Splenic responses to a series of repeated maximal static and dynamic apnoeas with whole‐body immersion in water

What is the central question of this study? Splenic contractions occur in response to apnoea‐induced hypoxia with and without face immersion in water. However, the splenic responses to a series of static or dynamic apnoeas with whole‐body water immersion in non‐divers and elite breath‐hold divers are unknown. What is the main finding and its importance? Static and dynamic apnoeas were equally effective in stimulating splenic contractions across non‐divers and elite breath‐hold divers. These findings demonstrate that the magnitude of the splenic response is largely dictated by the degree of the hypoxemic stress encountered during voluntary apnoeic epochs.


INTRODUCTION
The diving reflex is an oxygen conserving mechanism that is activated during the state of apnoea (Gooden, 1994). This reflex is primarily characterized by an initial bradycardic response which slows the depletion of bodily oxygen stores. This process is followed by a selective sympathetically-induced peripheral vasoconstriction in the body's extremities (arms and legs) and non-vital organs, with oxygenated blood being preferentially redistributed towards the vital organs (brain and heart) (Campbell, Gooden, & Horowitz, 1969;Kyhl et al., 2016;Shamsuzzaman et al., 2014;Sterba & Lundgren, 1988).
Therefore, the diving reflex serves a key role in effectively and economically managing bodily oxygen stores, enabling apnoeas to be sustained for prolonged durations until respiration is restored.

New Findings
• What is the central question of this study?
Splenic contractions occur in response to apnoeainduced hypoxia with and without face immersion in water. However, the splenic responses to a series of static or dynamic apnoeas with whole-body water immersion in non-divers and elite breathhold divers are unknown.
• What is the main finding and its importance?
Static and dynamic apnoeas were equally effective in stimulating splenic contractions across nondivers and elite breath-hold divers. These findings demonstrate that the magnitude of the splenic response is largely dictated by the degree of the hypoxemic stress encountered during voluntary apnoeic epochs. & Lundborg, 1977;Olsson, Kutti, Lundborg, & Fredén, 1976). It is therefore tempting to speculate that the collective effect of apnoeas and whole-body immersion would stimulate a greater splenic response.
To the best of our knowledge, there are no reports of the splenic responses to a series of dynamic apnoeas performed by either nondivers (ND) or elite breath-hold divers (EBHD). The physiological demands imposed by static and dynamic apnoeas differ substantially (Elia et al., 2019a;Overgaard, Friis, Pedersen, & Lykkeboe, 2006). The addition of contractile activity during the state of dynamic apnoea imposes a significant challenge to the diving reflex, where myocardial and skeletal muscle oxygen consumption is increased, and blood flow is redistributed to meet the competing needs of both the vital organs and recruited striated muscle. In addition, apnoea-induced physiological responses (e.g. the magnitude of the bradycardial response, erythropoietin release) vary across trained and untrained populations (Elia et al., 2019a;Joulia, Steinberg, Wolff, Gavarry, & Jammes, 2002;Lemaitre et al., 2005;Lemaitre, Buchheit, Joulia, Fontanari, & Tourny-Chollet, 2008). Evidence suggests that these differences across diving and non-diving populations are, at least in part, the result of a traininginduced stimulus (Joulia et al., 2003;Richardson et al., 2005;Schagatay, van Kampen, Emanuelsson, & Holm, 2000 Data are mean ± SD.
stronger hypoxemic stress compared with static apnoeas and that this will stimulate a greater splenic and systemic haematological response.

Ethical approval
Ethical approval for this human study was granted by the Leeds Beckett University Research Ethics Committee (52330), and all experimental procedures conformed to the Declaration of Helsinki, expect for registration in a database. All participants provided written informed consent before the study.

Participants
Twenty-six, healthy, non-smoking male participants volunteered for this study and were stratified into three groups including, EBHD (n = 8; height, 183 ± 1 cm; body mass, 84 ± 12 kg), ND (n = 10; height, 182 ± 1 cm; body mass, 85 ± 7 kg) and control (n = 8; height, 178 ± 1 cm; body mass, 82 ± 11 kg). All breath-hold divers were national team members (Table 1) and physically active individuals with no prior breath-hold diving experience were randomly assigned to the ND or control group. An independent control group was recruited due to the practical implications and time constraints of the study.

