Journal of Neurotrauma

41:2114–2124 (September 2024)

ª Mary Ann Liebert, Inc.
DOI: 10.1089/neu.2023.0353

Open camera or QR reader and

scan code to access this article
and other resources online.

ORIGINAL ARTICLE

Cardiorespiratory Responses to Acute Intermittent

Hypoxia in Humans With Chronic Spinal Cord Injury
Joseph F. Welch1,2,4,** Alicia K. Vose,1–3,5,** Kate Cavka,3 Gina Brunetti,3 Louis A. DeMark,3 Hannah Snyder,3
Clayton N. Wauneka,3 Geneva Tonuzi,3 Jayakrishnan Nair,1,2,6 Gordon S. Mitchell,1,2,*** and Emily J. Fox1,3,*,***

Abstract
Brief exposure to repeated episodes of low inspired oxygen, or acute intermittent hypoxia (AIH), is a promising
therapeutic modality to improve motor function after chronic, incomplete spinal cord injury (SCI). Although
therapeutic AIH is under extensive investigation in persons with SCI, limited data are available concerning car-
diorespiratory responses during and after AIH exposure despite implications for AIH safety and tolerability.
Thus, we recorded immediate (during treatment) and enduring (up to 30 min post-treatment) cardiorespira-
tory responses to AIH in 19 participants with chronic SCI (>1 year post-injury; injury levels C1 to T6; American
Spinal Injury Association Impairment Scale A to D; mean age = 33.8 – 14.1 years; 18 males). Participants com-
pleted a single AIH (15, 60-sec episodes, inspired O2 & 10%; 90-sec intervals breathing room air) and Sham
(inspired O2 & 21%) treatment, in random order. During hypoxic episodes: (1) arterial oxyhemoglobin satu-
ration decreased to 82.1 – 2.9% ( p < 0.001); (2) minute ventilation increased 3.83 – 2.29 L/min ( p = 0.008); and
(3) heart rate increased 4.77 – 6.82 bpm ( p = 0.010). Considerable variability in cardiorespiratory responses was
found among subjects; some individuals exhibited large hypoxic ventilatory responses (‡0.20 L/min/%, n = 11),
whereas others responded minimally (<0.20 L/min/%, n = 8). Apneas occurred frequently during AIH and/or
Sham protocols in multiple participants. All participants completed AIH treatment without difficulty. No sig-
nificant changes in ventilation, heart rate, or arterial blood pressure were found 30 min post-AIH p > 0.05).
In conclusion, therapeutic AIH is well tolerated, elicits variable chemoreflex activation, and does not cause per-
sistent changes in cardiorespiratory control/function 30 min post-treatment in persons with chronic SCI.
Keywords: breathing; cardiovascular; hypoxia; plasticity; respiratory; ventilation

Introduction threatening the regulation of arterial O2 and CO2 partial

The respiratory and cardiovascular control systems are pressures and the transport of gases to and from the tis-
integrated to ensure the metabolic needs of tissues are sues. Respiratory insufficiency and autonomic dysfunc-
met under varying conditions and demands.1 After spinal tion are major concerns after SCI that can impact
cord injury (SCI), cardiorespiratory control is impaired, health-related quality of life and survival.2

1
Breathing Research and Therapeutics Center and Department of Physical Therapy, 2McKnight Brain Institute, University of Florida, Gainesville, Florida, USA.
3
Brooks Rehabilitation, Jacksonville, Florida, USA.
4
School of Sport, Exercise and Rehabilitation Sciences, College of Life and Environmental Sciences, University of Birmingham, Edgbaston, Birmingham, United Kingdom.
5
Department of Neurology, College of Medicine–Jacksonville, University of Florida, Jacksonville, Florida, USA.
6
Department of Physical Therapy, Thomas Jefferson University, Philadelphia, Pennsylvania, USA.
**Contributed equally to this work..
***Co-senior author.

*Address correspondence to: Emily J. Fox, DPT, MHS, PhD, University of Florida, HPNP Building, Room 1130, PO Box 100154, Gainesville, FL 32610, USA E-mail: [email protected]fl.edu

