Journal Logo


Cerebral Blood Flow Responses to Aquatic Treadmill Exercise


Author Information
Medicine & Science in Sports & Exercise: July 2017 - Volume 49 - Issue 7 - p 1305-1312
doi: 10.1249/MSS.0000000000001230
  • Free


Aquatic-based treadmill (ATM) exercise allows for lower impact and increased resistance in comparison to land-based treadmill (LTM) exercise, therefore achieving decreased musculoskeletal loading of joints and providing potential for enhanced acute and chronic physiological adaptations (5). The increased hydrostatic pressure applied to the human body upon immersion in water leads to centralizing blood distribution within the body, which enhances cardiac performance and therefore increases tissue perfusion (32). Furthermore, the mechanical unloading and support of bodyweight due to buoyancy mean that ATMs are a useful tool in gait reeducation. Indeed, ATM exercise is used in the rehabilitation of spinal cord injuries (36) and stroke (28) patients. Positive effects of ATM exercise have also been noted in populations with joint conditions such as osteoarthritis (7) and those with coronary heart disease (13). At the other end of the spectrum, athletes use ATM exercise to maintain cardiorespiratory fitness while reducing the mechanical load when recovering from injury (30). Therefore, ATM exercise represents an effective form of therapy and rehabilitation for a range of healthy and diseased populations. Studies to date comparing physiological responses between ATM and LTM have mostly focused on cardiorespiratory responses (e.g., [17,35]), whereas no study has examined the effect that ATM exercise has on cerebrovascular responses and therefore explored the possibility of how this mode of therapy may optimize exercise-induced, stimulus–response adaptations leading to improved cerebrovascular function and ultimately brain health (11).

The effective regulation of blood flow to and within the brain is vital for optimal brain function. Regular exercise and higher cardiorespiratory fitness has been positively linked with cerebral blood flow (CBF) and its regulation (e.g., [3,6]), shown to offset the natural age-related decline in CBF (1), and reduce risk of neurodegenerative disease (e.g., dementia [21]). However, the mechanisms that underpin the neuroprotective benefits of exercise are yet to be established, meaning that the effectiveness of various exercise parameters such as mode, intensity, and duration are not yet understood (22). One suggested mechanism for exercise-induced improvements in vascular function is via shear stress–mediated increases in endothelium nitric oxide-dependent vasodilation of blood vessels (18,39), as a consequence of the recurrent mechanical force of blood flow on the walls of the arteries (i.e., shear stress) (27). Numerous studies (e.g., [9,15,19]) have reported the functional adaptation and structural changes to the vasculature that occur as a result of long-term exercise (18), albeit primarily provided from animal-based and cell culture studies (see [8]) or within the peripheral vasculature of humans. Researchers have begun to explore alternative methods (e.g., heat therapy [12,38]) that target this mechanism to improve vascular function. Extrapolating this to the brain, conditioning strategies that increase blood flow, either in combination or independent of exercise, may enhance shear stress–mediated adaptation of the cerebrovasculature. This has the potential to directly improve CBF and its regulation and therefore be used in the prevention and treatment of neurovascular disease.

Given the known physiological responses to water immersion, water-based activities may be one such strategy. Indeed, two studies have explored this possibility, examining CBF responses in water at rest (12) and during a box-stepping exercise protocol (31). Carter et al. (12) reported a positive correlation between middle and posterior CBF velocities during resting water immersion, linking the increase in CBF velocity during immersion to an increase in mean arterial pressure and arterial carbon dioxide content. From the same group, Pugh et al. (31) compared a 20-min bout of matched low-intensity stepping exercise (HR ≤ 100 bpm) in a water tank to that on land, finding CBF velocities to be augmented in water. These studies illustrate the potential for enhanced shear stress–mediated vascular adaptation by exercising in water, although important questions remain unanswered relating to the CBF profile during different exercise intensities and at different depths of water immersion. Such questions can be addressed using an ATM, which is already an established rehabilitation tool for several conditions (see above). Therefore, the primary purpose of this study was to compare changes in CBF (velocity) and HR responses during an incremental exercise test using an ATM and an LTM and to examine CBF and HR responses at different levels of immersion (midthigh, iliac crest [hip], and xiphoid process [midchest]) during ATM exercise. We hypothesized that 1) ATM exercise would augment the CBF response to an incremental intensity exercise test compared with LTM exercise and 2) increased water immersion would maintain CBF while lowering HR.



