β₁-Blockade increases maximal apnea duration in elite breath-hold divers

The governing bodies for international apnea competition, the Association Internationale pour le Développment de l’Apnée and La Confédération Mondaile des Activités Subaquatiques, have banned the use of β-blockers based on anecdotal reports that they improve apnea duration. Using a randomized placebo-controlled trial, we are the first to empirically confirm that β-blockade improves apnea duration. This improvement in apnea duration coincided with a reduced myocardial oxygen consumption.

Abstract

We hypothesized that the cardioselective β₁-adrenoreceptor antagonist esmolol would improve maximal apnea duration in elite breath-hold divers. In elite national-level divers (n = 9), maximal apneas were performed in a randomized and counterbalanced order while receiving either iv esmolol (150 μg·kg⁻¹·min⁻¹) or volume-matched saline (placebo). During apnea, heart rate (ECG), beat-by-beat blood pressure, stroke volume (SV), cardiac output (CO), and total peripheral resistance (TPR) were measured (finger photoplethysmography). Myocardial oxygen consumption (MV̇o₂) was estimated from rate pressure product. Cerebral blood flow through the internal carotid (ICA) and vertebral arteries (VA) was assessed using Duplex ultrasound. Apnea duration improved in the esmolol trial when compared with placebo (356 ± 57 vs. 323 ± 61 s, P < 0.01) despite similar end-apnea peripheral oxyhemoglobin saturation (71.8 ± 10.3 vs. 74.9 ± 9.5%, P = 0.10). The HR response to apnea was reduced by esmolol at 10–30% of apnea duration, whereas MAP was unaffected. Esmolol reduced SV (main effect, P < 0.05) and CO (main effect; P < 0.05) and increased TPR (main effect, P < 0.05) throughout apnea. Esmolol also reduced MV̇o₂ throughout apnea (main effect, P < 0.05). Cerebral blood flow through the ICA and VA was unchanged by esmolol at baseline and the last 30 s of apnea; however, global cerebral blood flow was reduced in the esmolol trial at end-apnea (P < 0.05). Our findings demonstrate that, in elite breath-hold divers, apnea breakpoint is improved by β₁-blockade, likely owing to an improved total body oxygen sparring through increased centralization of blood volume (↑TPR) and reduced MV̇o₂.

NEW & NOTEWORTHY The governing bodies for international apnea competition, the Association Internationale pour le Développment de l’Apnée and La Confédération Mondaile des Activités Subaquatiques, have banned the use of β-blockers based on anecdotal reports that they improve apnea duration. Using a randomized placebo-controlled trial, we are the first to empirically confirm that β-blockade improves apnea duration. This improvement in apnea duration coincided with a reduced myocardial oxygen consumption.

voluntary apnea elicits an integrative circulatory and neural response to maintain adequate perfusion of vital organs (e.g., the brain) while simultaneously reducing both flow to nonvital organs (e.g., skin and skeletal muscle) and total body oxygen consumption (9, 10, 20). Bradycardia, peripheral vasoconstriction, and centralization of blood volume are the primary cardiovascular responses to voluntary apnea, which together represent the mammalian diving reflex (20). During the initial phase (~25%) of a maximal apnea the extreme lung volumes mechanically compress the heart, whereas high intrathoracic pressures impede venous return, resulting in an abrupt drop in stroke volume (SV), cardiac output (CO), and consequently mean arterial pressure (MAP) and cerebral blood flow (CBF) (5, 8, 26). Concomitant unloading of the baroreceptors causes a transient increase in muscle sympathetic nerve activity (7), heart rate (22), and total peripheral resistance (TPR). Following this initial phase, bradycardia (via reduced baroreflex output in addition to increases in vagal tone; see Ref. 18) occurs alongside progressive increases in CO, MAP, TPR (5, 25), and CBF (23, 31). These elevations in CO, MAP, and TPR are thought to be mediated via marked elevations in sympathetic nerve activity (10- to 20-fold) (14, 29). Along with hypoxemic- and hypercapnic-mediated dilation of cerebral blood vessels, this increase in MAP aids in the augmenting the CBF response to maintain cerebral oxygen delivery throughout the apnea (31).

The precise mechanism(s) for apnea break point remains poorly understood (24) but in the motivated and elite breath-hold diver is likely related in part to chemoreflex stress (3, 4) and a critical oxygen level required to maintain consciousness (4, 31). Additional mechanisms have been theorized to contribute to break point in the untrained breath-holder, such as diaphragmatic afference (24); however, how this translates to the motivated and elite breath-holder is unknown.