2114
CARDIORESPIRATORY RESPONSES TO AIH IN SCI 2115

After decades of extensive research in rodents, interest International Standards for Neurological Classification
has grown in the potential therapeutic application of of Spinal Cord Injury (ISNCSCI). Enrollment criteria in-
repeated short-duration exposures to low inspired O2, or cluded adults with >20% impairment in maximal inspira-
acute intermittent hypoxia (AIH).3 In 2012, Trumbower tory or expiratory pressure, and either: (a) incomplete
and associates4 reported an 82% increase in plantar flex- SCI (C1 to T12) based on residual sensory and motor
ion torque and 43% increase in gastrocnemius electro- function below the level of injury or injury classification
myographic activity 30 min post-AIH in 13 individuals of American Spinal Injury Association (ASIA) Impair-
with chronic, incomplete SCI. Positive outcomes have ment Scale (AIS) B, C, or D; or (b) chronic complete
since been observed in many somatic motor functions, SCI (C4 to T12; AIS A). Participants required clearance
including: maximal inspiratory mouth pressure5; resting from a physician to ensure medical stability.
ventilation6,7; walking speed and endurance8,9; hand- Exclusion criteria consisted of: diagnosis of an addi-
grip strength10; and dynamic balance.11 Early investiga- tional neurological condition such as multiple sclerosis,
tions have focused primarily on changes in motor output Parkinson disease, stroke or brain injury; presence of
with less emphasis on the effects of AIH on cardiorespi- unstable or uncontrolled medical conditions such as car-
ratory control/function.12 diovascular or pulmonary disease, hypertension, hetero-
The mild-to-moderate AIH dose stimulates carotid topic ossification; infections; severe neuropathic pain;
body chemoreceptors and activates pontomedullary car- known pregnancy; severe recurrent autonomic dysre-
diorespiratory control centers that reflexively increase flexia; and/or history of seizure disorder. To ensure eligi-
ventilation and heart rate (HR) in able-bodied individuals bility, all participants were screened before initiating
and those with SCI.4–6,13–15 Although some differences in the study. Screening included an in-person participant
experimental protocols exist between studies (e.g. dura- interview, physical examination, and review of medical
tion of hypoxic episodes, presence of concurrent hyper- records.
capnia), blood pressure appears well regulated without
measurable change from baseline.5,8,13,14 Few studies, Design
however, have examined both the immediate (within- After enrollment, participants attended the laboratory
treatment exposure) and enduring (post-treatment expo- on two separate occasions separated by a minimum of
sure) cardiorespiratory responses to AIH and most lack three weeks. The two testing sessions were conducted
sufficient detail (e.g. measurement of end-tidal gases, at the same time of day (in most cases) and the order of
beat-by-beat blood pressure) to make conclusions regard- visits was randomized. Testing on both visits involved
ing AIH safety and tolerability in persons with SCI. the same breathing apparatus and experimental proto-
Given the therapeutic potential of AIH and known del- col except for the inspired gas fraction during AIH
eterious effects of SCI on respiratory and autonomic (FIO2 & 0.10) and Sham (FIO2 & 0.21). Fifteen cycles
function,16,17 assessment of immediate and enduring car- of 60-sec hypoxia or normoxia exposures were inter-
diorespiratory responses to AIH is warranted. To that spersed with 90-sec intervals breathing ambient air.5
end, we sought to characterize cardiorespiratory re- Ventilation, HR, arterial blood pressure, and oxyhemo-
sponses to a single AIH treatment in persons with chronic globin saturation were measured before (baseline), dur-
SCI. We tested the hypotheses that AIH: (a) increases ing, and 20–30 min after AIH and Sham. Participants
HR and ventilation without significantly altering arterial and assessors administering inspired gases were blinded
blood pressure during treatment exposure; and (b) does to the intervention. All experiments were performed at
not elicit persistent short-term effects on any measure Brooks Rehabilitation in Jacksonville, FL.
of cardiorespiratory control/function.
Spirometry
Methods On the first visit, participants were seated comfortably
Participants and instrumented with a nose clip and a mouthpiece
Individuals with chronic (>1-year post-injury) SCI were attached to a filter and calibrated pneumotachograph
recruited through an Institutional Review Board appro- (MLT1000L; ADInstruments, Colorado Springs, CO).
ved research recruitment registry, the clinicaltrials.gov After a few normal resting breaths, participants were
website (NCT03833674), flyers, and local healthcare instructed to inspire fully, and then rapidly and forcefully
providers and researchers. The University of Florida expire until no more air could be exhaled. A minimum of
Institutional Review Board approved the study, and all three forced vital capacity maneuvers were performed,
participants provided informed consent to participate. and the test was terminated when the three largest values
Data reported here constitute part of a larger phase 2 varied by <150 mL according to testing guidelines.18
clinical trial. Forced vital capacity is reported as the largest of three
Injuries were classified by a licensed physical therapist consistent trials. Participants were encouraged through-
with expertise in SCI rehabilitation according to the out all tests to ensure maximal effort.
2116 WELCH ET AL.

Breathing circuit and apparatus tory and cardiovascular variables were stable, AIH or
An illustration of the breathing circuit during AIH and Sham commenced. The AIH and Sham protocols were
Sham is shown in Figure 1. Participants were reclined identical except for the FIO2 level during episodes. Fifteen,
in a power wheelchair or placed in a long sitting position 1-min episodes of hypoxia (AIH; FIO2 & 0.10) or nor-
on a mat with the trunk reclined 45 degrees. Participants moxia (Sham, FIO2 & 0.21) interspersed with 90-sec
were instrumented with a face mask (7450 V2; Hans intervals breathing ambient air were delivered. Cardiores-
Rudolph Inc., Shawnee, KS, USA) connected to a two- piratory responses were recorded at baseline, during AIH/-
way non-rebreathing valve (2700; Hans Rudolph Inc.). Sham episodes, and 20–30 min post-AIH/Sham (i.e.
A heated, calibrated pneumotachograph (3818; Hans recovery). Immediately following AIH and Sham, partici-
Rudolph Inc.) was connected to the expired limb of the pants were relieved of the face mask for comfort purposes.
circuit. The inspired limb of the circuit was attached to During that time, participants were permitted to use the
a three-way stopcock (2100; Hans Rudolph Inc.) and restroom and drink water only.
hypoxia generator (HYP-123; Hypoxico, New York,
NY) with a 5 L reservoir bag. Another two-way non- Cardiovascular variables
rebreathe valve was connected between the hypoxia The HR, blood pressure, and arterial oxyhemoglobin sat-
generator and face mask for positive pressure from the uration (SpO2) were recorded continuously at baseline,
generator to escape the circuit and prevent bias flow. Fil- during AIH/Sham and during recovery. Heart rate was
ters were placed on the inspired and expired limbs of the obtained using 3-lead ECG (ML-132; ADInstruments).
circuit. Beat-by-beat blood pressure was recorded using finger
Mouth pressure was continuously recorded via a side photoplethysmography (Finapres; Finapres Medical Sys-
port in the face mask connected to a calibrated pressure tems BV, Arnhem, Netherlands). Finger arterial pressure
transducer (113253; Hans Rudolph Inc.). Breath-by- was calibrated to brachial pressure using an automated
breath inspired/expired O2 and CO2 concentrations sphygmomanometer. The SpO2 was measured using a
were recorded at the mouth via a side port in the face finger pulse oximeter (ML320/F; ADInstruments).
mask connected to a calibrated O2/CO2 gas analyzer
(GEMINI; CWE Inc., Ardmore, PA). Data analysis
Expired tidal volume (VT) was calculated by integrating
Acute intermittent hypoxia and Sham _ Respiratory cycle time (TTOT) was
the expired flow (V).
Approximately 10 min of resting breathing data were col- calculated as the time interval between the onset of suc-
lected before AIH and Sham (i.e. baseline). Once ventila- cessive breaths. Breathing frequency (fb) was calculated

