of heart rate and blood pressure variability during exercise. Respir Physiol Neurobiol 2004; 144: 91–98 4 Blain G, Meste O, Bouchard T, Bermon S. Assessment of ventilatory thresholds during graded and maximal exercise test using time varying analysis of respiratory sinus arrhythmia. Br J Sport Med 2005; 39: 448–452 5 Blain G, Meste O, Bermon S. Influences of breathing patterns on respiratory sinus arrhythmia during exercise. Am J Physiol 2005; 288: H887–H895 6 Bland JM, Altman DG. Statistical methods for assessing agreement between 2 methods of clinical measurement. Lancet 1986; 307–310 7 Buchheit M, Solano R, Millet GP. Heart-rate deflection point and the second heart-rate variability threshold during running exercise in trained boys. Pediatr Exerc Sci 2007; 19: 192–204 8 Clark JM, Hagerman FC, Gefland R. Breathing patterns during submaximal and maximal exercise in elite oarsmen. J Appl Physiol 1983; 55: 440–446 9 Cottin F, Médigue C, Papelier Y. Effect of heavy exercise on spectral baroreflex sensitivity, heart rate and blood pressure variability in well-trained humans. Am J Physiol 2008; 295: H1150–H1155 10 Cottin F, Médigue C, Lopes P, Leprêtre PM, Heubert R, Billat VL. Ventilatory thresholds assessment from heart rate variability during an incremental running test. Int J Sports Med 2007; 28: 287–294 11 Cottin F, Leprêtre PM, Lopes P, Papelier Y, Médigue C, Billat VL. Assessment of ventilatory thresholds from heart rate variability in welltrained subjects during cycling. Int J Sports Med 2006; 27: 959–967

12 Cottin F, Papelier Y, Escourrou P. Effects of exercise load and breathing frequency on heart rate and blood pressure variability during dynamic exercise. Int J Sports Med 1999; 20: 232–238 13 Gabor D. Theory of communication. J Int Elec Eng 1946; 93: 429–457 14 Harriss DJ, Atkinson G. International Journal of Sports Medicine – Ethical Standards in Sport and Exercise Science Research. Int J Sports Med 2009; 30: 701–702 15 Idema RN, Van Den Meiracker AH, Imholz BP, Man In’t Veld AJ, Ritsema Van Eck AD, Settels JJ, Schalekamp AD. Comparison of Finapres non invasive beat to beat finger blood pressure with intrabrachial artery pressures during and after bicycle ergometry. J Hypertens 1989; 7: S58–S59 16 Imholz BPM, Wieling W, Langewouters GJ, Montfrans GA. Continuous finger arterial pressure: utility in the cardiovascular laboratory. Clin Auton Res 1991; 1: 43–45 17 Macor F, Fagard R, Amery A. Power spectral analysis of RR interval and blood pressure short-term variability at rest and during dynamic exercise: Comparison between cyclists and controls. Int J Sports Med 1996; 17: 175–181 18 Reinhard U, Muller PH, Schmulling RM. Determination of anaerobic threshold by the ventilation equivalent in normal individuals. Respiration 1979; 38: 36–42 19 Rowell LB. Human Cardiovascular Control. New York: Oxford University Press; 1993; 42–43 20 Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Circulation 1996; 93: 1043–1065 21 Taylor JA, Eckberg DL. Fundamental relations between short-term R-R interval and arterial pressure oscillations in humans. Circulation 1996; 93: 1527–1532

Cottin F et al. Blood Pressure and Ventilatory Thresholds … Int J Sports Med 2010; 31: 604–609

609