β-Adrenoreceptor antagonists (β-blockers) have been used commonly in elite apnea competitions due to anecdotal reports of improved maximal apnea duration (personal correspondence, Drvis I, Croatian National Apnea Coach). As a result, the Association Internationale pour le Développment de l’Apnée (AIDA; https://www.aidainternational.org) and La Confédération Mondaile des Activités Subaquatiques (CMAS; http://www.cmas.org) prohibit the use of β-blockers in and out of competition. Given that recent work in elite apnea divers supports the presence of a critical oxygen tension as a major contributor to apnea break point (4), the present theory is that β-blockade improves maximal apnea duration via myocardial oxygen sparing, yet there are no published data demonstrating this effect. In the first randomized and placebo-controlled study of its kind, we determined the effects of β₁-blockade on maximal apnea duration and the circulatory parameters that are characteristic of extreme apnea. It was hypothesized that β₁-blockade would prolong apnea duration, owing to a reduced heart rate throughout the apnea and a consequently reduced rate of myocardial oxygen consumption (MV̇o₂). We further reasoned that the effects of β₁-blockade would be reflected in the cerebral vasculature through a lower CBF during baseline and end apnea compared with the placebo control due to reduced CO and MAP.

METHODS

Subjects

Elite breath-hold divers (n = 9; 1 female) were recruited from the National Croatian Apnea team to participate in this study. The divers were 29 ± 9 yr old, weighed 80 ± 13 kg, and were 184 ± 8 cm tall (body mass index = 23.5 ± 2.6). Personal record static apnea time was on average 394 s (range: 296–496), whereas divers had a mean of 4 yr (range: 2–11) of experience in competitive apnea training and a forced vital capacity (FVC) of 6.9 liters (range: 5.0–7.8). Following written informed consent, divers visited the laboratory on one occasion for the experimental session. Subjects arrived at either 8 AM or 12 PM, having abstained from alcohol, caffeine, and exercise for 24 h before arrival. All subjects were free from any cardiovascular, respiratory, and cerebrovascular disease at the time of study, as assessed by a screening questionnaire and spirometry. This study was approved by the University of British Columbia Clinical Research Ethics Board and by the Ethics Committee of the University of Split School of Medicine and conformed to the Declaration of Helsinki.

Experimental Design

Upon arrival to the laboratory, subjects first performed a spirometry test (Quark PFT; Cosmed, Rome, Italy) in the upright position. Subjects voided their bladders and were then instructed to rest supine, at which time a 20-gauge intravenous catheter was placed into an antecubital vein under local anesthesia (lidocaine 1.0%). Following cannulation, subjects rested for ≥20 min during the setup of experimental monitoring equipment.

This study implemented two experimental conditions: 1) cardiac-specific β₁-adrenergic receptor blockade (Esmolol, 10 mg/ml; Esmocard) and 2) a volume-matched placebo (saline; 0.9% NaCl) condition. Esmolol was first infused at a rate of 500 μg·kg⁻¹·min⁻¹ for 1 min, followed by 50 μg·kg⁻¹·min⁻¹ for 5 min. This was repeated two additional times before a steady maintenance infusion of 150μg·kg⁻¹·min⁻¹ for the remainder of the drug condition. These doses have previously been established to effectively reduce HR, systolic blood pressure, rate pressure product (RPP), and both left and right ventricular ejection fraction during rest and exercise (16). The order of infusions was randomized in a counterbalanced manner and separated by >5 half-lives (60 min) (30). The participants and apnea coach were blinded to the experimental condition.

All experimental breath holds (maximal breath holds) were completed in the presence of the breath hold divers’ coach, the Croatian national apnea coach. The coach was present to ensure that all divers were attaining a true maximal apnea. Each maximal breath hold was preceded by two practice breath holds. The preparatory breath holds included one at functional residual capacity lasting until seven involuntary breathing movements (IBMs) and, after two minutes of rest, a second breath hold at total lung capacity lasting until 10 IBMs. Subjects then rested quietly for 6 min in preparation for their maximal breath hold. The experimental breath hold was performed at TLC, whereas the extent of glossopharyngeal insufflation (lung packing) performed was based upon the individual capacity of each subject but standardized between trials.

Experimental Measures

Cardiovascular measures.

