2.1.

Protocol and Cohort Summary

This study had the Florida Institute of Technology (FIT) Institutional Research Board approval.^20,21 FIT undergraduate and graduate student subjects were randomly selected from an open solicitation for study participants. No screening for general physical or cardiovascular fitness was used; however, all subjects selected were healthy and had no recent concussive injuries. Subjects were provided with a visual summary protocol presentation and a verbal orientation and introduction to all instrumentation used in the study and fully understood the protocol before beginning the study.

Subjects completed a basic health questionnaire, signed informed consent documents, and were weighed and their heights measured by the supervising medical professional. Subjects were asked to sit upright in an adjustable-height padded chair with armrests in a well-lit lab room, offered a blanket, and connected to the breathing circuit and instrumentation described in the following sections by the supervising medical professional.

Each study began with a 5- or 10-min baseline period during which the subjects inhaled room air through the breathing circuit. All datasets were collected during this 5- or 10-min baseline period, which was used as a reference/control. The baseline period was followed by one to three hypercapnic challenges of 5 or 10 min duration. A 5- or 10-min recovery period during which subjects breathed room air followed each hypercapnic challenge. A total of 83 challenges in 31 studies on 26 healthy, nonconcussed, FIT student subjects were delivered. Mean age ($\text{average}\pm \text{standard deviation}$) was $22\pm 2.9\text{year}$. The cohort was composed of 10 females with a height of $161.8\pm 5.84\mathrm{cm}$, weight of $66.59\pm 7.71\mathrm{kg}$, and BMI of $25.5\pm 3.14$ and 16 males with a height of $176.5\pm 7.70\mathrm{cm}$, weight of $74.25\pm 18.86\mathrm{kg}$, and BMI of $23.7\pm 5.22$.

2.2.

Breathing Circuit

A breathing circuit was designed using North 5500 series half-mask respirators (Honeywell Safety Products, Smithfield, Rhode Island) connected to two 18 in. lengths of continuous positive airway pressure (CPAP) tubing (HC140C, AG Industries, St. Louis, Missouri) by in-house three-dimensional printed connectors as its base (Fig. 1). One length of CPAP tubing was connected to a SUBGEAR SG 100 first- and second-stage regulators (SUBGEAR, Johnson Outdoors, Wendelstein, Germany). The other length of CPAP tubing was alternately plugged to close the breathing circuit and switch the inhalation source to the ${\mathrm{CO}}_2$-compressed air mixture and unplugged to open the circuit to room air. The ${\mathrm{CO}}_2$-compressed air mixture was housed in a Catalina S80 ($77.4{\mathrm{ft}}^3$ capacity, 3000 psi) SCUBA cylinder, which was connected to the first-stage regulator to complete the breathing circuit. The regulator demand valve was loosened to free-float conditions to minimize the negative pressure required to draw air from the tank into the breathing circuit.

Fig. 1

Subject fully connected to breathing and monitoring circuit. (a) Nonin Equanox Advance 7600, (b) data acquisition computer, (c) Bionet BM3, (d) Smiths Medical Capnocheck II 8400, (e) Arduino Uno, (f) Nonin Equanox Advance 8004CA under cotton headband, (g) Nonin Equanox Advance 7600PA, (h) modified North 5500 half-mask respirator, (i) Smiths Medical airway adaptor/sample line, (j) ECG leads and electrodes, (k) blood pressure cuff, (l) room air end of circuit: CPAP tubing and rubber plug, (m) ${\mathrm{CO}}_2$-compressed air mixed gas delivery end of circuit: SUBGEAR SG 100 and CPAP tubing, and (n) pulse oximeter (${\mathrm{SpO}}_2$).

2.3.