Eleven healthy participants (seven females and four males, age = 27 ± 5 yr) were recruited for this study, which was approved by the University of Birmingham Science, Technology, Engineering and Mathematics Ethical Review Committee, and performed in accordance with the Declaration of Helsinki. After providing their written informed consent, all participants completed a general health questionnaire during an initial visit to the laboratory and declared that they were free of any cardiovascular, cerebrovascular, or respiratory disease; were not taking medication (not including contraceptive medications); or did not have injuries that would preclude treadmill-based exercise.

Study Design and Protocol

After the initial screening visit, participants completed exercise sessions on both an LTM and an ATM in a randomized, counterbalanced order. Each session lasted approximately 1 h, of which 15–25 min was exercise. First, participants completed incremental exercise starting from walking pace (4 km·h−1, immersed to iliac crest [aquatic], 6 km·h−1 [land]) and increasing 1 km·h−1 every 2 min up to 10 km·h−1 for aquatic (maximum belt speed) or 12 km·h−1 for land. On land, participants were then ramped to exhaustion (increased gradient 2° every min), whereas for aquatic exercise, participants completed two 2-min bouts of exercise immersed to midthigh and midchest at constant submaximal speed. During exercise at the different immersion depths on the ATM, the speed was held constant; eight participants completed this protocol at 10 km·h−1, whereas for three participants, who reached near maximal HR during the initial incremental intensity protocol, the treadmill speed was lowered (to 8 km·h−1 for 2 participants and 5 km·h−1 for 1 participant) and remained at this speed during the exercise at different immersion depths. Figure 1 provides a schematic for the ATM and LTM protocols.

Schematic outlining treadmill belt speed, land treadmill gradient, and ATM water depth for the ATM and LTM exercise protocols. Participants completed seven 2-min stages of incremental exercise intensity, induced via 1 km·h−1 increases in treadmill belt speed. Participants were then ramped to exhaustion (2° every minute) for the land-based protocol or completed two 2-min stages of exercise at two alterative water depths on the ATM. Numbers of participants completing each stage of the land-based protocol through to exhaustion are shown, along with the numbers of participants at each submaximal aquatic belt speed for the 2-min stages of different immersion depths. Room temperature was maintained at ~21°C for both, whereas water temperature was 32°C.

There were at least 48 h between sessions for each participant, with the majority of participants completing their second session within 2 wk of the first. Because of the timing of access to the ATM facility, the phase of menstrual cycle for female participants was not controlled for between exercise sessions. Before each session, participants were asked to refrain from eating a large meal for 4 h before arrival, although a light meal was permitted up to 2 h before arrival. To ensure adequate hydration status, participants were advised to drink 0.5 L of water within 4 h of beginning testing and 0.25 L of water within 15 min of testing, in accordance with the American College of Sports Medicine hydration guidelines (2). Participants were also asked to refrain from caffeine for 6 h before testing and refrain from vigorous exercise and the consumption of alcohol for 24 h before testing.

Equipment and Measurements

Exercise treadmills

A standard treadmill ergometer (Pulsar, H-P-Cosmos, Germany) was used for the land-based exercise protocol. The aquatic exercise session was completed on an ATM (FOCUS; HYDRO PHYSIOTM, Shropshire, UK) at the Optispine Physiotherapy Clinic in Birmingham, UK.

CBF velocity and HR measures

Bilateral blood flow velocity in the left and right middle cerebral arteries (MCAv) was measured using a 2-MHz transcranial Doppler (TCD) ultrasound system (Dopplerbox, DWL; Compumedics LTD, Germany), in accordance with search techniques described elsewhere (40). The two ultrasound probes were placed above each zygomatic arch on the left and right side of the head and secured via an adjustable headband that maintained a constant insonation angle throughout the testing session. A small amount of ultrasound gel was placed between the probe and the skin to obtain the highest quality images. The reliability of measuring CBF using TCD is operator dependent; thus, all measurements were taken by the same experienced sonographer (SJEL), with photographs of probe placement, and depth and filter settings recorded and kept constant between exercise sessions for each participant. Cerebrovascular data were acquired continuously via an analog-to-digital converter (PowerLab 8/30, ML870; ADInstruments, Dunedin, New Zealand) at 1 kHz. Data were displayed in real time and recorded for offline analysis using commercially available LabChart Pro software (v7, ADInstruments).