All cardiovascular and respiratory variables were sampled continuously at 1,000 Hz via an analog to digital acquisition system (Powerlab; ADInstruments, Colarado Springs, CO). Finger photoplethysmography (Finometer, Finapres Medical Systems, Amsterdam, The Netherlands) was used to measure beat-by-beat blood pressure and estimate cardiac output (CO), stroke volume (SV), and total peripheral resistance (TPR), whereas standard three-lead electrocardiogram (ECG) was used to measure heart rate. Peripheral oxyhemoglobin saturation ($\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$) was measured during baseline and throughout the apneas (Poet II; Criticare). The IBM onset was recorded in real time during the apnea by the apnea coach and verified from a chest plethysmography belt. All data were interfaced with LabChart (version 7; ADInstruments) on a laboratory computer and stored for offline analysis.

Myocardial oxygen consumption.

For each trial, MV̇o₂ of the left ventricle was estimated from RPP (the product of systolic blood pressure and HR) and left ventricular mass. To calculate MV̇o₂ per 100 g tissue, we used the equation reported by Gobel et al. (12):

$$M\stackrel{․}{V}{O}_2\left(ml\cdot 100g^{-1}\cdot mi{n}^{-1}\right)=0.08\left(RPP\times {10}^{-2}\right)-0.15$$ (1)

Left ventricular mass was estimated at baseline using echocardiography from a parasternal short-axis view, in accordance with the recommendations of the American Society of Echocardiography (19) using the following equation:

$$\text{LV mass}\left(\text{g}\right)=0.8\times 10.4\times \left[{\left(IVS+LVID+PWT\right)}^3-LVI{D}^3\right]+0.6g$$ (2)

where IVS and PWT are the diameters of the ventricular septum and inferolateral wall during diastole, respectively, and LVID is the internal diameter of the LV during diastole.

To then calculate absolute MV̇o₂ for each subject, left ventricular mass was incorporated into the following equation on an individual basis:

$$M\stackrel{․}{V}{O}_2\left(ml\cdot mi{n}^{-1}\right)=\frac{\text{left ventricular mass}\left(\text{g}\right)}{100g}\left(0.08\left(RPP\times {10}^{-2}\right)-0.15\right)$$ (3)

The cumulative myocardial oxygen consumption throughout apnea was then calculated by estimating MV̇o₂ from RPP on a beat-by-beat basis. The volume cost of oxygen per beat was calculated by dividing beat-by-beat MV̇o₂ by the R-R interval for each heart-beat. Total myocardial oxygen consumption for the entire apnea was calculated as the area under the curve of the beat-to-beat O₂ cost.

Cerebrovascular measures.

Blood velocity in the right middle cerebral artery (MCAv) and left posterior cerebral artery (PCAv) were measured using a 2M-Hz transcranial Doppler ultrasound (TCD; Spencer Technologies, Seattle, WA). The TCD probes were attached bilaterally to a specialized commercial headband (model M600 bilateral head frame; Spencer Technologies) and secured in place. Insonation of the MCA and PCA was performed through the transtemporal window using previously described location and standardization techniques (32).

Blood flow through extracranial cerebral vessels was measured using a 10-MHz, multifrequency, linear array vascular ultrasound (Terason T3200; Teratech, Burlington, MA). The internal carotid (ICA) and vertebral artery (VA) were insonated ipsilateral to the MCA and PCA, respectively. Arterial diameter was measured via B-mode imaging, whereas peak blood velocity was simultaneously measured with pulse-wave mode. Measurements of ICA diameter and velocity (ICAv) were acquired ~1.5 cm distal to the common carotid bifurcation, with no evidence of turbulent or retrograde flow present during recording. Vertebral artery diameter and velocity (VAv) were acquired between either the C4–C5 or C5–C6 vertebral segment. This location was determined on an individual basis in an attempt to select the most reproducible measures, with the same location repeated within subjects. Measurement location and insonation angle were standardized within subjects and between the esmolol and placebo conditions. An angle correction of 60° was used for all subjects, whereas the velocity sample volume was placed in the center of the vessel and adjusted to span across the entire luminal diameter. Baseline recordings were made in all subjects (n = 9); however, the sample size for VA measures throughout an entire breath hold was reduced (n = 6) due to strict exclusion criteria such as obvious angle changes.