Vital Signs

Vital signs including heart rate and blood pressure were monitored using a Bionet BM3 (Bionet America, Inc., Tustin, California). The BM3 was connected to three electrocardiograph (ECG) leads connected to RedDot Ag/AgCl monitoring electrodes with clear tape and solid gel (9642, 3M, St. Paul, Minnesota) in a standard Einthoven’s triangle configuration, a Y-clip ${\mathrm{SpO}}_2$ sensor on the right index finger, and a blood pressure cuff on the left arm. Data for heart rate and ${\mathrm{SpO}}_2$ were acquired at a rate of 1 Hz using the companion B-link automated record keeping software (Bionet America, Inc., Tustin, California).

2.4.

Partial Pressure of End-Tidal ${\mathrm{CO}}_2$ and Respiration Rate

${\mathrm{PETCO}}_2$ (mmHg) and respiration rate (breaths per minute, bpm) were monitored with a Smiths Medical Capnocheck II 8400 Hand-Held Capnograph/Oximeter (Smiths Medical, St. Paul, Minnesota). The Capnocheck II provided four-breath averages for both ${\mathrm{PETCO}}_2$ and respiration rate, with an accuracy of $\pm 2\mathrm{mmHg}$ or 4% of the reading, whichever was greater, for ${\mathrm{PETCO}}_2$ and $\pm 1\mathrm{bpm}$ for respiration rate. Precision was 1 mmHg or 1 bpm.

2.5.

Regional Cerebral Oxygen Saturation

A Nonin Equanox Advance 7600 system and 8004CA NIRS (Nonin Medical, Inc., Plymouth, Minnesota) sensor containing differentially spaced light emitters and detectors embedded in a forehead patch were used to monitor ${\mathrm{rSO}}_2$ throughout the duration of each study. The patented layout of the 76-mm-wide 8004CA sensor contained two near-infrared (NIR) light-emitting diodes with wavelengths and maximum average powers of 730 nm at 3.0 mW, 760 nm at 4.5 mW, 810 nm at 3.2 mW, and 880 nm at 4.5 mW and spaced 60 mm apart, bookending two detection sensors spaced 20 mm apart.^7,22 Signals from each of the two emitters were received by both a detector spaced 20 mm from the emitter and a detector spaced 40 mm from the detector. This provided two 20 mm paths and two 40 mm paths in each 8004CA sensor.⁷ Two sensors were used, one on each side of the forehead. The 20-mm spacing measured extracranial tissue oxygen saturation while the 40-mm spacing measured both extracranial and intracranial tissue oxygen saturation.⁷ The device uses the 20-mm path to subtract the extracranial component of the signal from the 40-mm path in order to isolate intracranial signals.⁷ The purpose of using two extracranial (20 mm) and two overlapping extracranial + intracranial (40 mm) paths is to correct for differences in skin and skull optical properties between each emitter and its far (40 mm) detector.⁷

Data were obtained for the standard 1 update per 4 s rate via the Equanox direct link (low resolution dataset) and were also processed by Nonin Medical, Inc. for a high-resolution dataset of 18.75 updates per second. The high-resolution dataset is reported in this paper. The manufacturer-published accuracy of the device for ${\mathrm{rSO}}_2$ values in the range of 50% to 100% was $\pm 5.1$ $A_{\mathrm{rms}}$ (root mean square accuracy) absolute and $\pm 3.4$ $A_{\mathrm{rms}}$ trending during hypercapnia with a sensor repeatability (precision) of $\pm 2$ digits of $A_{\mathrm{rms}}$.²¹ The reference values used by the manufacturer for the root mean square accuracy were average oxygen saturation of arterial blood (${\mathrm{SavO}}_2$, %) calculated from the average of oxygen saturation of arterial blood samples (${\mathrm{SaO}}_2$, %) and oxygen saturation of jugular venous blood samples (${\mathrm{SjvO}}_2$, %) as ${\mathrm{SavO}}_2=(0.7\times {\mathrm{SjvO}}_2)+(0.3\times {\mathrm{SaO}}_2)$.^7,22

2.6.

Plethysmograph

A custom-made plethysmograph was used to measure cuff pressure changes in the right arm at a rate of 40 samples per second.

3.1.

