The brain is encapsulated by the skull, where the Monro-Kellie doctrine states the sum of brain parenchyma, cerebral spinal fluid, and cerebral blood flow (CBF) must remain constant to ensure intracranial pressure homeostasis (Monro 1783, Wilson 2016). However, following concussion, some individuals may experience an exacerbation of physiologically-based symptoms during physical exertion that is absent in healthy populations (Leddy et al 2007, O'Brien et al 2017, Graham et al 2021). This could potentially be attributed to deficits in autoregulatory and autonomic function, which could lead to a greater relative CBF increase during exercise, and further exacerbate these symptoms via the Monro-Kellie doctrine (Tan et al 2014). Therefore, developing novel techniques and exercise modalities that enable individuals with persisting post-concussion symptoms to obtain the positive benefits associated with exercise (e.g. elevated brain-derived neurotrophic factor, neural stem cell proliferation, decreased neuronal apoptosis) (Lucas et al 2015, Liu and Nusslock 2018, Gorski and De Bock 2019, Bliss et al 2020, Calverley et al 2020), while reducing symptom burden. This has been identified as a top research priority from the perspective of clinicians (Osmond et al 2023).

During exercise, a multitude of physiological changes occur; however, cerebral blood velocity (CBv; index of cerebral blood flow), generally follows a parabolic trend due to the change in chemical stimuli circulating within the cerebrovasculature (i.e. carbon dioxide) (Hellström et al 1996). CBv increases during mild to moderate-intensity, before peaking at ∼60%–70% of an individual's maximal oxygen uptake. This coincides with an intensity slightly before an individual crosses into anaerobic metabolism at a respiratory exchange ratio of ∼1.0 (Beaver et al 1986). Here, the build-up of carbon dioxide is the greatest, which results in a ∼20% increase in CBv (Hellström et al 1996).

This elevation in CBv, in conjunction with impaired autoregulation and autonomic function, may underlie the increased symptomology of headache, pressure in the head, dizziness, and so forth with engagement in physical exertion (Giza and Hovda 2001, 2014). This can nonetheless be blunted through external means (Yoshimoto et al 1994, Goswami et al 2018). For example, one way to artificially decrease CBv is through lower body negative pressure (LBNP), which redistributes blood flow into the venous system of the lower extremities (Crystal and Salem 2015, Goswami et al 2018). Pilot work from the author group using transcranial Doppler ultrasound demonstrated the ~20% increase in cerebral blood velocity during cycling exercise can be attenuated with the application of LBNP (Miutz et al 2023). This data within ∼30 individuals demonstrated CBv was elevated compared to baseline at 0 and −20 Torr, slightly elevated at −40, and not different at −60 and −70 Torr levels of LBNP (Miutz et al 2023). Previous studies have additionally used tilt table methodological approaches to manipulate CBF, which utilizes gravity to shift blood from upper to lower body compartments (Alperin et al 2005,van Campen et al 2018, van Campen et al 2020). While both LBNP and tilt are possibilities to reduce CBF/CBv, the combinatorial effect to attenuate the CBv increase during exercise has not been elucidated in the literature.

A dose-response relationship has been proposed with respect to the aforementioned benefits of acute exercise (Herold et al 2019). Individuals with PPCS may experience transient symptoms with physical exertion, and thus creating techniques where they could exercise for slightly longer and/or at a higher intensity, may help improve recovery trajectories in those slow to recover from concussion. This would also enable them to obtain a greater extent of the systemic effects of exercise on the organs and cerebrovasculature (e.g. increased insulin sensitivity, reduced arterial stiffness, elevated mitochondrial oxidative capacity, etc) (Lucas et al 2015, Calverley et al 2020). Moreover, increases in intracranial pressure are thought to occur following concussion due to tissue swelling (Haider et al 2018), which could additionally be blunted with external means. Therefore, this study aims to better understand the relationship between LBNP and tilt with supine cycling in a healthy population to determine the feasibility of employing this following concussion, which as stated is a top research priority identified by clinicians who treat concussion (Osmond et al 2023). The current protocol was first applied to a healthy population to lessen the burden of refining a methodological protocol in individuals with a concussion for safety and ethical reasons. A previous study by Tymko et al (2016), found CBv did not differ in steady-state conditions (without exercise) with 45 degrees head-up and head-down tilt independently; however, CBv was reduced when these were combined with −50 Torr LBNP. Therefore, it is hypothesized the combination of −40 Torr LBNP, and 45 degrees tilt would blunt the known ∼20% increase in CBv associated with submaximal exercise slightly below anaerobic threshold (Hellström et al 1996).