Ultrasound images of the extracranial vessels were recorded as video files at 30 Hz and stored for offline analysis. Concurrent values for arterial diameter and peak blood velocity were acquired at 30 Hz, using customized edge detection software designed to mitigate observer bias (33). Blood flow was subsequently calculated for each vessel using the following formula:

$$Q_{ICA}\text{or}{Q}_{VA}=\frac{\text{peak velocity}}{\text{2}}\times \text{π}{\left(\frac{\text{diameter}}{2}\right)}^2$$ (4)

where Q_ICA represents ICA blood flow and Q_VA represents VA blood flow. For the maximal apneas, Q_ICA and Q_VA were averaged for a 1-min baseline and the last 30 s of apnea. Furthermore, values were also calculated on an individual basis to represent 10% increments in total apnea duration. No-flow calculation included <12 consecutive cardiac cycles, and where possible it included the entire averaging period. Once individual vessel flow was calculated, global cerebral blood flow (gCBF) was estimated for each apnea time point using the following formula:

$$gCBF=2(Q_{ICA}+Q_{VA})$$ (5)

During the latter half of each maximal apnea (i.e., struggle phase), IBMs resulted in large diaphragmatic movements and neck muscle recruitment (i.e., sternocleidomastoid contraction). As such, reliable velocity traces were no longer attainable. Once measures of Q_ICA and Q_VA were no longer attainable due to poor velocity traces, flow was estimated using the following formulas in subsequent fashion for the remainder of the apnea:

$$eICAv\text{or}eVAv=PreIBM\hspace{0.17em}ICAv\text{or}VAv\times \%\Delta MCAv\text{or}\%\Delta PCAv$$ (6)

$$Q_{ICA}\text{or}{Q}_{VA}=eICAv\text{or}eVAv\times \pi {\left(\frac{d}{2}\right)}^2$$ (7)

where changes in MCAv and PCAv were used to calculate an estimated ICA (eICAv) and VA velocity (eVAv), respectively (Eq. 3). The term “pre-IBM ICAv or VAv” represents the average velocity value of the last 30 s before the IBMs that precluded the measurement of velocity occurred, and “%ΔMCAv or %ΔPCAv” represents the relative change in MCAv and PCAv from the same pre-IBM stage. As such, this multiplication gives a reliable estimate of neck vessel blood velocity during the struggle phase of apnea in the event that changes in downstream vessel (MCA and PCA) velocity represent upstream vessel (ICA and VA) velocity changes. As reliable diameter measurements (d) were still attainable throughout IBMs, changes in vessel diameter were accounted for in our estimation of volumetric blood flow (Eq. 4). Volumetric flow through the neck vessels (Q_ICA and Q_VA) was then calculated from the estimated velocity values and vessel cross-sectional area $\left(\pi {\left(\frac{d}{2}\right)}^2\right)$. To exemplify the feasibility of this estimation, we ran a Pearson’s correlation between the percent changes in ICAv and VAv with those of MCAv and PCAv, respectively, throughout apnea until velocity measures were unattainable. The average within-subject r² values between ICAv and MCAv were 0.88 ± 0.12 and 0.90 ± 0.10 during the placebo and esmolol trials, respectively. For VAv and PCAv, the r² average was 0.88 ± 0.09 and 0.95 ± 0.03 for the placebo and esmolol trials, respectively. The strong correlation between blood velocity of confluent intra- and extracranial blood vessels supports the accuracy of our estimation. Furthermore, we have demonstrated previously that MCAv and ICAv reactivity do not differ statistically over a wide range of end-tidal CO₂ (rest to +9 mmHg) (15), whereas work in our laboratory also indicates that ICAv and VAv reactivity do not differ from MCAv and PCAv reactivity, respectively, during hypoxia at an SaO₂ of 98, 90, 80, and 70% (14a).

Statistical Analyses

Differences in apnea duration and time to IBM onset between esmolol and placebo were compared using two-tailed paired t-tests. To analyze the effect of esmolol over the progression of apnea, apnea duration was normalized to 100%, with comparisons being made between placebo and esmolol conditions at 10% relative increments in apnea duration using a two-way repeated-measures ANOVA (factors: esmolol and apnea duration). When significant F-ratios were detected, post hoc comparisons were made using Tukey’s honest significant difference test. End-apnea variables (final 30 s) in the placebo and esmolol trials were compared using two-tailed paired t-tests (Table 1). Data are presented as means ± SD.

Table 1.