FIG. 1. Breathing circuit. See text for description. Pm, mouth pressure; PO2, partial pressure of O2; PCO2,
partial pressure of CO2, NRBV, non-rebreathe valve; PNT, pneumotachograph.
CARDIORESPIRATORY RESPONSES TO AIH IN SCI 2117

from TTOT. Minute ventilation (V_ E) was calculated as

fbVT. Partial pressures of respired O2 and CO2 (PO2
and PCO2) were measured continuously at the mouth,
and end-tidal partial pressures were calculated as
PET = FET(Pbar-47), where PET is the end-tidal gas pres-
sure, FET is the fractional end-tidal gas concentration,
Pbar is the ambient barometric pressure, and 47 is satu-
rated water vapor pressure at 37C. End-tidal PO2
(PETO2) and PCO2 (PETCO2) were calculated breath-
by-breath as the nadir in PO2 and peak in PCO2 wave-
forms. MAP was calculated as 1/3 SBP +2/3 DBP.
Cardiorespiratory variables were averaged over the
final 2–3 min of baseline conditions. During AIH and
Sham exposures, cardiorespiratory variables were aver-
aged over the final 20–30 sec of each hypoxic/normoxic
episode and the normoxic intervals. Cardiorespiratory
variables were averaged over the final 2–3 min of the
post-AIH/Sham recovery period. The hypoxic ventilatory
response during AIH was calculated as DV_ E/DSpO2 in
L/min/%, where the change in V_ E and SpO2 are reported
as differences between hypoxic episodes and normoxic
intervals. All data were sampled at 200 Hz using Power-
Lab data acquisition hardware (16/35; ADInstruments)
FIG. 2. Example of cardiorespiratory responses
and monitored online using LabChart software (V8.1;
to acute intermittent hypoxia (AIH). Example
ADInstruments). An example of cardiorespiratory res-
breath-by-breath and beat-by-beat traces from
ponses to AIH is shown in Figure 2.
one participant (SCI_15). Note the increases in
heart rate (HR) and minute ventilation (V_ E ) and
Statistics
Changes in cardiorespiratory variables were tested at decreases in O2 saturation (SpO2) and end-tidal
baseline and during AIH/Sham using a linear mixed O2 and CO2 partial pressures (PETO2, PETCO2)
model. Fixed effects were condition, time, and the inter- concomitant to hypoxic episodes. Participant
action between condition and time. Participant was a ran- informed experimenters of bladder discomfort
dom effect. Condition consisted of two levels (AIH at the onset of autonomic dysreflexia (AD),
and Sham), and time consisted of 30 levels (baseline, demarcated by an arrow. AP; arterial pressure.
episodes 1 to 15, and intervals 1 to 14). Mean changes AD was not caused by exposure to AIH.
from baseline in cardiorespiratory variables during AIH
versus Sham exposure and during recovery after AIH
versus Sham were compared by paired t tests. data were omitted from analyses: five individuals for
Associations between the magnitude of change in ven- end-tidal gases during AIH; two individuals for end-
tilatory parameters (VT, fb, and V_ E) from baseline versus tidal gases during Sham; five individuals for beat-by-
recovery after AIH and the hypoxic ventilatory response beat blood pressure during AIH; and 10 individuals for
during AIH were tested by the Pearson correlation. Pear- beat-by-beat blood pressure during Sham. Statistical
son correlation coefficient was classified as: negligible tests were performed using jamovi (V2.2.5; the jamovi
(r £ 0.29); low (r = 0.30-0.49); moderate (r = 0.50-0.69); project). Data are reported as means – one standard devi-
high (r = 0.70-0.89); and very high (r ‡ 0.90).19 As a sec- ation. Statistical significance was set at p £ 0.05.
ondary analysis, comparisons of the hypoxic ventilatory
response were made between individuals with chronic
SCI and data from healthy individuals collected in a Results
prior study (see Fig. 2 from13 for poikilocapnic AIH) Participant characteristics
using an independent samples t test. Twenty individuals met the inclusion criteria and were
All data were inspected for normal distribution of enrolled. One participant withdrew because of a health
residuals via Q-Q plots and histograms. Bonferroni cor- condition unrelated to the study. Thus, 19 participants
rections were applied when necessary to account for mul- (18 males; age = 33.8 – 14.1 years) were studied with var-
tiple comparisons. Because of equipment failure and/or ied injury characteristics: neurological level of injury C1
difficulties instrumenting participants, the following to T6; injury severity AIS A (n = 8), B (n = 4), C (n = 4),
2118 WELCH ET AL.