Ethical approval

The Conjoint Health Research Ethics Board at the University of Calgary (REB20-1662 and REB20-2112) provided the ethical approval for the current investigation. Prior to participation, participants provided written informed consent, researchers answered all questions, and thoroughly explained all protocols. Protocols were followed accordingly to institutional guidelines and standards set by the Declaration of Helsinki (aside from registering the study within a database).

Participants

Participants consisted of 23 healthy adults (11 females and 12 males) aged 20–33, who reported no diagnosis of concussion in the past 6 months. Participant demographics and environmental conditions are displayed in table 1. All participants were asked to complete the physical activity readiness questionnaire (PAR-Q+) (Warburton et al 2019), which requires participants to self-report any potential health complications/contraindications that may arise during exercise. No participants recorded any complications and thus were cleared to engage in maximal exercise. Before all testing sessions, participants were asked to refrain from exercise, smoking, alcohol, vaping, and caffeine for 8 h prior to maintain consistency between testing days and ensure results were accurate (Burma et al 2020b, Burma et al 2020c). No participants were regular consumers of either tobacco or vaping products (Husten 2009).

Table 1. Demographics of participants and environmental conditions across the three testing sessions.

VariableTotal (n = 23)Females (n = 12)Males (n = 11)Height (cm)170.7 ± 9.0165.2 ± 5.7175.8 ± 8.6Weight (kg)71.8 ± 11.963.5 ± 8.379.3 ± 9.5Body mass index (kg m−2)24.5 ± 2.723.2 ± 2.325.7 ± 2.5Age (years)24.0 ± 4.122.4 ± 2.625.6 ± 4.7Wattage of maximal MCAv117.0 ± 35.293.7 ± 28.1138.4 ± 26.8Barometric pressure (mmHg)   Day 1666.6 ± 4.3666.6 ± 4.1666.7 ± 4.9Day 2668.2 ± 6.7666.7 ± 4.9668.0 ± 6.7Day 3668.0 ± 2.6667.7 ± 2.4667.7 ± 2.4Room temperature (°C)   Day 121.2 ± 0.921.0 ± 1.021.4 ± 0.8Day 221.1 ± 0.821.0 ± 1.021.2 ± 0.6Day 321.1 ± 0.821.0 ± 0.921.4 ± 0.7Room humidity (%)   Day 142.8 ± 10.542.5 ± 11.242.0 ± 9.8Day 245.6 ± 7.344.1 ± 9.246.4 ± 4.6Day 346.8 ± 5.746.4 ± 5.546.7 ± 6.7

Data are displayed as mean ± standard deviation. Centimetres (cm), degrees Celsius (°C) kilograms (kg), middle cerebral artery velocity (MCAv), meters (m), millimeters of mercury (mmHg), percent (%).

Protocol

Participants completed three testing sessions within a week timespan, using a randomized crossover design (figure 1). The first visit consisted of a maximal ramp supine cycling test in order to identify the wattage that correlated with the peak CBv as previously described (Miutz et al 2023). This identified an individualized wattage associated with each participant's peak CBv to be used in the subsequent visits. Initial height and weight measurements were taken to determine each participant's stage increase in wattage during each minute of the exercise test until volitional fatigue. This wattage was calculated using the following formula (Miutz et al 2023), which has been demonstrated to elicit a robust parabolic MCAv response during supine cycling:

Figure 1. A visual representation of the three testing days with the exercise test occurring on the first, followed by the second and third in a randomized order with manipulation of lower body negative pressure (LBNP) and head-up tilt.