Summary of Relationship Between $\mathrm{\Delta }{\mathrm{rSO}}_2$ and $\mathrm{\Delta }{\mathrm{PETCO}}_2$

Change in ${\mathrm{PETCO}}_2$ ($\mathrm{\Delta }{\mathrm{PETCO}}_2$) during ${\mathrm{CO}}_2$ inhalation was used as a correlate of the degree of hypercapnic challenge delivered. Change in ${\mathrm{rSO}}_2$ ($\mathrm{\Delta }{\mathrm{rSO}}_2$) during ${\mathrm{CO}}_2$ inhalation positively correlated to $\mathrm{\Delta }{\mathrm{PETCO}}_2$ ($R^2=0.40$). The mean $\text{response}\pm \text{standard error}$ in $\mathrm{\Delta }{\mathrm{PETCO}}_2$ was $6.39\pm 0.52\mathrm{mmHg}$, and the mean $\text{response}\pm \text{standard error}$ in $\mathrm{\Delta }{\mathrm{rSO}}_2$ was $2.22\pm 0.30\%$ (Fig. 2).

Fig. 2

Relationship between change in ${\mathrm{rSO}}_2$ versus change in ${\mathrm{PETCO}}_2$ during hypercapnia challenges as compared to prechallenge values.

3.2.

Influence of Self-Reported Exercise (h/week) on $\mathrm{\Delta }{\mathrm{rSO}}_2$ and $\mathrm{\Delta }{\mathrm{PETCO}}_2$

Exercise in hours per week (h/week) was self-reported on the intake questionnaire. Subjects were grouped into three exercise factor levels as follows: (1) 0 h/week, (2) $>0\mathrm{h}/\text{week}$ and $<10\mathrm{hr}/\text{week}$, and (3) $\ge 10\mathrm{h}/\text{week}$. Subjects grouped in exercise factor level 3 had a significantly greater mean $\mathrm{\Delta }{\mathrm{rSO}}_2$ response to ${\mathrm{CO}}_2$ inhalation than those in exercise factor level 2 [2.82% greater mean $\mathrm{\Delta }{\mathrm{rSO}}_2$ response, 95% confidence interval (CI): 1.00% to 4.65% greater mean $\mathrm{\Delta }{\mathrm{rSO}}_2$ response, $p=0.001$] and level 1 (3.39% greater mean $\mathrm{\Delta }{\mathrm{rSO}}_2$ response, 95% CI: 1.67% to 5.12% greater mean $\mathrm{\Delta }{\mathrm{rSO}}_2$ response, $p=0.00003$). No significant difference in subject mean $\mathrm{\Delta }{\mathrm{rSO}}_2$ responses was observed between exercise factor levels 2 and 1 (Fig. 3).

Fig. 3

Box plot of $\mathrm{\Delta }{\mathrm{rSO}}_2$ response during ${\mathrm{CO}}_2$ challenges for exercise factor levels. Subjects grouped in exercise factor level 3 had a significantly higher $\mathrm{\Delta }{\mathrm{rSO}}_2$ response than those in exercise factor levels 2 ($p=0.001$) and 1 ($p=0.00003$). Diamonds denote mean values, whiskers show 5th/95th percentiles, and small circles show outliers.

No significant differences in mean $\mathrm{\Delta }{\mathrm{PETCO}}_2$ were observed between the subjects in the three factor levels. This indicates that while subject $\mathrm{\Delta }{\mathrm{PETCO}}_2$ response to ${\mathrm{CO}}_2$ inhalation did not vary with exercise factor level, $\mathrm{\Delta }{\mathrm{rSO}}_2$ response did, and subjects exercising for greater than 10 h/week showed a different $\mathrm{\Delta }{\mathrm{rSO}}_2$ response for the same average $\mathrm{\Delta }{\mathrm{PETCO}}_2$ response to ${\mathrm{CO}}_2$ inhalation than subjects exercising less than 10 h/week. The subjects grouped in exercise level 3 demonstrated a greater CVR as measured by ${\mathrm{rSO}}_2$ than did the subjects grouped into exercise factor levels 2 and 1. These results identify a relationship between CVR as measured by ${\mathrm{rSO}}_2$ in this study and exercise frequency in healthy young adults.