Table 1. Participant Information

Age (y) Time since FVC

Participant and sex NLI AIS injury (months) (%-predicted)

SCI_01 42 M C4 B 180 86.82

SCI_02 20 M C4 C 31 62.12
SCI_03 60 M C8 D 93 68.92
SCI_05 41 M C3 A 84 36.44
SCI_06 23 M T6 B 12 100.62
SCI_08 24 M T6 A 39 73.23
SCI_09 54 M C3 B 336 72.54
SCI_10 20 M C7 D 14 93.17
SCI_11 19 M C6 A 16 50.52
SCI_12 25 F C5 A 14 57.82
SCI_13 67 M C3 A 58 81.57
SCI_14 21 M C1 B 14 67.92
SCI_15 26 M C4 A 20 70.95
SCI_17 21 M T5 C 26 104.52
SCI_18 39 M T5 A 12 64.35
SCI_19 34 M T5 A 16 108.77
SCI_22 44 M C2 C 24 86.15
SCI_23 37 M C4 D 234 87.20
SCI_24 26 M C8 C 11.5 82.27
Mean 34 - - 65 76.63
SD 14 - - 87 18.11

NLI, neurological level of injury; AIS, American Spinal Injury Associ-

ation (ASIA) Impairment Scale; FVC, forced vital capacity; SCI, spinal
cord injury; SD, standard deviation.

and D (n = 3); and average time since injury of 5.4 –

7.5 years (range = 1–28 years). Forced vital capacity
was 3.8 – 1.1 L (range = 1.70–5.75 L, 36–109%-predicted). FIG. 3. Respiratory responses to acute
Full participant information and injury details are pro- intermittent hypoxia (AIH) and Sham. Figure
vided in Table 1. shows respiratory responses to AIH and Sham in
absolute values. B, baseline; fb, breathing
Respiratory responses to AIH and Sham frequency; PETCO2, end-tidal partial pressure of
Immediate (within-treatment exposure) respiratory CO2; PETO2, end-tidal partial pressure of O2; V_ E ,
responses to AIH and Sham are shown in Figure 3 and minute ventilation; VT, tidal volume. Statistics:
Table 2A. A main effect of time was found for VT, fb, Linear mixed model (condition · time). *Effect of
V_ E and PETO2. A main effect of condition was found for time ( p < 0.05); {effect of condition ( p < 0.05);
{
VT, V_ E, PETO2 and PETCO2. An interaction effect was interaction effect p < 0.05).
found for VT (F = 1.6, p = 0.026), V_ E (F = 3.0, p < 0.001),
and PETO2 (F = 63.1, p < 0.001). Full statistical compari-
sons and results are presented in Supplementary Table S1.

Table 2A. Cardiorespiratory Responses at Baseline and During Exposures to Acute Intermittent Hypoxia and Sham

AIH Sham

Variable Baseline Hypoxia %-D Baseline Normoxia %-D AIH vs. Sham p (%-D)

Respiratory
VT (L) 0.63 – 0.19 0.81 – 0.26 34.58 – 41.42 0.60 – 0.67 0.67 – 0.19 13.40 – 32.47 0.186
fb (breaths/min) 12.57 – 3.06 14.35 – 2.70 18.09 – 25.42 13.51 – 3.19 13.39 – 2.49 1.15 – 14.00 0.048
V_ E (L/min) 7.53 – 1.99 11.36 – 3.41 51.54 – 27.83 8.07 – 2.25 8.83 – 2.86 13.30 – 34.34 0.008
PETO2 (mm Hg) 101.98 – 7.76 55.12 – 7.21 -46.00 – 5.09 103.32 – 8.68 108.51 – 9.67 5.10 – 5.62 < 0.001
PETCO2 (mm Hg) 37.18 – 3.09 35.48 – 4.34 -4.63 – 7.89 35.22 – 4.73 33.32 – 5.17 -5.35 – 8.21 0.722
Cardiovascular
SBP (mm Hg) 119.15 – 12.26 123.60 – 14.01 3.75 – 6.06 118.84 – 17.70 121.53 – 12.15 3.09 – 8.81 0.930
DBP (mm Hg) 75.12 – 10.66 79.31 – 13.52 5.39 – 9.21 73.99 – 8.36 76.92 – 9.71 4.23 – 9.46 0.803
MAP (mm Hg) 89.80 – 10.97 94.08 – 13.38 4.69 – 7.60 88.94 – 10.97 91.79 – 10.36 3.60 – 7.91 0.891
HR (bpm) 68.55 – 8.17 73.32 – 10.54 7.02 – 10.12 68.11 – 12.72 67.76 – 12.46 -0.37 – 6.02 0.010
SpO2 (%) 95.18 – 1.94 82.11 – 2.90 -13.72 – 2.68 95.56 – 1.21 95.82 – 0.79 0.28 – 1.16 < 0.001
CARDIORESPIRATORY RESPONSES TO AIH IN SCI 2119