Methodology
Participants reported at Leeds Beckett University after a 12 h fast and abstinence from caffeine-and alcohol-containing beverages. In addition, participants were instructed to refrain from physical activity and apnoea-related activities for 24 h prior to and during each testing day (i.e. preliminary measures, apnoeic and eupnoeic protocols). Centrifuge, London, UK). Plasma and blood volume changes for each post apnoeic time point were determined using the methods of Dill and Costill (1974). Prior to collecting any blood samples, the participant's fingers were cleaned and dried with a towel to avoid any influence of water on the results.

Familiarization session
Within 24 h of completing the baseline measurements, participants completed a familiarization session that introduced them to the apnoeic disciplines and testing environment. Participants were familiarized with the trial conditions,requirements and were introduced to the static apnoea position (i.e. seated position immersed up to the neck) and the dynamic apnoea technique (i.e. horizontal underwater breaststroke swimming).

Apnoea protocols
Within a week from completing the familiarization session, participants reported at the swimming pool (~28 • C). Participants entered the swimming pool without wearing any wet or dry suits and performed, on separate occasions (i.e. separated a week apart), one set of five maximal static or dynamic apnoeas with a 2 min seated rest between each apnoea.
Participants were instructed to hold their breath after a deep but not maximal inspiration, and both hyperventilation and lung packing were prohibited. Participants received a 1 min warning prior to commencing each apnoea, received a nose clip 30 s prior to the apnoea and a 10 s countdown was provided prior to immersing their face underwater and commencing their maximal apnoeic attempt.
During the static apnoea protocol the participants' heart rate and SpO 2 were monitored at 10 s intervals until 30 s post the termination of their maximal apnoeic attempt (Fagoni et al., 2017). During the dynamic apnoea protocol the participants' heart rate and SpO 2 were measured only up to 30 s after the termination of each maximal attempt, due to practical constraints. At the completion of each apnoea the participants' splenic volumes were assessed and a finger capillary blood sample was collected for the identification of haemoglobin, haematocrit and blood lactate concentrations. After each apnoea the participants underwent a two-minute resting period during which they were allowed to relax and breathe normally in a seated position whilst remaining immersed in water up to the waist. This procedure was repeated five times for each protocol, with apnoeic duration (static and dynamic protocols) and distance swam (dynamic protocol) measured during each maximal apnoeic attempt.

Control protocol
To control against any possible effects of whole-body immersion in water and diurnal variation on splenic volume and haematology, a control group performed a static eupnoeic (normal breathing) protocol.
The static eupnoeic protocol replicated the water exposure times, resting periods and data collection time points of the static apnoea protocol and replaced apnoeas with normal breathing periods. The static apnoea protocol was chosen to construct the static eupnoeic protocol as the water exposure periods were longer compared with the dynamic apnoea protocol.
Participants reported to the swimming pool testing site, at same time period as for the apnoea measurements and were immersed up to the neck level.

Statistical analysis
All participants completed the protocols successfully, and all data were statistically analysed using the SPSS statistics software v.21 (IBM, NY, USA). The Shapiro-Wilk test was used to assess normality, whereas homogeneity was assessed using Levene's test. Sphericity was assessed using Mauchly's test of sphericity; where the assumption of sphericity was violated, the Greenhouse-Geisser correction was applied. Repeated-measures ANOVA with Bonferroni post hoc contrast comparisons were used to assess differences between and within groups for baseline measurements and other collection time points for splenic volume, haemoglobin, haematocrit, blood lactate, HR, heart rate minimum (HR min ), SpO 2 , plasma volume and blood volume. Time to HR min was expressed as a relative percentage time course of static apnoeas for both groups (e.g. beginning of apnoea, 0% and termination of apnoea, 100%) and were compared using repeatedmeasures ANOVA with post hoc contrast comparisons to assess differences between and within groups. MANOVAs were used to assess differences in collection time points between groups (EBHD versus ND) and conditions (static versus dynamic). Pearson's correlation was used to assess for relationships between splenic volumes, performance levels, HR min and SpO 2 . Data were reported as mean ± SD. and significance was accepted at P < 0.05, and P = 0.000 was reported as P < 0.001. GraphPad Prism v.7.0c (GraphPad Software, CA, USA) was used to construct figures.

Apnoeic performances
The EBHD attained significantly longer (68%) static apnoeic durations during each successive maximal attempt (P < 0.001; Table 2). A significant difference between groups in the distance travelled during each dynamic apnoeic bout was observed (P < 0.001), with EBHD covering significantly greater distances (67%) at all time points compared with ND (Table 2).