Self-reported exercise frequency was considered to be a more significant indicator of physical fitness than BMI because all members of exercise factor group 3 were competitive athletes. One difficulty in using this method is that individuals that may participate in activities (such as walking) at a frequency and intensity to be considered nonsedentary may self-report as sedentary. While age, height, weight, and BMI were more tightly clustered for exercise factor group 3 than for groups 1 or 2, there were no significant differences in these variables between the groups (Table 1).

Table 1

Subject average age, height, weight, and BMI for each exercise factor group.

Were any of these variables used to group subjects for comparison or analyzed for covariance, the large variance of these variables, particularly in exercise factor groups 1 and 2, could have erroneously categorized relatively thin, sedentary subjects in the same group as muscular, endurance trained competitive athletes. Both exercise factor groups 1 and 2 had more members with BMIs below 20 and lower minimum BMIs than exercise factor group 3 members. Many studies have investigated the relationship between physical fitness and cerebrovascular health. Age studies typically group subjects as young adult,¹⁷ middle aged,²³ elderly sedentary/masters athlete,¹⁷ or simply young and older.²⁴ All subjects in this study were young adults. Exercise studies have grouped subjects based on self-reported/verified exercise frequency and/or type rather than BMI.^17,23

3.3.

Vital Sign Results

Continuous vital sign data were recorded for 75 of the 83 hypercapnic challenges. Data from each baseline, challenge, and recovery period were analyzed for respiration rate and heart rate against $\mathrm{\Delta }{\mathrm{rSO}}_2$ and $\mathrm{\Delta }{\mathrm{ETCO}}_2$, but no statistically significant relationships between these parameters could be established. Average heart rate during each postchallenge recovery period showed no significant difference from the prechallenge period in 61 of these 75 challenges. Large heart rate variability was observed. Relationships between respiration rate and other variables were complicated due to the fact that some subjects breathed spontaneously throughout the study, some paced their breathing during ${\mathrm{CO}}_2$ challenge periods and breathed spontaneously during prechallenge and recovery periods, and others paced their breathing throughout the duration of the study.

4.1.

Exercise and Cerebrovascular Reactivity

The response of ${\mathrm{PaCO}}_2$ to exercise intensity is known to be parabolic. ${\mathrm{PaCO}}_2$ typically increases from low intensity through moderate intensity exercise and then decreases during heavy exercise due to hyperventilation.^8,25 Rasmussen et al. proposed that CBF is dominantly controlled by ${\mathrm{PaCO}}_2$ and CVR during exercise based on the results of a 2006 study in which ${\mathrm{PETCO}}_2$, tidal volume ($V_{\mathrm{T}}$), respiration rate, and middle cerebral artery mean velocity (${\mathrm{MCAV}}_{\text{mean}}$) were monitored using an online gas analyzer and TCD before, during, 30 min after, and 24 h after exercise on a cycle ergometer.²⁵ The study identified a significant CVR increase above resting levels during exercise in normothermic conditions and a greater increase during exercise under hyperthermic condtions.²⁵ The subject cohort ($n=6$) was comprised of men ages 26 to 30 year, and ${\mathrm{PaCO}}_2$ was calculated from ${\mathrm{PETCO}}_2$ and $V_{\mathrm{T}}$.²⁵ A 2007 review article identified that ${\mathrm{MCAV}}_{\text{mean}}$ has been found to increase during exercise and subsequently decrease to resting levels upon exercise cessation, whereas MAP has been found to increase during exercise and then remain elevated above resting levels postexercise, leading to the conclusion that MAP may not be the dominant influence on increased CBF during exercise.⁸