Mean changes in VT, fb, and V_ E during hypoxic epi- d = 0.13 vs. Sham), or DPETCO2 ( p = 0.920, d = 0.02 vs.
sodes from baseline were +0.18 – 0.21 L (35 – 41%; Sham) expressed as %-change from baseline. No associ-
p = 0.186, d = 0.33 vs. Sham), +1.78 – 2.72 breaths/min ations were found between DVT (r = 0.322, p = 0.179),
(18 – 25%; p = 0.048, d = 0.50 vs. Sham), and +3.83 – Dfb (r = -0.393, p = 0.096), or DV_ E (r = -0.147, p = 0.547),
2.29 L/min (52 – 28%; p = 0.008, d = 0.70 vs. Sham), and the hypoxic ventilatory response.
respectively. On average, PETO2 decreased to 55.1 –
7.2 mm Hg (-46 – 5%; p < 0.001, d = 6.73 vs. Sham), Cardiovascular responses to AIH and Sham
and PETCO2 decreased to 35.5 – 4.3 mm Hg (-5 – 8%; Immediate (within-treatment exposure) cardiovascular
p = 0.722, d = 0.10 vs. Sham) during hypoxic episodes responses to AIH and Sham are shown in Figure 6 and
from baseline. Considerable interindividual variability Table 2A. A main effect of time was found for SpO2,
in AIH responses was observed (Fig. 4). The hypoxic SBP, and MAP. A main effect of condition was found
ventilatory response during AIH was greater in persons for SpO2, HR, and DBP. An interaction effect was
with SCI versus healthy controls (0.25 L/min/% vs. found for SpO2 (F = 90.3, p < 0.001). Full statistical com-
0.12 L/min/%, p = 0.008), as shown in Figure 5. parisons and results are presented in Supplementary
Enduring (post-treatment exposure) respiratory res- Table 1.
ponses to AIH and Sham are provided in Table 2B. No Mean changes in SpO2, HR, SBP, DBP, and MAP
differences were found in DVT ( p = 0.684, d = 0.10 during hypoxic episodes from baseline were -13.06 –
vs. Sham), Dfb p = 0.111, d = 0.40 vs. Sham), DV_ E 2.58% ( p < 0.001, d = 5.92 vs. Sham), +4.77 – 6.82 bpm
( p = 0.218, d = 0.30 vs. Sham), DPETO2 ( p = 0.589, (7 – 10%; p = 0.010, d = 0.76 vs. Sham), +4.45 –

FIG. 4. Variability in respiratory responses to acute intermittent hypoxia (AIH). Figure demonstrates vast
differences in respiratory responses to AIH between participants. SCI_12 is an example of an individual with
normal chemoreflex modulation of breathing. SCI_02 is an example of an individual without evidence of
chemoreflex modulation of breathing. SCI_13 is an example of an individual with multiple episodes of
apnea followed by hyperventilation.
2120 WELCH ET AL.

FIG. 5. Comparison of hypoxic ventilatory response between healthy controls and people with chronic
spinal cord injury (SCI). Panel A shows the change in arterial O2 saturation (SpO2) and minute ventilation
(V_ E ) during hypoxic episodes relative to normoxic intervals in persons with chronic SCI (filled circles) and
healthy controls (open circles). Panel B shows the hypoxic ventilatory response (HVR) in chronic SCI and
healthy controls. Statistics: Independent samples t test. *p < 0.05.

6.72 mm Hg (3.8 – 6%; p = 0.930, d = 0.03 vs. Sham), Discussion

+4.19 – 6.28 mm Hg (5 – 9%; p = 0.803, d = 0.09 vs. Main findings
Sham), and +4.28 – 6.26 mm Hg (4 – 8%; p = 0.891, The purpose of the present study was to characterize car-
d = 0.05 vs. Sham), respectively. On average, SpO2 diorespiratory responses to AIH in persons with chronic
decreased to 82.1 – 2.9% and HR increased to 73.3 – SCI. Four major findings were: (1) AIH activates chemo-
10.5 bpm during hypoxic episodes from baseline. reflexes that increase V_ E and HR during hypoxic epi-
Enduring (post-treatment exposure) cardiovascular sodes, and these chemoreflexes are greater in persons
responses to AIH and Sham are provided in Table 2B. with SCI versus uninjured individuals; (2) respiratory
No differences were found in DSpO2 ( p = 0.622, and cardiovascular responses to AIH are highly variable
d = 0.12 vs. Sham), DHR ( p = 0.847, d = 0.05 vs. Sham), between individuals; (3) mild-to-moderate AIH does
DSBP ( p = 0.591, d = 0.20 vs. Sham), DDBP ( p = 0.683, not elicit persistent short-term (up to 30 min post-AIH)
d = 0.15 vs. Sham) or DMAP ( p = 0.638, d = 0.17 vs. changes in BP, HR, or ventilation; and (4) all participants
Sham) expressed as %-change from baseline. were able to complete the full AIH treatment.