End-apnea hemodynamic variables

Units	Placebo	Esmolol
HR, beats/min	54.2 ± 16.9	52.2 ± 13.3
MAP, mmHg	139.0 ± 18.6	129.6 ± 19.1
SV, ml	92.7 ± 21.8	78.5 ± 21.3†
CO, l/min	5.4 ± 2.4	4.4 ± 1.3
RPP, mmHg·beats−1·min−1	13,192 ± 3,882	10,124 ± 3,318†
Systolic BP, mmHg	247.9 ± 37.6	195.7 ± 25.8†
$\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$, %	74.9 ± 9.5	71.8 ± 10.3

Values are means ± SE. HR, heart rate; MAP, mean arterial pressure; SV, stroke volume; CO, cardiac output; RPP, rate pressure product; BP, blood pressure; $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$, peripheral oxyhemoglobin saturation.

^†Significant difference between placebo and esmolol, P < 0.05.

RESULTS

Apnea Duration

Apnea duration increased 10 ± 10% from 323.4 ± 60.6 s during the placebo trial to 355.9 ± 56.8 s after β₁-blockade with esmolol (P < 0.01; Fig. 1). The time to IBM onset was delayed by 21 ± 15% (from 160.4 ± 41.3 s during the placebo trial to 192.0 ± 45.3 s after treatment with esmolol, P < 0.01). There was no relationship between the absolute (r² = 0.06, P = 0.53) or percent (r² = 0.21, P = 0.22) change in IBM onset and apnea duration. Despite the prolonged apnea duration, $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ during the last 30 s of apnea was not significantly different between placebo (74.9 ± 9.5%) and esmolol trials (71.8 ± 10.3%) (P = 0.10).

Fig. 1.

Maximal apnea duration during placebo and esmolol trials. Individual data are represented by ○ and solid lines, whereas mean data are represented by ■ and the dashed line. Overall, there was a 32 ± 25-s increase in apnea duration with esmolol treatment. *Significant difference in apnea duration (P < 0.01).

Hemodynamics

Values and statistical output for HR, MAP, SV, CO, and TPR during each relative time point throughout apnea are presented in Fig. 2. During the initial phase of apnea, HR was elevated in both placebo and esmolol trials, whereas esmolol reduced HR compared with placebo at 10–30% of apnea duration. In both trials, MAP was elevated from baseline. Throughout apnea, SV and CO were reduced from baseline and lower (main effect: both P < 0.01) during the esmolol compared with the placebo trial from 40% to max apnea and 20% to max apnea, respectively. End-apnea (final 30 s; Table 1) HR and MAP were not different between esmolol and placebo, whereas SV and CO were reduced by 15 and 18%, respectively, in the esmolol trial; however, the difference in CO did not reach statistical significance (P = 0.06). Esmolol increased TPR (main effect: P = 0.04) compared with placebo, whereas within-condition TPR was elevated above baseline from 20 to 100% of apnea duration.

Fig. 2.

The hemodynamic response to maximal apnea during placebo and esmolol infusion. Data are normalized within subjects to represent 10% increments in apnea duration. ●, Data from the placebo trial; □, data from the esmolol trial. Data were analyzed using a 2-way repeated-measures ANOVA with the factors of apnea duration and esmolol treatment. *Significant difference from baseline P < 0.05; †main effect of treatment, P < 0.05; ‡interaction between apnea time and treatment. HR, heart rate; MAP, mean arterial pressure; SV, stroke volume; CO, cardiac output; TPR, total peripheral resistance.

There was a main effect of esmolol on MV̇o₂ (P < 0.01; Fig. 3), which was lower in the esmolol trial. In both trials MV̇o₂ was elevated from baseline at 30, 40, and 90%. There was a 16.2 ± 4.2% reduction in MV̇o₂ across all stages (range: 8.0–21.6%) in the esmolol trial. However, there was no difference in the cumulative myocardial oxygen consumption between the placebo (76.3 ± 19.9 ml) and esmolol (70.2 ± 21.7 ml) trials (P = 0.27; Fig. 3B) due to the prolonged duration of the esmolol apnea.

Fig. 3.

Myocardial oxygen consumption (MV̇o₂) during maximal apnea in elite breath-hold divers. A: data are normalized within subjects to represent 10% increments in apnea duration. ●, Data from the placebo trial; □, data from the esmolol trial. Data were analyzed using a 2-way repeated-measures ANOVA with the factors of apnea duration and esmolol treatment. *Significant difference from baseline, P < 0.05; †main effect of treatment, P < 0.05. B: total MV̇o₂ throughout apnea was unaltered between placebo and esmolol trials. Individual data are represented by ○ and solid lines, whereas mean data are represented by ■ and the dashed line.