HR was monitored using telemetry (Polar, Finland) via a belt fitted around the chest of the participant, as well as derived from the beat-by-beat MCAv waveform. Steady-state measures for HR were recorded at each stage of the incremental protocols and for the different immersion depths.

Data analysis and statistical approach

Mean values for MCAv and HR at each 2-min stage were determined using an average of the final 30 s of each stage and used to calculate change from resting (seated) baseline for each measured time point. Because the ATM had a maximum belt speed of 10 km·h−1, we intended to use the initial 14 min of each protocol to compare the responses between the ATM and the LTM exercises. We also independently compared two exercise intensities, specifically walking and the recommended public health guideline of moderate exercise intensity (65% V˙O2max), which was from estimated from HR measures. Walking pace in the ATM was 4 km·h−1, whereas on the LTM, it was 6 km·h−1, but both represented the only speed for which all participants were walking; i.e., participants started jogging in the ATM at 5 km·h−1. The 65% V˙O2max intensity was estimated at 79% HRmax (37), and treadmill speeds that induced an HR response closet to this target were selected (range for land: 7–11 km·h−1, range for aquatic: 5–10 km·h−1).

Two-way repeated-measures ANOVA were used to compare changes in MCAvmean and HR across the incremental protocol (time × treadmill) and at walking and 65% V˙O2max intensity (intensity × treadmill), whereas a one-way ANOVA compared changes in MCAvmean and HR across the three different immersion levels while running at the same speed in the ATM. Post hoc comparisons were done using pairwise comparisons (Bonferroni corrected) to show where effects occurred. Paired t-tests were used to determine whether significant differences existed between comparable data sets of interest (e.g., resting and peak HR/MCAv, MCAv during walking vs 65% V˙O2max). All statistical analysis was conducted using the Statistical Package for the Social Sciences (version 22; SPSS Inc., Chicago, IL), with a priori statistical significance set at P ≤ 0.05. Data are presented as mean ± SD.


All 11 participants who began the exercise sessions completed both protocols. All eleven participants reached the maximum belt speed in the ATM (10 km·h−1), therefore completing all 14 min of incremental aquatic exercise. For the LTM protocol, one participant stopped at the completion of the 10-km·h−1 stage (at 10 min), and two participants stopped after the 11-km·h−1 stage (at 12 min) because of reaching voluntary exhaustion. Consequentially, the two-way ANOVA for the comparison of incremental exercise protocols used the first five stages of exercise for which all 11 participants had paired data sets for (as indicated in Fig. 1).

There was no significant difference (P = 0.79) between left and right MCAvmean in participants that had a TCD signal on both sides throughout testing in both sessions (n = 6); therefore, data were pooled and presented as a combined mean value. In four aquatic sessions and one land-based session, the TCD signal on one side was either lost during exercise or not found initially; for these trials, the remaining side was used as the mean value.

There was a small, but significant, difference for baseline resting (seated) MCAvmean between land and aquatic testing sessions (70 ± 9 cm·s−1 vs 66 ± 9 cm·s−1, respectively; P = 0.023), whereas resting HR was similar (70 ± 13 vs 69 ± 14 bpm, P = 0.738).

Effects of increased exercise intensity on cerebrovascular and HR responses

As illustrated in Figure 2A, there was a significant main effect of time (P = 0.004) and treadmill type (P = 0.003) on the change in MCAvmean from baseline over the initial 10 min of each protocol. The pooled difference across the 10 min for MCAvmean between the treadmill protocols was ~6 cm·s−1, with the largest difference occurring at the 4-min stage (~11 cm·s−1). The 4-min stage also represented the peak change in MCAvmean from baseline in the water (~16 cm·s−1), which was maintained to within 3 cm·s−1 for the remainder of the protocol. On land, however, the peak change in MCAvmean (~12 cm·s−1) did not occur until the 10th minute. Nevertheless, this difference in the pattern of increase did not reach statistical significance (interaction effect: P = 0.073).

Changes in MCAvmean (A) and HR (B) from resting (seated) baseline values over the initial 10 min of ATM and LTM exercise protocols. Data are means ± SD. *Significant difference between treadmills. #Significant difference between preceding stage.