Download figure:

Standard image High-resolution image

As previously mentioned, peak MCAv is known to occur slightly before the anaerobic threshold. To minimize the potential impact of aerobic deconditioning and thus different energy systems contributing to the exercise condition (e.g. anaerobic versus aerobic metabolism), the second and third visits occurred within a 48–72 h time span. These consisted of two exercise sessions with progressively increasing head-up tilt that were completed in a randomized order. These were completed at the same time of day to minimize diurnal variation (Burma et al 2020a). During the second and third visits, 5 min of resting baseline data were collected. This was followed by 7 min exercise stages during successively increasing stages of tilt of 0, 15, 30, and 45 degrees during concurrent cycling (figures 1 and 2), for a total exercise duration of 28 min. The protocol for the two visits were identical, except one day was completed during −40 Torr of LBNP (experimental condition) while the other had no LBNP applied (control condition). During all stages of exercise, participants cycled at a cadence of 60–80 revolutions per minute (rpm).

Figure 2. A visual representation of the lower body negative pressure chamber containing the cycle ergometer inside. A participant is set up with the instrumentation utilized at the four stages of tilt used within the current investigation.

Download figure:

Standard image High-resolution image Instrumentation

The LBNP chamber was custom-built with a supine cycle ergometer situated inside. This also contained a bicycle seat with adjusting pedal lengths. At the opening of the chamber, the participant was fitted with a wooden plank around the iliac crest and then secured around the waist with a plastic sleeve and neoprene bands in order to ensure a tight pressure seal. During all exercise tests, bilateral middle cerebral artery velocities (MCAv), heart rate, blood pressure, and end-tidal carbon dioxide (PETCO2) were monitored and collected. The MCAv was measured using transcranial Doppler ultrasound (TCD; DWL USA, Inc., San Juan Capistrano, CA, USA), where the velocity and depth were recorded for each participant during the first test and matched for the subsequent visits based upon depth and velocity. Two 2 MHz probes were placed at the transtemporal window using a headframe fitted to each participant (DWL USA, Inc., San Juan Capistrano, CA, USA). Carotid compressions were used to ensure the artery insonated was fed from the internal carotid artery (Willie et al 2014). Heart rate was monitored using a Polar heart rate monitor fitted using a chest strap (Polar H10, Kempele, Finland). Additional measures of heart rate were collected using a 3-lead ECG (FE 231 BioAmp; AD Instruments, Colorado Springs, CO, USA). Finger photoplethysmography was used to collect blood pressure across the cardiac cycle and was corrected at heart level (Finometer NOVA; Finapres Medical Systems, Amsterdam, The Netherlands) (Omboni et al 1993, Sammons et al 2007). PETCO2 was collected through capnography with an inline gas analyzer (ML206; AD Instruments). All data were collected using LabChart and stored offline analysis with this commercially available software (LabChart Pro Version 8, AD Instruments).

Data processing

Bilateral MCAv's were taken as a precautionary measure in case one of the two probes experienced movement error during exercise. Therefore, for the final measure of CBv, the stronger of the two MCAv's was used to ensure the conclusions were based upon physiological differences between conditions, rather than measurement/movement and/or sonographer error. However, the within-participant comparisons were completed from the MCAv on the same side of the head. During the maximal exercise test, the last 40 s of each minute-long stage were averaged to determine MCAv. During the second and third sessions, the last 3 min of the 7 min stages were used to determine the average MCAv for each stage. Previous work has noted that supine-to-stand orthostatic fluid shifts generally stabilize by ∼60 s in the majority of individuals (Harms et al 2020). As a tilt change of 15 degrees is substantially less than a 90 degrees change from supine-to-standing, the initial 4 min buffer at the start of each stage ensured participants reached a steady-state following the fluid shift associated with each tilt stage. While absolute MCAv were recorded, the relative metric for each stage was calculated as the percent change compared to the baseline average, as seen below.

Sample size calculation

A sample size calculation was completed using G*Power (Version 3.1.9.6) based on the findings from Tymko and colleagues (2016), who compared MCAv responses during 5 min of −50 Torr LBNP within three positions: head-up 45 degree tilt, head-down 45 degree tilt, and supine. These authors noted a 15 cm s−1 reduction in MCAv when comparing between the supine baseline and after 5 min LBNP within the head-up tilt position, which is nearly identical to the stimulus used in the current investigation. This magnitude difference led to an f effect size of 0.91, which with an alpha and beta of 0.05 and 0.80, respectively, produced a required sample size of 10 participants. However, previous research has identified orthostatic intolerance differences between males and females (Cheng et al 2011). Therefore, the sample was doubled to enable a comprehensive comparison between LBNP and no LBNP, while also exploring the potential confounding influence sex has on the MCAv response during the experimental protocols.