Heart rate
A bradycardial response was evident in both groups during the static apnoea protocol, with an earlier response evident in the EBHD group, although this only approached significance (P = 0.067; Table 3). During the static apnoea protocol, the HR min was significantly lower in the EBHD group compared with the ND group (P = 0.001; Table 3). When static apnoeas were expressed as a biphasic percentage (i.e. beginning of apnoea, 0% and termination of apnoea, 100%) the time to HR min was not significantly different between groups (P = 0.086). There was a strong negative correlation (r = -0.98, P < 0.001) between the HR min and the static apnoea duration. There was also a strong negative correlation (r = −0.91, P < 0.001) between the time to HR min and the static apnoea duration.
For both groups, the end-apnoeic HR for each maximal apnoeic repetition was not different post the static apnoea protocol when compared with baseline (EBHD, P = 0.585; ND, P = 0.179) or when compared between groups (P = 0.585; Table 4). The end-apnoeic HR for maximal dynamic repetition was significantly higher than baseline for both groups post the dynamic apnoea protocol (EBHD, P < 0.001; ND, P < 0.001), however, there was no significant difference between groups (P = 0.342; Table 4). For both groups, end-apnoeic HR was higher post each successive dynamic apnoea attempt compared with the static apnoea protocol (EBHD, P < 0.001; ND, P < 0.001).

TA B L E 3
Heart rate responses to each successive maximal static apnoeic attempt Data are means ± SD. Significant (P < 0.05) between group differences are denoted as * . Abbreviations: EBHD, elite breath-hold divers; HR min , heart rate minimum; and ND, non-divers.   F I G U R E 2 Mean (±SD) end-apnoeic SpO 2 , relative volume of spleen, blood volume and plasma volume for dynamic apnoeas. Abbreviations: EBHD, elite breath-hold divers; ND, non-divers; and SpO 2 , peripheral oxygen saturation. Significance (P < 0.05) from baseline is denoted as * , between group differences are denoted as † (P < 0.05) the dynamic apnoea protocol versus control (P < 0.001) but not between the static apnoeas (statics versus control) (P = 0.231). The mean SpO 2 was significantly lower than baseline during each static apnoeic repetition in the EBHD group (P = 0.002), but not in the ND group (P = 0.176; Table 4). EBHD reached significantly lower SpO 2 levels at all static apnoeic repetitions than the ND group (P = 0.002; Table 4). Dynamic apnoeas induced a significant decrease in mean end-apnoeic SpO 2 from baseline in both groups (EBHD, P < 0.001; ND, P < 0.001), with the EBHD reaching lower SpO 2 at all apnoeic repetitions when compared with the ND group (P < 0.001; Table 4).

Static apnoea repetitions
When the end-apnoeic SpO 2 of EBHD and ND were compared between the apnoeic protocols, the dynamic apnoea protocol elicited significantly lower SpO 2 in EBHDs (P = 0.004) and ND (P < 0.001), respectively.

Spleen
When end-apnoeic splenic volumes were compared between the
Additionally, no differences were observed for either protocol when end-apnoeic haemoglobin concentrations were compared between groups (EBHD, P = 0.630; ND, P = 0.149), protocols (static, P = 0.406; dynamic, P = 0.102) or control intervention (P < 0.992;  Dynamics ND F I G U R E 3 Relationship between: (a) resting splenic volume and mean best static apnoeic performance for each participant, (b) resting splenic volume and mean best dynamic apnoeic performance for each participant, (c) average end-apnoeic SpO 2 and end-apnoeic splenic volume for each apnoeic repetition. Abbreviations: EBHD, elite breath-hold divers; ND, non-divers; and SpO 2 , peripheral oxygen saturation

Blood lactate
Mean end-apnoeic blood lactate concentrations were significantly higher than baseline for both groups during the static (EBHD, P < 0.001; ND, P < 0.001) and the dynamic apnoea protocols (EBHD, P < 0.001; ND, P < 0.001; Table 4). Significantly higher blood lactate concentrations were attained for both groups during the dynamic apnoea versus static apnoea protocol (EBHD, P < 0.001; ND, P < 0.001), with the EBHD achieving significantly higher lactate concentrations during both protocols compared with the ND (static, P = 0.008; dynamic, P = 0.004; Table 4).