For HR, there was an interaction effect for the change in HR across time (P = 0.020). Post hoc analysis revealed that HR increased for each incremental stage except for the transition between 5 and 6 km·h−1 (minutes 4 and 6) on the ATM, and HR was significantly higher on the LTM than the ATM except for at 4 min (7 and 5 km·h−1, respectively, see Fig. 2B). Overall, and in contrast to the MCAvmean observations, HR was higher with land-based exercise compared with aquatic exercise (pooled difference: ~11 bpm greater for land; main effect: P = 0.028). Further, the peak MCAvmean during the aquatic incremental protocol tended to be at a lower percentage of HRmax (determined during the land-based protocol) compared with land-based incremental exercise (75% ± 12% vs 84% ± 15% of HRmax for aquatic and land, respectively; P = 0.069).

Figure 3 shows the comparison between walking and moderate intensity running (at 65% V˙O2max) on each treadmill for MCAvmean and HR responses. Both walking and running at 65% V˙O2max elicited a greater increase in MCAvmean during ATM as compared with LTM (main effect: P = 0.003). Interestingly, although there was a main effect of intensity (P = 0.022) for MCAvmean, subsequent analysis revealed that while MCAvmean increased similarly within each treadmill modality (interaction effect: P = 0.628), there was no difference between MCAvmean for ATM walking and LTM running at 65% V˙O2max (paired t-test: P = 0.563).

Mean change in MCAvmean (left panel) and HR (right panel) from resting baseline for walking and moderate intensity (65% V˙O2max) running exercise using LTM and ATM. Data are means ± SD. N = 11. *Significant difference between treadmills. #Significant difference between preceding stage (i.e., walking).

Collectively, these data indicate that water-based exercise across a range of intensities stimulates greater increases in MCAv for a relatively lower HR response compared with land-based exercise, and that water-based walking elicits a similar increase in blood flow (velocity) as running on land at 65% V˙O2max.

Effect of immersion level on cerebrovascular and HR responses

Figure 4 illustrates that HR decreased with greater levels of water immersion on the ATM while the treadmill belt speed remained constant. Post hoc analysis showed that the mean decrease in HR from the water level at midthigh to iliac crest was ~18 bpm (P = 0.001), and from iliac crest to xiphoid process was ~21 bpm (P = 0.002). This was in contrast to MCAvmean, with the change from resting baseline not different between immersion levels (P = 0.371). Finally, the 2-min exercise bout at midthigh water depth elicited near maximal HR (95% ± 5% of HRmax; see Fig. 4).

Changes in middle CBF velocity (MCAvmean) and HR from resting (seated) baseline values during constant speed ATM exercise immersed to midthigh, iliac crest, and xiphoid process. Data are means ± SD for 11 participants. *Different from midthigh (P < 0.05). †Different from iliac crest (P < 0.05).


The aim of this study was to examine CBF responses during incremental exercise on an ATM as compared with an LTM, and while exercising at different levels of water immersion on the ATM. Our main novel findings were that: 1) MCAvmean was augmented during ATM exercise compared with LTM exercise across the range of exercise intensities tested, and this augmented MCAvmean was associated with a relatively lower HR response; 2) walking on an ATM elicited a similar increase in MCAvmean to that of running at moderate intensity (65% V˙O2max) on land; and 3) immersion depth altered HR while maintaining MCAvmean during exercise at a constant ATM speed. Collectively, these data indicate that ATM exercise augments CBF. Further, although we have not quantified differences in shear stress per se, the elevated flow velocity demonstrates the potential for ATM exercise to enhance shear stress–mediated cerebrovascular adaptation and thus optimize exercise-induced adaptations linked with improved brain health.

The findings of the current study are consistent with previous research reporting elevated MCAv during exercise (e.g., [10,23]), and an augmented MCAv response when in water (observed at rest and during light intensity exercise [12,31]). Here, we show for the first time that this augmented MCAv response occurs across a range of exercise intensities and that MCAv can be maintained while exercising at lower intensities with greater depths of water immersion. One notable observation was that walking on an ATM elicited a similar increase in MCAv (~10 cm·s−1) to that of running on land at the exercise intensity promoted by current public health guidelines (i.e., 65% of aerobic capacity for 150 min·wk−1). Furthermore, the profile of exercise-induced changes in MCAv was different between the protocols, with ATM exercise producing maximal gains in MCAv within 4 min of starting the protocol, a time point that also represented the greatest difference between treadmill protocols, whereas MCAv during LTM exercise increased linearly across the incremental protocol yet remained lower (see Fig. 2).