Table 2B. Cardiorespiratory Responses at Baseline and 30 Minutes after Acute Intermittent Hypoxia and Sham (Recovery)

AIH Sham

Variable Baseline Recovery %-D Baseline Recovery %-D AIH vs. Sham p (%-D)

Respiratory
VT (L) 0.63 – 0.19 0.57 – 0.14 -7.42 – 16.53 0.60 – 0.14 0.56 – 0.12 -5.10 – 21.29 0.684
fb (breaths/min) 12.57 – 3.06 14.37 – 3.46 16.28 – 23.82 13.51 – 3.19 13.81 – 2.66 4.15 – 13.99 0.111
V_ E (L/min) 7.53 – 1.99 7.89 – 2.10 6.03 – 20.97 8.07 – 2.25 7.59 – 1.73 -2.44 – 20.59 0.218
PETO2 (mm Hg) 101.98 – 7.76 100.79 – 8.12 1.67 – 5.04 103.32 – 8.68 103.48 – 9.37 0.84 – 3.69 0.589
PETCO2 (mm Hg) 37.08 – 3.09 36.78 – 3.90 -0.77 – 7.11 35.22 – 4.73 34.75 – 4.67 -0.42 – 4.95 0.920
Cardiovascular
SBP (mm Hg) 119.15 – 12.26 121.89 – 14.97 2.46 – 9.31 118.84 – 17.70 120.59 – 14.23 2.62 – 14.52 0.591
DBP (mm Hg) 75.12 – 10.66 74.59 – 10.25 0.03 – 11.87 73.99 – 8.36 73.06 – 9.04 -0.50 – 14.15 0.683
MAP (mm Hg) 89.80 – 10.97 90.35 – 11.47 1.04 – 10.09 88.94 – 10.97 88.90 – 10.33 0.80 – 13.73 0.638
HR (bpm) 68.55 – 8.17 68.65 – 8.57 0.48 – 9.93 68.11 – 12.72 68.62 – 13.80 1.15 – 12.35 0.847
SpO2 (%) 95.18 – 1.94 94.86 – 2.14 -0.31 – 2.32 95.56 – 1.21 95.60 – 1.16 0.06 – 1.74 0.622

AIH, acute intermittent hypoxia; VT, tidal volume; fb, breathing frequency; VE, minute ventilation; PETO2, end-tidal partial pressure of O2; SBP, systolic
blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; HR, heart rate; SpO2, peripheral oxyhemoglobin saturation.
Statistics: Paired t test.
CARDIORESPIRATORY RESPONSES TO AIH IN SCI 2121

study that reported greater hypoxic ventilatory responses

in persons with SCI.6 The hypoxic ventilatory response
(during AIH), however, was not correlated with the mag-
nitude of change in ventilation post-AIH versus baseline.
Further study is necessary because clear trends in the
relationships between the hypoxic ventilatory response
and changes in VT and fb were apparent.
During hypoxia, PETO2 and PETCO2 (respectively)
decreased to 55 and 35 mm Hg. To our knowledge,
this is the first study to record PETO2 and PETCO2 dur-
ing poikilocapnic AIH in persons with SCI. Although
previous comparisons are not available, values found
here are within ranges reported for healthy young
adults.13,15
Unexpectedly, increases in V_ E during Sham exposures
(+0.76 L/min) caused hypocapnia (DPETCO2 from
baseline = -1.90 mm Hg). This modest hyperventilation
did not persist during recovery. We posit that an anticipa-
tory feedforward response may have occurred, similar to
the increase in HR and ventilation observed in humans at
the onset of physical exercise.25
Large variability in AIH responses was observed
(Fig. 4). Such variable responses are important for three
reasons. First, limited chemoreflex activation during
AIH may diminish potential therapeutic benefits because
FIG. 6. Cardiovascular responses to acute weaker chemoafferent stimulation predicts lesser activa-
intermittent hypoxia (AIH) and Sham. Figure tion of the serotonergic raphe neurons that drive respira-
shows cardiovascular responses to AIH and tory motor plasticity.3 When released in the spinal cord,
serotonin binds to 5-HT2 receptors on phrenic motor neu-
Sham in absolute values. B, baseline; DBP,
rons, initiating intracellular signaling cascades leading to
diastolic blood pressure; HR, heart rate; MAP,
new protein synthesis that increases phrenic motor output
mean arterial blood pressure; SBP, systolic blood
(i.e., phrenic motor plasticity).22–24
pressure; SpO2, arterial O2 saturation. Statistics:
Second, irregular breathing or apnea may alter the
Linear mixed model (condition · time). *Effect of
magnitude of fall in arterial and spinal tissue PO2 during
time p < 0.05); {effect of condition ( p < 0.05); hypoxic episodes. With longer apneas, arterial and spinal
{
interaction effect ( p < 0.05). tissue PO2 could fall to levels sufficient to trigger glial
adenosine triphosphate release and spinal adenosine
accumulation, undermining AIH-induced phrenic motor
plasticity.26–28
We interpret these findings to mean that AIH is safe Third, apnea combined with hypoxia may induce
and tolerable in persons with chronic SCI. Because of asphyxia (hypercapnia plus hypoxia). While the effects
large variability in AIH responses, capture of beat-by- of asphyxia on respiratory motor plasticity in persons
beat and breath-by-breath data may provide a useful with chronic SCI are unknown, acute intermittent
index to tailor AIH treatment to individual responses, hypercapnic-hypoxia is a more potent stimulus to ventila-
thus mitigating possible occurrences of confounding fac- tory long-term facilitation12,29 and enhanced cortico-
tors that disrupt the neurotherapeutic potential of AIH, diaphragmatic neurotransmission versus AIH alone in
such as recurrent apneas.3,20–24 healthy able-bodied individuals14,30; however, asphyxia
is also a hallmark feature of sleep apnea.
Immediate AIH effects on respiratory control Sleep disordered breathing has been implicated in
On average, hypoxia increased V_ E by 3.83 L/min ver- adverse responses to SCI rehabilitation,31 and chronic
sus baseline (Fig. 3), an effect driven by increases in intermittent hypoxia increases ventilatory loop gain,
both VT and fb. Increases in V_ E were significantly greater leading to increased frequency of apneic events during
than in healthy young adults from a previous study using sleep in persons with obstructive sleep apnea.32 Thus,
a similar poikilocapnic AIH protocol (+1.79 L/min).13 additional work is needed to determine whether our
Our observations are in accordance with an earlier mild-to-moderate AIH protocol elicits changes in apneic
2122 WELCH ET AL.