Reports of the effects of age and exercise on CVR and CBF are conflicting. Thomas et al. performed a 3T MRI study and identified that masters athletes (MA) ($\text{mean age}=74\text{year}$) had lower CVR but higher resting CBF than sedentary elderly subjects,¹⁷ which may be due to the fact that ${\mathrm{PaCO}}_2$ increases during moderate aerobic exercise^8,15 and implies the possibility of a dampening of CVR due to lifetime exposure.¹⁷ Zhu et al. performed a TCD study and reported no difference in CVR between MA and sedentary elderly subjects, stating “life-long aerobic exercise training appears to have minimal effects on these age-related differences in hemodynamics.”²⁶ Sedentary elderly were found to have lower baseline CBF than MA in this study.²⁶ The discrepancy between these two studies^17,26 also highlights the possibility that MRI and TCD measures of CVR may not always agree. In another TCD study that measured CVR using MCAV, previously sedentary individuals completed a 12-week duration exercise training program.²⁴ No information is provided to qualify “previously sedentary” with regard to lifetime activity and ${\mathrm{CO}}_2$ exposure in the elderly population; however, the young were found to have a greater response to exercise than the old.²⁴ Many studies of the relationship of CVR to age and exercise do not consider subjects’ lifetime exposure to ${\mathrm{CO}}_2$.²⁴

Tarumi et al. used TCD and considered three measures of CVR: hypocapnic–hypercapnic, normocapnic–hypocapnic, and normocapnic–hypercapnic.²³ That study found that endurance trained middle-aged adults had significantly higher hypocapnic–hypercapnic CVR than sedentary middle-aged adults but identified no significant differences between the two groups in normocapnic–hypocapnic and normocapnic–hypercapnic CVR measures.²³ In the reported study, we identified that healthy young adult athletes that exercised more than 10 h/week had significantly higher normocapnic–hypercapnic CVR as measured with NIRS-based ${\mathrm{rSO}}_2$ than healthy young adults who exercised 0 h/week and $>0$ and $<10\mathrm{h}/\text{week}$, and those who exercised $>0\mathrm{h}/\text{week}$ and $<10\mathrm{h}/\text{week}$ had significantly higher normocapnic–hypercapnic CVR than those who exercised 0 h/week.

CVR may be dependent on numerous interacting factors, including age, exercise intensity, lifetime ${\mathrm{CO}}_2$ exposure, and other health conditions. Additionally, populations with the highest frequency of mTBI occurrences such as contact-sports athletes and military personnel frequently perform moderate to heavy exercise that may alter their CVR during their lifetime. Since CBF dysfunction in the form of altered CVR may be a marker of mTBI, it may be beneficial to establish baseline CVR for all individuals likely to experience repeated or prolonged ${\mathrm{CO}}_2$ exposure or to engage in activities that may result in head trauma or that have conditions known to affect CBF or CVR and regularly monitor them for CVR changes.

4.2.

Relationship of ${\mathrm{PETCO}}_2$ to ${\mathrm{PaCO}}_2$

Noninvasively measured ${\mathrm{PETCO}}_2$ is not interchangeable with invasively measured ${\mathrm{PaCO}}_2$,^3,8,9 and there are conflicting reports of the relationship between the two measurements. Peebles et al. identified that ${\mathrm{PETCO}}_2$ overestimates ${\mathrm{PaCO}}_2$ during normocapnia and hypercapnia²⁷ while other studies have found ${\mathrm{PETCO}}_2$ to underestimate ${\mathrm{PaCO}}_2$ as a function of the ratio of physiologic dead space in the lungs ($V_{\mathrm{D}}$) to $V_{\mathrm{T}}$²⁸ and by as much as 13 mmHg on average.²⁹ Regression relationships based on both ${\mathrm{PETCO}}_2$ and $V_{\mathrm{T}}$^3,25,30 and on ${\mathrm{PETCO}}_2$ only²⁷ have also been published. Since the subject study only employed noninvasive methods, and ${\mathrm{PaCO}}_2$, $V_{\mathrm{T}}$, and $V_{\mathrm{D}}$ were not measured, no conclusions can be drawn as to the response of ${\mathrm{PaCO}}_2$ to the ${\mathrm{CO}}_2$ inhalation challenges in this study.