It is widely reported that the greatest exercise-induced elevation in CBF is achieved at moderate exercise intensity (~65% V˙O2max), as above this threshold CBF will decrease back toward resting values as a result of hyperventilation-induced cerebral vasoconstriction due to lower PaCO2 (24). Our data indicate that ATM exercise may have a different exercise-induced CBF profile to this commonly reported profile, most of which come from cycling-based exercise protocols. The findings of the current study indicate that optimal CBF gains may be achieved at lower exercise intensities in water than on land (see Fig. 2A), and even at higher exercise intensities (induced via less water immersion), MCAv is consistently elevated above resting measures (see Fig. 4). On the basis of these findings, it could be suggested that changes in arterial carbon dioxide above anaerobic threshold during ATM exercise have less influence on CBF relative to other factors involved in CBF regulation (e.g., blood pressure and cardiac output). Unfortunately, we were unable to measure end-tidal PCO2 during our study because of equipment unavailability. Thus, we can only speculate about the relation between PCO2 and CBF during ATM exercise. Further research is needed to determine the influence of this key regulator of CBF during ATM exercise of increasing exercise intensity.

The regulation of CBF during exercise is multifactorial and complex (25), with an integrative combination of exercise-induced changes in brain metabolic and neuronal activity, blood pressure, cardiac output, and arterial PCO2 all likely to contribute to changes in cerebral perfusion during any exercise paradigm. On the basis of previous water immersion studies (12,31), an elevated PCO2 likely explains some of the difference in MCAvmean with aquatic exercise as compared with land-based exercise. Further, given the linear relation between cardiac output and CBF (26), another likely contributor to the augmented blood flow velocity is related to the well-documented increases in cardiac output during water immersion (29) because of the effects of increased hydrostatic pressure centralizing blood within the trunk and increasing stroke volume (5). Interestingly, an elevated stroke volume would appear to be the key mediator of this increased cardiac output because a reduction in HR (as we observed) during water immersion is also well documented (4,16,20). In contrast, Pugh et al. (31) reported no significant difference in HR between water and land during their low-intensity exercise. However, it is worth noting that the box-stepping exercises in their study were matched for HR between land and water protocols to compare similar intensities. Importantly, regardless of the mechanisms regulating CBF during this modality of exercise, our observed differences in exercising MCAv between our treadmill protocols demonstrates that ATM exercise produces higher blood flow velocity across a range of intensities, and particularly so at lower exercise intensities (i.e., walking/light jogging), thus illustrating the potential for an enhanced shear stress–mediated pathway for cerebrovascular adaptation following repeated exposure (i.e., training).

The decrease in HR associated with increasing immersion levels noted in this study is supported by previous studies that have reported a continuous decrease in HR from hip level up to head-out immersion in water at rest (4,20). Our findings illustrate that a similar elevation in MCAvmean can be achieved with greater water immersion for a comparatively lower HR. As such, ATM exercise training at higher levels of water immersion could optimize shear stress–mediated vasculature adaptations, while lowering the risk of a cardiac event in populations with elevated risk.

The water temperature used in this study (32°C) is representative of conditions regularly used in rehabilitative therapy, and is within 3°C of the temperatures used in previous studies. Although changes in water temperature have been reported to translate into changes in the cardiac response (5), the relatively small variation (<2°C) in temperature between this study and recent research (12,31) is unlikely to impact the relative changes in MCAvmean and HR observed here. Another consideration is the different heat conduction capacities of water versus air, which may differentially alter exercise-induced changes in body core temperature. As such, measures of body core temperature would be of value in future studies to assess differences between modalities and potential effects of thermal stress–related adaptations during the ATM exercise. Indeed, one further possibility for this form of exercise therapy is to alter the water temperature to investigate the potential additive therapeutic effect of thermal stress, which may further optimize the stimulus–response interaction and promote greater neuroprotection against neurodegenerative diseases (11).