Statistical analysis

All statistical analyses were performed using R-Studio (Version 2022.7.1.554) (R Core Team 2020). Differences between days and stages for absolute and relative MCAv were assessed using a two-factorial repeated measures analysis of variance (ANOVA): days (pressure and no pressure), stage (baseline, 0 degrees, 15 degrees, 30 degrees, and 45 degrees tilt), and the interaction between days and stages (Blanca et al 2017). Levene's tests were run to determine if any comparisons violated the homogeneity assumption between groups. Tukey's post-hoc pairwise comparisons were conducted to assess differences between groups in the case of a significant omnibus test. Sex comparisons at each stage were assessed using Wilcoxon rank sum tests. Based on literature suggesting making inferences based upon a binary p-value may not be the most appropriate for biomedical/physiological literature, inferences were made based on p-values and the associated effect size (Amrhein et al 2019, Panagiotakos 2008, Halsey 2019, Bakeman 2005). For the ANOVAs, generalized eta squared (η2G ) coefficients were used with thresholds of <0.02 (negligible), 0.02–0.14 (small), 0.14–0.26 (moderate), and >0.26 (large) (Bakeman 2005). Cohen's d coefficients were used for the Tukey post-hoc comparisons with thresholds of <0.20 (negligible), small (0.20–0.50), moderate (0.50–0.80), and large (>0.80) were used (Lakens 2013). Wilcoxon effect sizes (r) were used for the sex comparisons with thresholds of: <0.10 (negligible), 0.10–0.30 (small), 0.30–0.50 (moderate), and >0.50 (large) (Maher et al 2013). To understand if body weight confounded the outcome metrics of interest, basic linear regressions (R2) were completed for all stages of tilt across both exercise conditions. These were conducted on the relative MCAv metrics and stratified by biological sex, as lower body weight has been shown to result in a greater predisposition to orthostatic intolerance (Rutan et al 1992). Females and males differ in body weight and mass, and therefore these regressions were stratified by sex to control for both potential modifying factors of sex and body weight. While R2 estimates have no agreed-upon thresholds as these will vary depending upon the field of study, a priori cut-offs consistent with previous literature were used: <0.10 (negligible), 0.10–0.30 (small), 0.30–0.50 (moderate), 0.50–0.80 (large), and 0.80–1.00 (very large) (Burma et al 2022). Variability between baseline physiological variables was assessed through coefficient of variation (CoV) estimates (Hopkins 2000). Alpha was set a priori at 0.05.

Physiological parameters

All cardiorespiratory, cerebrovascular, and cardiovascular parameters are displayed in figure 3 across both days and all tilt stages. Baseline values between days are shown in table 2, with excellent reliability as displayed through the low coefficient of variation values (all <8.0%) and the MCAv being noted at <3.0%.

Figure 3. Average with 95% confidence intervals for all cardiovascular and respiratory parameters during each stage of tilt in 23 total participants (12 females and 11 males) during lower body negative pressure (LBNP) and no LBNP conditions. Beats per minute (bpm), breaths per minute (brpm), millimeters of mercury (mmHg).

Download figure:

Standard image High-resolution image

Table 2. Cerebrovascular, cardiovascular, and respiratory parameters across both days and each stage of tilt, stratified by sex (23 total: 12 females and 11 males). The coefficient of variation (CoV) was calculated to determine the between-day variability for baseline values. This ensured any differences were due to the lower body negative pressure (LBNP) stimulus and not natural between-day variations.