Cerebral Blood Flow

Measurements of Q_ICA, Q_VA, gCBF, MCAv, and PCAv were not different between placebo and esmolol at baseline (Table 2). At the end of apnea (final 30 s), the sample sizes for gCBF and Q_VA were reduced to six participants. In these six participants, gCBF was lower at the end of apnea during the esmolol trial compared with placebo (Table 2).

Table 2.

Baseline and end-apnea cerebrovascular variables

	Baseline		Final 30 s
Units	Placebo	Esmolol	Placebo	Esmolol
QICA, ml/min	222.9 ± 33.5	201.5 ± 33.0	384.8 ± 77.1	347.6 ± 66.3
QVA, ml/min	74.3 ± 35.5	72.6 ± 38.7	144.5 ± 61.9	133.5 ± 57.3 (n = 6)
gCBF, ml/min	583.6 ± 99.8	531.1 ± 88.2	935.2 ± 415.0 (n = 6)	808.8 ± 373.1 (n = 6)†
MCAv, cm/s	57.3 ± 11.5	54.8 ± 10.3	94.8 ± 11.8	92.2 ± 12.0
PCAv, cm/s	41.0 ± 13.4 (n = 8)	39.2 ± 6.3 (n = 8)	68.4 ± 17.6 (n = 8)	64.6 ± 15.5 (n = 8)

Values are means ± SE. Q_ICA, internal carotid artery blood flow; ICAv, internal carotid artery blood velocity; D_ICA, internal carotid artery diameter; Q_VA, vertebral artery blood flow; VAv, vertebral artery blood velocity; D_VA, vertebral artery diameter; gCBF, global cerebral blood flow; MCAv, middle cerebral artery blood velocity; PCAv, posterior artery blood velocity. As noted in methods, the values for Q_ICA, Q_VA, and gCBF are based off of estimated velocity values combined with real diameter measurements.

^†Significant difference between placebo and esmolol, P < 0.05; n = 9 unless otherwise specified.

DISCUSSION

The primary findings of this study are as follows: 1) β₁-blockade with esmolol improves maximal apnea duration in elite breath-hold divers by ~10%; 2) esmolol reduces MV̇o₂ at each time point throughout maximal apnea by ~16%, whereas the cumulative myocardial oxygen consumption is unaltered; 3) there are marked changes in the hemodynamic response (TPR and CO) to apnea between the esmolol and placebo trials; and 4) end-apnea $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ was not different despite the prolonged apnea in the esmolol trial. Collectively, these data support the hypothesis that apnea break point is influenced by β₁-blockade but ultimately governed by a critical oxygen tension.

β₁-Blockade and Apnea Duration

It is now recognized that β-blocker drugs act as an ergogenic aid, and therefore, they are banned from international apnea competition by the AIDA. In support of this rule, using a placebo-controlled and blinded design, this study is the first to empirically demonstrate the increased maximal apnea time with β₁-blockade in elite breath-hold divers.

Hemodynamic Response to Apnea

β₁-blockade affects hemodynamics, and thus CO, through chronotropic and inotropic means (2, 16). In the present study, we observed a predominantly negative inotropic influence of β₁-blockade with esmolol on CO (reduced SV), with only slight chronotropic changes. With the exception of a lower HR during the initial hypotensive phase of apnea (first 10–30% apnea duration), esmolol did not have an appreciable effect on HR; MAP was not different from placebo at any relative time point during the esmolol trial apnea. However, esmolol did cause a reduction in SV throughout the apnea (likely through reduced contractility, i.e., negative inotropy), which caused a concomitant reduction in CO compared with the placebo trial. This reduction in CO would likely have acted to further baroreceptor unloading and augment the reflex increase in sympathetic output (and TPR) throughout the apnea (14) to maintain the MAP response. Despite a reduced CO, increased centralization of blood volume (due to ↑TPR) in the esmolol trial likely reduced nonvital organ perfusion while aiding in the selective maintenance of oxygen delivery to vital organs.

Esmolol reduced left ventricular MV̇o₂ throughout apnea when compared with the placebo trial, indicating that whole heart MV̇o₂ would likely have been reduced (Fig. 3). Reduced MV̇o₂ would further spare oxygen and also contribute in part to the prolonged apnea duration associated with β₁-blockade. Despite a lower MV̇o₂ throughout apnea, the cumulative myocardial oxygen consumption was not different between trials, owing to the prolonged apnea in the esmolol trial.