events during subsequent sleep. Alternatively, pairing hypercapnic-hypoxia, but not AIH alone, however,
AIH with hypercapnia may prevent apnea during treat- enhances corticodiaphragmatic conduction by a central
ment by increasing ventilatory drive and conscious neural mechanism of respiratory motor plasticity.13
awareness of breathing, while also enhancing respiratory Thirty minutes after AIH, HR had returned to base-
motor plasticity. line and BP remained unaltered. These findings are in
line with previous studies in humans with chronic SCI,
Immediate AIH effects on cardiovascular control which report minimal (or no) persistent AIH effects on
On average, SpO2 decreased to 82% during hypoxic epi- HR or BP regulation.6,8,10 Although prolonged effects
sodes, similar to that of previous studies of mild-to- were not observed, intermittent hypercapnic-hypoxia
moderate AIH in persons with SCI4,5 (Fig. 6). During (or asphyxia) experienced during sleep apnea elicits
hypoxic episodes, HR increased by 5 bpm on average, long-lasting pathology.50
reflecting modest chemoreflex activation. Our results Numerous studies report sustained increases in muscle
are in accord with previous work reporting increases in sympathetic nerve activity after various intermittent
HR during hypoxic episodes in persons with SCI.5,8 hypoxia protocols51,52; however, increased muscle sym-
While HR increased during hypoxic episodes, it is pathetic nerve activity is not always associated with
unknown whether stroke volume, cardiac output, or peri- changes in BP.53 Indeed, reductions in BP have been
pheral vascular resistance was altered. observed after AIH in healthy individuals54 and per-
No changes in SBP, DBP, or MAP were observed dur- sons with chronic SCI given the non-steroidal anti-
ing hypoxic episodes, although responses were highly inflammatory drug, ibuprofen.55
variable. Persons with SCI often exhibit diminished sym- A recent study in healthy humans found that intermit-
pathetic control of BP,33,34 leading to hypotension at rest tent asphyxia elicited long-lasting increases in muscle
and increased risk of syncope during exercise35,36 or sympathetic nerve activity, MAP, and the translation of
changes in posture.37 A recent study found impaired BP increased sympathetic outflow to changes in DBP; yet,
responses during the cold-pressor test in individuals forearm vascular conductance was unchanged, suggest-
with cervical SCI versus healthy able-bodied controls, ing regional differences in sympathetic neurovascular
supporting the notion that SCI alters the ability to regu- transduction after intermittent asphyxia.56 In rats with
late BP in response to environmental stress.38 cervical spinal hemisection, moderate AIH elicits renal
Nevertheless, similar to results reported here, BP does but not splanchnic sympathetic long-term facilitation.57
not appear to change during acute poikilocapnic hypox- Thus, it remains to be determined whether AIH can be
emia in healthy humans,39 perhaps indicating that our harnessed to improve autonomic function and orthostatic
therapeutic AIH dose does not present a major challenge tolerance post-SCI. Future studies are needed to test
to BP regulation. In support of this postulate, more severe whether AIH alters sympathetic outflow, cardiac contrac-
intermittent hypoxia protocols do report increases in arte- tility, endothelial function, sympathovagal balance, and
rial BP.40,41 In persons with obstructive sleep apnea and hemodynamics, in addition to establishing whether con-
hypertension, daily intermittent hypoxia increases SBP comitant hypercapnia enhances these responses, as is
during treatment, but lowers resting SBP between treat- known with breathing.
ments.42,43 A lack of data investigating mechanisms of
altered cardiovascular control during and after AIH limits Limitations
interpretation of our data. We studied a heterogeneous sample of 19 individuals
with chronic SCI, with injury levels ranging from C1–
Persistent AIH effects on cardiorespiratory control T6 and injury severities from AIS A–D. In addition, the
Table 2 provides an overview of group mean averages for age range of subjects was large (19–67 years) and
various indices of cardiorespiratory control at baseline included only one female participant. Although these
and 30 min post-AIH. No persistent short-term changes broad inclusion criteria generated a large dataset, they
in respiratory or cardiovascular responses were found, limit our ability to isolate specific subgroup effects.
suggesting that AIH per se has limited sustained impact While our predominantly male cohort aligns with the de-
on cardiorespiratory control/function post-SCI. mographic distribution of individuals living with chronic
Evidence has emerged demonstrating that poikilo- SCI,58 we acknowledge that the limited female represen-
capnic or isocapnic AIH does not elicit ventilatory tation prohibits generalizability of our findings. Future
long-term facilitation in healthy humans.44–46 Human studies are needed to determine whether AIH responses
ventilatory long-term facilitation has been observed differ on the basis of injury level/severity, age, sex,
most often when AIH is paired with hypercapnia, either and/or genetic biomarkers of plasticity.59
sustained47,48 or intermittent.49 Mechanisms differenti- Some data were omitted (described in Methods) because
ating poikilocapnic versus hypercapnic AIH have not of unreliable beat-by-beat and/or breath-by-breath record-
been comprehensively studied. Acute intermittent ings caused by equipment malfunction or difficulties
CARDIORESPIRATORY RESPONSES TO AIH IN SCI 2123