Speed limitations of the ATM prevented a full comparison between treadmill modalities for an incremental test to exhaustion. Nevertheless, both protocols started at a walking pace and increased at the same rate (1 km·h−1 every 2 min), which resulted in a similar rate of increase in HR and therefore allowed for a meaningful comparison between the ATM and the LTM exercises across a range of exercise intensities. This study design meant we were unable to compare matched HR responses across all intensities. We acknowledge that the differences in cardiorespiratory responses may influence the absolute values we show here, but ultimately will not affect the pattern of MCAv that we observed across the range of exercise intensities we tested. Measurements of V˙O2 were originally planned in addition to HR to further quantify the cardiorespiratory strain and energy expenditure during both protocols, but this was not possible due to equipment unavailability. However, similar decreases in HR and V˙O2 between land running and deep and shallow water running have previously been noted (16), indicating that HR alone can adequately reflect measures of exercise intensity on land as compared with in water. It is also acknowledged that the reduction in HR alone does not necessarily reflect a reduction in cardiac work (as reflected by myocardial V˙O2). Given the linear relationship that V˙O2 and cardiac output share (14), cardiac work can be indexed via the combined measures of HR, stroke volume, and blood pressure (i.e., HR × SV × systolic BP [or MAP] [33]). However, we chose not to fit a blood pressure measuring device (e.g., finometer) so that participants could perform the exercise in the water as naturally as possible (i.e., fitting this device would have required them to hold their arm up out of the water). In addition to providing a measure of stroke volume (e.g., via Beatscope software) to determine cardiac work, measures of BP would have also provided insightful data regarding the influence of blood pressure on CBF for these different exercise modalities. Nevertheless, our primary question was to examine the CBF (velocity) differences between these modalities across a range of exercise intensities and at different immersion depths. The mechanism(s) that underpin these differences were not the primary focus of our study, but based on previous studies (12,32), differences in BP (and PETCO2) would likely be involved.

As mentioned earlier, we were also unable to measure end-tidal gas content and therefore assess the influence of PETCO2 on MCAv during our exercise protocols. Nevertheless, given the earlier peak in MCAvmean during the initial incremental protocol on the ATM and the consistently elevated MCAvmean during near maximal exercise in water (see Figs. 2 and 4), the typical influence of changes in PaCO2 on CBF during exercise would appear to be different for ATM exercise. Future studies that include measures of both V˙O2 and PETCO2 are needed to confirm the hypothesis that there may be an altered relation between CBF and PCO2 during ATM exercise.

We measured blood flow velocity using TCD as an index of CBF. The validity of this approach and the likelihood of vessel diameter changes affecting interpretations of these measures should be considered, especially given the likely changes in blood pressure and PCO2 associated with exercise. Nevertheless, given the differential pattern of MCAv changes between the ATM and the LTM protocols while changes in PETCO2 and BP were likely similar (albeit elevated in water [12,31]), it seems unlikely that changes in MCA diameter would affect the interpretation of the findings here. Further, TCD is the ideal brain imaging tool to use in this setting, whereas other approaches are not feasible or realistic (e.g., MRI or duplex Doppler).

Finally, the relatively small and demographically limited sample population (mostly young university students) should be taken into account. As such, whether similar responses occur in older and clinical populations remain to be determined.


On the basis of the findings of the current study, ATM exercise could provide an ideal exercise modality to maximize the stimulus–response for shear stress–mediated adaptation of the cerebrovasculature in clinical and nonclinical populations. Research is now needed to establish whether this augmented acute response translates into permanent adaptation of the cerebrovasculature and how such training may improve other aspects of brain structure and function.

Exercise training is recommended in clinical populations with elevated risk of neurodegenerative disease to aid rehabilitation (e.g., stroke [34]); however, physical disability may affect on the effectiveness of traditional exercise programs to improve vascular health via shear stress–mediated adaptation. Our findings demonstrate the potential for ATM exercise to optimize this stimulus for vascular adaptation at exercise intensities (e.g., walking) that are feasible for clinical populations with impaired physical function (e.g., stroke survivors). As such, the utility of ATMs for brain-targeted exercise training may be another important reason to promote such a rehabilitation approach.


ATM exercise augments CBF velocity across a range of intensities, and particularly so at lower exercise intensities (i.e., walking/light jogging). This elevated blood flow has the potential to enhance shear stress–mediated cerebrovascular adaptation and thus optimize exercise-induced adaptations linked with improved brain health. Research is now needed to establish whether this augmented acute response translates into permanent adaptation of the cerebrovasculature, and how such training may improve other aspects of brain structure and function.