VariableSexDayBaseline0153045Baseline BD COVMCAv (cm s−1)FemaleLBNP78.0 ± 12.682.1 ± 14.977.6 ± 14.574.5 ± 14.969.5 ± 13.52.7 (95% CI: 1.7–3.7)  No LBNP76.8 ± 12.284.8 ± 14.682.1 ± 13.279.4 ± 12.777.4 ± 13.1  MaleLBNP63.9 ± 8.7169.7 ± 8.0266.2 ± 7.3863.6 ± 7.5660.8 ± 7.502.3 (95% CI: 1.2–3.4)  No LBNP63.5 ± 10.372.5 ± 9.4168.9 ± 8.1267.8 ± 8.9765.7 ± 9.06  TotalLBNP70.7 ± 12.875.6 ± 13.271.6 ± 12.568.8 ± 12.664.9 ± 11.42.5 (95% CI: 1.8–3.2)  No LBNP69.9 ± 12.978.4 ± 13.575.2 ± 12.673.4 ± 12.271.3 ± 12.4 Heart rate (bpm)FemaleLBNP70.3 ± 12.7116.9 ± 17.4129.7 ± 22.2138.6 ± 25.4147.9 ± 30.35.7 (95% CI: 0.6–10.8)  No LBNP67.6 ± 12.8108.0 ± 18.9111.5 ± 22.7115.4 ± 20.4119.2 ± 19.6  MaleLBNP66.2 ± 10.9112.1 ± 17.1115.2 ± 19.7120.9 ± 22.0129.0 ± 22.05.5 (95% CI: 1.7–9.3)  No LBNP65.5 ± 13.4112.7 ± 15.5115.5 ± 17.4117.2 ± 16.9116.4 ± 18.4  TotalLBNP68.1 ± 11.7114.4 ± 17.0122.1 ± 21.8129.4 ± 24.8138.0 ± 27.45.6 (95% CI: 2.7–8.5)  No LBNP66.5 ± 12.9110.5 ± 17.0113.6 ± 19.8116.3 ± 18.3117.7 ± 18.6 Mean arterial pressure (mmHg)FemaleLBNP84.5 ± 8.5787.0 ± 8.1983.2 ± 7.2385.0 ± 7.2686.1 ± 10.66.7 (95% CI: 4.2–9.2)  No LBNP81.7 ± 8.6293.8 ± 10.687.8 ± 9.3985.4 ± 10.385.3 ± 12.9  MaleLBNP82.9 ± 14.991.8 ± 20.696.2 ± 16.794.4 ± 11.997.6 ± 10.27.1 (95% CI: 3.4–10.8)  No LBNP80.6 ± 16.695.1 ± 20.791.5 ± 16.694.3 ± 17.894.7 ± 14.7  TotalLBNP83.7 ± 12.089.5 ± 15.889.9 ± 14.489.9 ± 10.992.1 ± 11.86.9 (95% CI: 4.9–9.0)  No LBNP81.1 ± 13.194.5 ± 16.389.7 ± 13.590.0 ± 15.190.2 ± 14.4 Systolic blood pressure (mmHg)FemaleLBNP122.8 ± 10.8144.5 ± 11.8134.2 ± 14.5134.0 ± 16.9131.3 ± 19.95.5 (95% CI: 3.0–7.9)  No LBNP122.6 ± 14.2148.7 ± 15.2141.4 ± 16.5139.2 ± 22.1139.2 ± 21.8  MaleLBNP124.4 ± 16.8155.9 ± 29.0155.9 ± 24.3153.7 ± 23.9148.5 ± 21.36.0 (95% CI: 3.1–8.9)  No LBNP126.6 ± 20.5171.8 ± 35.5167.1 ± 33.3163.5 ± 28.5161.3 ± 20.6  TotalLBNP123.6 ± 13.9150.4 ± 22.7145.5 ± 22.6144.3 ± 22.7140.3 ± 22.05.8 (95% CI: 4.0–7.5)  No LBNP124.7 ± 17.5160.7 ± 29.6154.9 ± 29.2151.9 ± 28.0150.7 ± 23.6 Diastolic blood pressure (mmHg)FemaleLBNP64.7 ± 6.7062.6 ± 6.7561.8 ± 6.8863.8 ± 5.9663.6 ± 8.386.0 (95% CI: 4.1–7.8)  No LBNP66.2 ± 7.4470.2 ± 8.2665.2 ± 8.2863.8 ± 8.3965.2 ± 9.50  MaleLBNP67.4 ± 10.471.4 ± 15.170.8 ± 9.8570.5 ± 7.8770.4 ± 8.227.6 (95% CI: 3.7–11.4)  No LBNP71.9 ± 12.478.9 ± 16.474.3 ± 14.274.9 ± 14.774.3 ± 13.3  TotalLBNP66.1 ± 8.7367.2 ± 12.566.5 ± 9.5667.3 ± 7.6567.1 ± 8.816.8 (95% CI: 4.7–8.9)  No LBNP69.2 ± 10.574.7 ± 13.669.9 ± 12.469.6 ± 13.170.0 ± 12.3 PETCO2 (mmHg)FemaleLBNP37.6 ± 4.5938.2 ± 2.4336.9 ± 2.4435.8 ± 2.6034.6 ± 3.515.6 (95% CI: 1.9–9.3)  No LBNP38.6 ± 4.5937.3 ± 4.4036.3 ± 4.3036.4 ± 4.2536.5 ± 3.95  MaleLBNP39.1 ± 3.9140.7 ± 3.9836.6 ± 4.6837.4 ± 4.5336.4 ± 4.657.4 (95% CI: 1.7–13.1)  No LBNP38.7 ± 5.5240.0 ± 4.7136.4 ± 4.8938.6 ± 4.9738.2 ± 5.09  TotalLBNP38.4 ± 4.2239.5 ± 3.5036.8 ± 3.7936.7 ± 3.7435.5 ± 4.156.6 (95% CI: 3.3–9.8)  No LBNP38.6 ± 4.9838.7 ± 4.6736.4 ± 4.6337.6 ± 4.6737.4 ± 4.56 Respiration rate (brpm)FemaleLBNP14.1 ± 5.3227.3 ± 4.6129.4 ± 5.4731.2 ± 4.7332.5 ± 4.868.7 (95% CI: 3.5–13.9)  No LBNP14.1 ± 5.2828.1 ± 8.8530.0 ± 8.2229.4 ± 8.1829.4 ± 7.53  MaleLBNP13.5 ± 3.2225.1 ± 4.8027.3 ± 5.4928.0 ± 5.2629.4 ± 5.758.9 (95% CI: 4.2–13.6)  No LBNP14.2 ± 2.5324.3 ± 5.3026.4 ± 5.0826.4 ± 4.6426.4 ± 4.83  TotalLBNP13.8 ± 4.2626.2 ± 4.7428.3 ± 5.4729.5 ± 5.1730.9 ± 5.458.8 (95% CI: 5.6–12.0)  No LBNP14.1 ± 3.9926.1 ± 7.3128.1 ± 6.8527.9 ± 6.5927.9 ± 6.30 