Plasma and blood volume
Plasma volume and blood volume did not change for either protocol or group (P = 0.140; Figures 1 and 2).

DISCUSSION
This study made the first investigations into the splenic responses to a series of repeated maximal static and dynamic apnoeas with wholebody water immersion in EBHD and ND. The novel findings signify that relative to static apnoeas, dynamic apnoeas induced a stronger hypoxemic stress and this was associated with, (i) a higher end-apnoeic HR, (ii) a lower end-apnoeic SpO 2 , (iii) a higher blood lactate concentration and, (iv) a greater splenic contraction (i.e. in the EBHDs only), but with a similar erythrocyte release. EBHDattained greater apnoeic performances and reached lower SpO 2 than ND during both apnoeic protocols, but post-apnoeic splenic responses were similar across groups. These findings demonstrate that the magnitude of the splenic response is largely dictated by the magnitude of the hypoxemic stress encountered during apnoeic epochs.
An earlier bradycardic response and a significantly lower HR min were evident during the static apnoea protocol in the EBHD when compared with the ND. Interestingly, when time to HR min was reported as a relative biphasic percentage, a faster but not significantly different time to HR min was observed in the EBHD group compared with ND (Table 3). Our findings are in line with the literature (Ferretti et al., 1991;Lemaitre et al., 2005Lemaitre et al., , 2008 and provide further evidence that apnoeic training augments the magnitude of the apnoea-induced bradycardial response (Joulia et al., 2002(Joulia et al., , 2003Schagatay et al., 2000).
Additionally, we identified a significant strong negative correlation between static apnoeic durations and HR min (r = −0.98) and between static apnoeic durations and time to HR min (r = −0.91), which reinforces the relationship between apnoeic durations and the magnitude of the diving reflex-induced oxygen-conserving mechanism. Collectively, these findings point to a stronger diving reflex response and a more efficient oxygen-conserving mechanism in the EBHD than ND.
A lower end-apnoeic SpO 2 was evident in both groups during the dynamic apnoea protocol compared with static apnoeas. During both protocols, EBHD attained lower end-apnoeic SpO 2 levels than ND (Table 4). Our findings agree with Overgaard et al. (2006) observations but are contrary to Breskovic et al. (2011) that reported similar end-apnoeic SpO 2 post static (two repetitions) and dynamic apnoeas (one bout). These discrepancies might be attributed to the fundamental differences between the protocols utilised (i.e. number of apnoeic repetitions, pre-apnoeic breathing protocol, resting periods). Additionally, a higher blood lactate concentration was observed in both groups during the dynamic apnoea protocol compared with the static apnoea protocol in our study (Table 4). These findings suggest that the addition of contractile activity during apnoeic attempts upregulates the consumption of bodily oxygen stores and progressively increases the reliance on anaerobic metabolism, evidenced by the concurrent accumulation of lactate. Therefore, our study provides further evidence that maximal dynamic apnoeas induce a greater hypoxemic stress compared with maximal static apnoeas.
It is well accepted in the literature that the spleen plays an important role during apnoeic conditions, with its capacity to store oxygen-rich erythrocytes and release them into the systemic circulation during oxygen-deprived conditions (Hurford et al., 1990;Schagatay et al., 2001;Stewart & McKenzie, 2002). We observed a significant positive correlation between apnoeic performance levels and resting splenic volumes, which suggest that a larger splenic volume with capacity to hold a greater amount of erythrocytes is advantageous in an apnoeic context. These data signify that splenic size might serve as a strong predictor of apnoeic capabilities. Moreover, in line with earlier publications we failed to observe any between group differences in resting splenic volumes (Baković et al., 2003;Elia et al., 2019b;Prommer et al., 2007). Ilardo et al. (2018)  Thus, splenic size may be governed by a complex interplay between apnoeic training and genetics.
These observations are in agreement with earlier studies that assessed splenic responses following static apnoeas with or without face immersion in water (Bakovic et al., 2003;Schagatay, 2009).