Mechanisms Governing Apnea Termination

The physiological apnea break point (i.e., cessation of breath hold) in untrained breath holders, which is observed as IBM onset in elite breath-hold divers, is governed largely by chemoreceptor afferents (6, 11). Breskovic et al. (6) suggested that IBM onset is related predominantly to a $\mathrm{P}{\mathrm{a}}_{\mathrm{C}\mathrm{O}}{}_{{}_{\mathrm{2}}}$ threshold level of 6.5 ± 0.5 kPa (48.8 mmHg). A similar $\mathrm{P}{\mathrm{a}}_{\mathrm{C}\mathrm{O}}{}_{{}_{\mathrm{2}}}$ threshold for IBM onset has been reported in most (1, 21) but not all (4) earlier studies, highlighting the importance of $\mathrm{P}{\mathrm{a}}_{\mathrm{C}\mathrm{O}}{}_{{}_{\mathrm{2}}}$ in initiating the physiological break point. Nevertheless, the prevailing oxygen tension will impact the level of $\mathrm{P}{\mathrm{a}}_{\mathrm{C}\mathrm{O}}{}_{{}_{\mathrm{2}}}$ at which IBM onset occurs (6). These chemoreflex interactions, along with a modulatory role from lung afferents (reviewed in Ref. 24), indicate that it is the integration of signals (i.e., peripheral and central chemoafference) that collectively govern the physiological break point.

Concurrent decreases in O₂ depletion and CO₂ production, both of which are a consequence of a reduced metabolism, will act to attenuate arterial blood gas changes and hence, chemoreflex stimulation. Indeed, our findings show that IBM onset was delayed by ~32 s but nevertheless occurred at the same level of $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ between trials (placebo = 97.8 ± 1.3%; esmolol = 96.4 ± 2.2%). Under the assumption that the respiratory exchange ratio was unaltered between trials, a similar reduction in $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ signifies that metabolic CO₂ production likely did not differ and was also similar between trials at IBM onset. Therefore, despite the delayed IBM onset with esmolol treatment, IBMs likely occurred at the same level of chemoreflex stress with no influence from altered chemosensitivity (13) or CO₂ washout in the brainstem (i.e., Q_VA, an index of brainstem flow, was similar between trials). Therefore, it is likely that the delay in reaching a chemoreflex threshold was responsible for delayed IBM onset (or physiological break point). In contrast to the physiological break point, the actual apnea termination in elite breath-hold divers appears to be only modestly affected by suppression of peripheral chemoreflex drive (4).

Typically, it is accepted that man is incapable of sustaining a voluntary apnea to the point of losing consciousness (27). However, many elite apnea divers are able to suppress their ventilatory drive during an apnea to the loss of consciousness (unpublished observations). This latter point is reflected in that AIDA includes loss of consciousness as a criterion for disqualification during apnea competition. At this point, the apnea “break point” is governed to a potentially large extent by a threshold oxygen tension to maintain cerebral functioning and, therefore, consciousness. Current literature indicates this threshold occurs at around 25–35 mmHg $\mathrm{P}{\mathrm{a}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ (4, 31), which coincides with P50. Despite a prolonged apnea duration following esmolol in the current study, $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ was not significantly different between trials, indicating similar end-apnea hypoxic stress. Based upon the commonly used Severinghaus equation (28) and our pulse-oximetry measurements, it appears that end-apnea $\mathrm{P}{\mathrm{a}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ would have been ~35–40 mmHg in both the placebo and esmolol trials. Although the achieved hypoxic stimulus does not appear as severe as that of a previous study indicating the potential of an oxygen threshold for apnea termination (4), the similar end-apnea $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ indicates some role of hypoxia in mediating apnea break point in both trials. Of note is that data from our previous study (4), where concurrent $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ (finger pulse-oximetry) and $\mathrm{S}{\mathrm{a}}_{{\mathrm{O}}_{\mathrm{2}}}$ (radial artery blood draw) measurements were collected, indicate that $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ overestimates $\mathrm{S}{\mathrm{a}}_{{\mathrm{O}}_{\mathrm{2}}}$ by a mean value of ~5% and that this over estimation is largest below an $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ of 70% (Fig. 4). Therefore, it is likely that the end-apnea hypoxic stimulus was similar to our previous studies in elite breath-hold divers (4, 31). These data are similar to our previous findings in elite breath-hold divers of similar end-apnea $\mathrm{P}{\mathrm{a}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ despite moderate prolongation (but in some cases reductions) in maximal apnea time following peripheral chemoreflex blunting with low-dose dopamine (4).