Data are displayed as mean ± standard deviation or mean (95% confidence interval [CI]). Middle cerebral artery velocity (MCAv), partial pressure end-tidal values of carbon dioxide (PETCO2).

Absolute CBv

Day (F(1,220) = 3.90, p = 0.049, η2G = 0.02 [small]) and stage (F(4,220) = 3.31, p = 0.012, η2G = 0.06 [small]) main effects were significant for absolute MCAv metrics, whereas the stage-by-day interaction was not (F(4,220) = 0.51, p = 0.729, η2G < 0.01 [negligible]) (figure 4).

Figure 4. Absolute and relative changes in middle cerebral artery (MCA) velocity across all stages in 23 total participants (12 females and 11 males) during lower body negative pressure (LBNP) and no LBNP conditions. Two-factorial repeated measures analysis of variance were used to determine differences between days and tilt stages. The Phi symbol (Φ) denotes a stage that differed compared to baseline. The Psi symbol (Ψ) denotes a stage that differed compared to 0 degrees tilt. The Upsilon symbol (ϒ) denotes a stage that differed compared to 15 degrees tilt. The Sigma symbol (Σ) denotes a stage that differed compared to 30 degrees tilt. Centimetres per second (cm/s), percent (%).

Download figure:

Standard image High-resolution image

The post-hoc comparisons revealed MCAv was lower on the negative pressure day compared to the no-pressure day (p = 0.049, d = 0.26 [small]) (figure 4). For stage, a greater MCAv was found during exercise at 0 degrees of tilt compared to tilt at 45 degrees (p = 0.008, d = 0.70 [large]) (figure 4). However, no other tilt stages differed from each other (all p > 0.082, all d ≤ 0.52 [negligible-to-moderate]) (figure 4).

Relative CBv

Main effects for all relative MCAv metrics were found to be significant (figure 4). Measured effects were day (F(1,220) = 56.1, p < 0.001, η2G = 0.20 [moderate]), stage (F(4,220) = 30.3, p < 0.001, η2G = 0.36 [large]), and stage-by-day (F(4,220) = 4.47, p = 0.002, η2G = 0.07 [small]) (figure 4).

The post-hoc comparisons