This study was funded by the School of Sport, Exercise and Rehabilitation Sciences, University of Birmingham. The authors extend thanks to the participants for taking part in the study and to Dr. Rebekah Lucas for her feedback and critical insight during the preparation of the manuscript. They also thank Mr. Davinder Chatha for making his water treadmill available to conduct this study. There were no conflicts of interest. R. P., M. H., and S. J. E. L. were involved in the conception and design of research. R. P. and S. J. E. L. conducted the experiment, performed data analysis and interpretation, and contributed to the drafting the manuscript. M. H. provided critical review of the manuscript. All authors approved the final version of this manuscript. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The authors also acknowledge that the results of the present study do not constitute endorsement by the American College of Sports Medicine.


1. Ainslie PN, Cotter JD, George KP, et al. Elevation in cerebral blood flow velocity with aerobic fitness throughout healthy human ageing. J Physiol. 2008;586(16):4005–10.
2. American College of Sports Medicine. Selecting and Effectively Using Hydration for Fitness. 2011. Available from:
3. Bailey DM, Marley CJ, Brugniaux JV, et al. Elevated aerobic fitness sustained throughout the adult lifespan is associated with improved cerebral hemodynamics. Stroke. 2013;44(11):3235–8.
4. Barbosa TM, Garrido MF, Bragada J. Physiological adaptations to head-out aquatic exercises with different levels of body immersion. J Strength Cond Res. 2007;21(4):1255–9.
5. Barbosa TM, Marinho DA, Reis VM, Silva AJ, Bragada JA. Physiological assessment of head-out aquatic exercises in healthy subjects: a qualitative review. J Sports Sci Med. 2009;8(2):179–89.
6. Barnes JN, Taylor JL, Kluck BN, Johnson CP, Joyner MJ. Cerebrovascular reactivity is associated with maximal aerobic capacity in healthy older adults. J Appl Physiol (1985). 2013;114(10):1383–7.
7. Bartels EM, Lund H, Hagen KB, Dagfinrud H, Christensen R, Danneskiold-Samsøe B. Aquatic exercise for the treatment of knee and hip osteoarthritis. Cochrane Database Syst Rev. 2007;4:CD005523.
8. Bolduc V, Thorin-Trescases N, Thorin E. Endothelium-dependent control of cerebrovascular functions through age: exercise for healthy cerebrovascular aging. Am J Physiol Heart Circ Physiol. 2013;305(5):H620–33.
9. Brendle DC, Joseph LJO, Corretti MC, Gardner AW, Katzel LI. Effects of exercise rehabilitation on endothelial reactivity in older patients with peripheral arterial disease. Am J Cardiol. 2001;87(3):324–9.
10. Brugniaux JV, Marley CJ, Hodson DA, New KJ, Bailey DM. Acute exercise stress reveals cerebrovascular benefits associated with moderate gains in cardiorespiratory fitness. J Cereb Blood Flow Metab. 2014;34(12):1873–6.
11. Burley CV, Bailey DM, Marley CJ, Lucas SJ. Brain train to combat brain drain; focus on exercise strategies that optimise neuroprotection. Exp Physiol. 2016;101(9):1178–84.
12. Carter HH, Spence AL, Pugh CJ, Ainslie P, Naylor LH, Green DJ. Cardiovascular responses to water immersion in humans: impact on cerebral perfusion. Am J Physiol Regul Integr Comp Physiol. 2014;306(9):R636–40.
13. Choi JH, Kim BR, Joo SJ, et al. Comparison of cardiorespiratory responses during aquatic and land treadmill exercise in patients with coronary artery disease. J Cardiopulm Rehabil Prev. 2015;35(2):140–6.
14. Damato AN, Galante JG, Smith WM. Hemodynamic response to treadmill exercise in normal subjects. J Appl Physiol. 1966;21(3):959–66.
15. Dinenno FA, Tanaka H, Monahan KD, et al. Regular endurance exercise induces expansive arterial remodelling in the trained limbs of healthy men. J Physiol. 2001;534(1):287–95.
16. Dowzer CN, Reilly T, Cable NT, Neville A. Maximal physiological responses to deep and shallow water running. Ergonomics. 1999;42(2):275–81.