When end-apnoeic splenic volumes were compared between the apnoeic protocols, significantly greater splenic contractions were only observed during the dynamic apnoea protocol in the EBHD. The spleen contains~98% sympathetic fibres and represents a constitutive part of the SNS (Stewart & McKenzie, 2002). In both mammals and humans, the spleen has been observed to contract in response to sympathetic nervous stimulation and hypoxia-induced increases in sympathetic output (Bakovic et al., 2013;Donald & Aarhus, 1974;Greenway, Lawson, & Stark, 1968;Hoka, Bosnjak, Arimura, & Kampine, 1989;Hurford et al., 1996;Stewart & McKenzie, 2002). Since the degree of hypoxemia is a potent stimulus for evoking splenic contractions, the lower SpO 2 attained by the EBHD during the dynamic apnoea protocol compared with ND would have served as a stronger stimulus for evoking splenic contractions. Thus, providing a partial reasoning for the greater splenic volume contractions observed in EBHD in response to the dynamic apnoea protocol.
During maximal apnoeic epochs the human body is subjected to extreme chemoreflex stimulations, with a number of studies noting significant increases in arterial blood pressure and carbon dioxide (CO 2 ) levels (Breskovic et al., 2011;Joulia et al., 2002Joulia et al., , 2003Sieber et al., 2009). In addition, as a consequence of sustaining longer apnoeic durations, EBHD are subjected to a greater degree of hypercapnic (i.e. higher end-apnoeic arterial CO 2 levels) and hypoxemic stress (i.e. lower end-apnoeic arterial O 2 levels) compared with ND (Breskovic et al., 2012;Joulia et al., 2002;Willie et al., 2015). Interestingly, Richardson et al. (2012) demonstrated that hypercapnia (i.e. prebreathing 5% CO 2 in O 2 ) facilitated a greater degree of splenic contractions during a series of three repeated maximal static apnoeas (−33% from control) compared with hypocapnia (+13%), normocapnia (−9%) and eupneic hypercapnia (+30%) at similar end-apnoeic arterial haemoglobin saturation levels. Accentuating that hypercapnia, acts as an independent stimulus for invoking splenic contractions-likely through interacting with central medullary and peripheral carotid body chemoreceptors . Therefore, the greater splenic contractions observed in our EBHD group during the dynamic apnoea protocol (i.e. compared with the static apnoea protocol) may indicate that this group was exposed to a greater magnitude of chemoreflex stress than the ND group, which consequently served as a stronger stimulus for evoking splenic contractions. However,since we did not evaluate end-apnoeic arterial CO 2 or blood pressure levels, we are unable to fully elucidate the underlining mechanisms that dictated these group differences and thus further rationalize our findings.
To the best of our knowledge, this is the first study to assess the splenic responses to a series of repeated maximal dynamic apnoeas performed by either EBHD or ND. Our study demonstrated that dynamic apnoeas elicited, in both groups, splenic contraction and this was associated with a significant increase in haemoglobin concentration.Since no plasma or blood volume changes were reported during the dynamic apnoea protocol, it can be reasoned that the significant increases in haemoglobin concentrations were likely derived from the dynamic apnoea-associated splenic contractions and not evoked by water immersion or haemoconcentration. Interestingly, our EBHD groups' post-apnoeic haemoglobin increases (+5 g/L , +4%; haematocrit unchanged) are greater than those previously reported in divers by Schagatay, Andersson, and Nielsen (2007)  However, our values are lower than those reported by Hurford et al. (1990) in Korean Ama divers (+11 g/L, +9.5%; +3.6%, +10.5%) after a routine diving shift (174 ± 46 min, depths of ∼5-7 m). An explanation for the higher haemoglobin and haematocrit concentrations reported by Hurford et al. (1990) could be dehydration, hypovolaemia or extravascular volume displacement in connection with prolonged exercise and insufficient hydration (Harrison et al., 1986). Similarly our ND group's haemoglobin increases (+8 g/L, +4.90%; haematocrit unchanged) were greater than those reported in untrained individuals by Richardson et al. (2005)  we were unable to collect venous blood samples. Thus, the presently recorded post-apnoeic haemoglobin concentrations (i.e. from fingertip sampling) might be an underestimation of the true magnitude of the haematological fluctuations induced by the splenic response.
In conclusion, the present study demonstrated that repeated maximal static and dynamic apnoeas with whole-body immersion are effective in stimulating splenic contractions in both ND and EBHD. Moreover, dynamic apnoeas, in comparison with static apnoeas, elicited greater splenic contractions in EBHD only. In addition, haemoglobin increases were only observed following the dynamic apnoea protocol in both ND and EBHD, whereas haematocrit concentrations were unchanged across groups and apnoeic protocols.
Lastly, our findings signify that the magnitude of the apnoea-induced splenic response is largely dictated by the degree of the hypoxemic stress experienced during apnoeic epochs.