Fig. 4.

A Bland-Altman plot of $\mathrm{S}{\mathrm{a}}_{{\mathrm{O}}_{\mathrm{2}}}$ and $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ at the end of a maximal apnea. This figure demonstrates an overestimation of $\mathrm{S}{\mathrm{a}}_{{\mathrm{O}}_{\mathrm{2}}}$ (radial blood draw) by $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ (finger pulse-oximetry) at the end of apnea in elite breath-hold divers using data from our previous study (4). Dotted line represents the mean difference (5.3), whereas the dashed lines represent the upper and lower limits of agreement (16.1 and −5.6, respectively). Of note is that the overestimation of $\mathrm{S}{\mathrm{a}}_{{\mathrm{O}}_{\mathrm{2}}}$ by $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ is most evident below a $\mathrm{S}{\mathrm{a}}_{{\mathrm{O}}_{\mathrm{2}}}$ of 70%. Each data point represents an individual subject; n = 14.

In the current study, centralization of blood volume and, therefore, a reduction of nonvital organ perfusion but maintenance of vital organ oxygen delivery, in combination with myocardial oxygen sparring, likely reduced both the hypoxic and acidotic stresses’ characteristic of extreme apnea. Together, these factors likely underscore the increased apnea duration with esmolol.

Methodological Considerations

In the current study, we have used RPP and echocardiographic assessment of left ventricular mass to estimate left ventricular MV̇o₂ from the equation reported by Gobel et al. (12). Although RPP has been shown to be an excellent (r = 0.85) correlate of MV̇o₂ during exercise pre- and post-β-blockade (propranolol) (17), we acknowledge that this validation has not been made during the specific conditions of prolonged apnea. Despite this, it is unlikely that our estimates of RPP and MV̇o₂ would differ between conditions and discredit our findings. Furthermore, although we used doses comparable with those in other studies (16), we must acknowledge that we did not assess the completeness of our β₁-blockade. However, the presence of an incomplete block would have led to an underestimation of the magnitude by which esmolol impacts apnea duration.

We used $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ to quantify the hypoxic stimulus during prolonged breath hold during placebo (saline) and esmolol infusion. As demonstrated in Fig. 4, $\mathrm{S}{\mathrm{p}}_{\mathrm{O}}{}_{{}_{\mathrm{2}}}$ overestimates $\mathrm{S}{\mathrm{a}}_{{\mathrm{O}}_{\mathrm{2}}}$substantially below a $\mathrm{S}{\mathrm{a}}_{{\mathrm{O}}_{\mathrm{2}}}$ of ~70%. However, given the overestimation of $\mathrm{S}{\mathrm{a}}_{{\mathrm{O}}_{\mathrm{2}}}$, our data should be interpreted as presenting end-apnea hypoxic stress similar to that in previous investigations by our group (4, 31).

Conclusion

This study contributes to the recent body of literature (4, 31) supporting the emerging hypothesis that apnea break point in elite breath-hold divers is governed to a large extent by an oxygen threshold for the maintenance of consciousness. Cardiac-specific β₁-blockade with esmolol improved apnea duration in the current study and thus acts as an effective ergogenic aid in apnea competition. The likely mechanisms by which esmolol increased apnea duration are through increased centralization of blood flow (due to ↑TPR), maintenance of vital organ perfusion, and a reduced MV̇o₂. Together, these two factors would contribute to improved total body oxygen sparring and reduced metabolic CO₂ production.

GRANTS

This study was funded through a Canadian Research Chair and NSERC Discovery grant held by P. N. Ainslie. Z. Dujic, O. Barak, and P. N. Ainslie were also funded through the Croatian Science Foundation (Grant No. IP-2014-09-1937).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

R.L.H., P.N.A., A.R.B., D.B.M., M.S., I.D., D.M., O.F.B., D.M.M., and Z.D. performed experiments; R.L.H., P.N.A., and A.R.B. analyzed data; R.L.H., P.N.A., and A.R.B. interpreted results of experiments; R.L.H. prepared figures; R.L.H. drafted manuscript; R.L.H., P.N.A., A.R.B., D.B.M., M.S., I.D., D.M., O.F.B., D.M.M., and Z.D. edited and revised manuscript; R.L.H., P.N.A., A.R.B., D.B.M., M.S., I.D., D.M., O.F.B., D.M.M., and Z.D. approved final version of manuscript; P.N.A., A.R.B., D.B.M., I.D., and Z.D. conception and design of research.