17. Fujishima K, Shimizu T. Body temperature, oxygen uptake and heart rate during walking in water and on land at an exercise intensity based on RPE in elderly men. J Physiol Anthropol Appl Human Sci. 2003;22(2):83–8.
18. Green DJ, Maiorana A, O'Driscoll G, Taylor R. Effect of exercise training on endothelium-derived nitric oxide function in humans. J Physiol. 2004;561(1):1–25.
19. Hambrecht R, Adams V, Erbs S, et al. Regular physical activity improves endothelial function in patients with coronary artery disease by increasing phosphorylation of endothelial nitric oxide synthase. Circulation. 2003;107(25):3152–8.
20. Kruel LF, Peyré-Tartaruga LA, Coertjens M, Dias AB, Da Silva RC, Rangel AC. Using heart rate to prescribe physical exercise during head-out water immersion. J Strength Cond Res. 2014;28(1):281–9.
21. Lautenschlager NT, Cox K, Cyarto EV. The influence of exercise on brain aging and dementia. Biochim Biophys Acta. 2012;1822(3):474–81.
22. Lucas SJE, Cotter JD, Brassard P, Bailey DM. High-intensity interval exercise and cerebrovascular health: curiosity, cause and consequence. J Cereb Blood Flow Metab. 2015;35(6):902–11.
23. Moraine JJ, Lamotte M, Berré J, Niset G, Leduc A, Naeije R. Relationship of middle cerebral artery blood flow velocity to intensity during dynamic exercise in normal subjects. Eur J Appl Physiol Occup Physiol. 1993;67(1):35–8.
24. Ogoh S, Ainslie PN. Cerebral blood flow during exercise: mechanisms of regulation. J Appl Physiol (1985). 2009;107(5):1370–80.
25. Ogoh S, Ainslie PN. Regulatory mechanisms of cerebral blood flow during exercise: new concepts. Exerc Sport Sci Rev. 2009;37(3):123–9.
26. Ogoh S, Brothers RM, Barnes Q, et al. The effect of changes in cardiac output on middle cerebral artery mean blood velocity at rest and during exercise. J Physiol. 2005;569(2):697–704.
27. Papaioannou TG, Stefanadis C. Vascular wall shear stress: basic principles and methods. Hellenic J Cardiol. 2005;46(1):9–15.
28. Park SW, Lee KJ, Shin DC, Shin SH, Lee MM, Song CH. The effect of underwater gait training on balance ability of stroke patients. J Phys Ther Sci. 2014;26(6):899–903.
29. Pendergast DR, Moon RE, Krasney JJ, Held HE, Zamparo P. Human physiology in an aquatic environment. Compr Physiol. 2015;5(4):1705–50.
30. Prins J, Cutner D. Aquatic therapy in the rehabilitation of athletic injuries. Clin Sports Med. 1999;18(2):447–61.
31. Pugh CJ, Sprung VS, Ono K, et al. The effect of water immersion during exercise on cerebral blood flow. Med Sci Sports Exerc. 2015;47(2):299–306.
32. Reilly T, Dowzer CN, Cable NT. The physiology of deep-water running. J Sports Sci. 2003;21(12):959–72.
33. Rooke GA, Feigl EO. Work as a correlate of canine left ventricular oxygen consumption, and the problem of catecholamine oxygen wasting. Circ Res. 1982;50(2):273–86.
34. Saunders DH, Sanderson M, Hayes S, et al. Physical fitness training for stroke patients. Cochrane Database Syst Rev. 2016;3:CD003316.
35. Silvers WM, Rutledge ER, Dolny DG. Peak cardiorespiratory responses during aquatic and land treadmill exercise. Med Sci Sports Exerc. 2007;39(6):969–75.
36. Stevens SL, Morgan DW. Heart rate response during underwater treadmill training in adults with incomplete spinal cord injury. Top Spinal Cord Inj Rehabil. 2015;21(1):40–8.
37. Swain DP, Abernathy KS, Smith CS, Lee SJ, Bunn SA. Target heart rates for the development of cardiorespiratory fitness. Med Sci Sports Exerc. 1994;26(1):112–6.
38. Thomas KN, van Rij AM, Lucas SJ, Gray AR, Cotter JD. Substantive hemodynamic and thermal strain upon completing lower-limb hot-water immersion; comparisons with treadmill running. Temperature (Austin). 2016;3(2):286–97.
39. Tinken TM, Thijssen DH, Hopkins N, et al. Impact of shear rate modulation on vascular function in humans. Hypertension. 2009;54(2):278–85.
40. Willie CK, Colino FL, Bailey DM, et al. Utility of transcranial Doppler ultrasound for the integrative assessment of cerebrovascular function. J Neurosci Methods. 2011;196(2):221–37.


© 2017 American College of Sports Medicine