instrumenting participants because of (in part) injury char- ology, validation; LAD: investigation, supervision; HS:
acteristics and symptoms, such as hyperreflexia/spasticity. investigation, supervision; CNW: investigation, method-
Our data are limited to some degree by the narrow time ology, supervision; GJ: investigation, supervision; JN:
window studied post-AIH despite being consistent with investigation, methodology; GSM: conceptualization,
previous literature describing enhanced motor functions methodology, validation, resources, writing – review
in persons with SCI.4,5,7,55,60 and editing, project administration, supervision, projec-
ting administration, funding acquisition; EJF: concep-
Conclusions tualization, methodology, validation, resources, data
We investigated immediate and enduring cardiorespi- curation, writing – review and editing, project adminis-
ratory responses to AIH in persons with chronic SCI. tration, supervision, projecting administration, funding
Although AIH elicited robust chemoreflex activation, acquisition.
increasing V_ E and HR during hypoxic episodes, BP
remained unaltered. No changes in ventilation, HR, or BP Funding Information
were found 30-min after AIH relative to baseline. Thus, University of Florida (UF) McKnight Brain Institute,
we conclude that mild-to-moderate therapeutic AIH is Craig H. Neilsen Foundation (SCIRTS), United States
safe, well tolerated, and does not significantly affect cardio- Department of Defense (SCIRP W81XWH-17), National
respiratory control/function in persons with SCI. Institutes of Health (R01 HL147554), Brooks Rehabilita-
tion/UF College of Public Health and Health Professions
Research Collaboration.
Transparency, Rigor
and Reproducibility Summary Author Disclosure Statement
The current study is a substudy of a larger clinical trial No competing financial interests exist.
that was pre-registered with clinicaltrials.gov
(NCT03833674). As such, the pre-specified analysis
Supplementary Material
plan, statistical power, and sample size calculations Supplementary Table S1
were made based on primary outcomes from the parent
clinical trial. The current study is a secondary analysis References
with a planned sample size of 20 participants. Observed 1. Daly MdB. Interactions between respiration and circulation. In: Handbook
of Physiology, Section 3. (Cherniack NS, Widdicombe JG. eds.) American
effect sizes (Cohen’s d) for mean changes in V_ E and Physiological Society: Bethesda, MD; 1986
HR during AIH vs. Sham were 0.87 and 1.01, respec- 2. Cardenas DD, Hoffman JM, Kirshblum S, et al. Etiology and incidence of
tively. Participants were randomly assigned to groups rehospitalization after traumatic spinal cord injury: a multicenter anal-
ysis. Arch Phys Med Rehabil 2004;85(11):1757–1763
using a concealed, random number assignment. Investi- 3. Vose AK, Welch JF, Nair J, et al. Therapeutic acute intermittent hypoxia: a
gators and participants were blinded to group assignment translational roadmap for spinal cord injury and neuromuscular disease.
Exp Neurol 2022;347:113891
by use of an identically appearing treatment. Key inclu- 4. Trumbower RD, Jayaraman A, Mitchell GS, et al. Exposure to acute inter-
sion criteria and clinical outcomes were assessed by in- mittent hypoxia augments somatic motor function in humans with
vestigators with professional qualifications and multiple incomplete spinal cord injury. Neurorehabil Neural Repair 2012;26(2):
163–172
years of experience. All data were inspected for normal 5. Sutor TW, Cavka K, Vose AK, et al. Single-session effects of acute inter-
distribution of residuals via Q-Q plots and histograms. mittent hypoxia on breathing function after human spinal cord injury.
Exp Neurol 2021;342:113735
Bonferroni corrections were applied when necessary to 6. Sankari A, Bascom AT, Hiehani A, et al. Tetraplegia is associated with
account for multiple comparisons. Analytic code used enhanced peripheral chemoreflex sensitivity and ventilatory long-term
to conduct the analyses presented in this study may be facilitation. J Appl Physiol 2015;119(10):1183–1193
7. Tester NJ, Fuller DD, Fromm JS, et al. Long-term facilitation of ventilation
available by emailing the corresponding author. in humans with chronic spinal cord injury. Am J Respir Crit Care Med
2014;189(1):57–65
8. Hayes HB, Jayaraman A, Herrmann M, et al. Daily intermittent hypoxia
Acknowledgments enhances walking after chronic spinal cord injury: a randomized trial.
We thank our research participants for their time and Neurology 2014;82(2):104–113
dedication. 9. Navarrete-Opazo A, Alcayaga J, Sepulveda O, et al. Reptitive intermittent
hypoxia and locomotor training enhances walking function in incom-
plete spinal cord injury subjects: a randomized, triple-blind, placebo-
Authors’ Contributions controlled trial. J Neurotrauma 2017;34(9):1803–1812
10. Trumbower RD, Hayes HB, Mitchell GS, et al. Effects of acute intermittent
JFW: conceptualization, methodology, software, valida- hypoxia on hand use after spinal cord trauma: a preliminary study.
tion, investigation, formal analysis, data curation, writing Neurology 2017;89(18):1904–1907
11. Navarrete-Opazo A, Alcayaga J, Sepulveda O, et al. Intermittent hypoxia
– original draft, writing – review and editing, visualiza- and locomotor training enhances dynamic but not standing balance in
tion, supervision; AKV: conceptualization, methodology, patients with incomplete spinal cord injury. Arch Phys Med Rehabil
software, validation, investigation, formal analysis, 2017;98(3):415–424
12. Puri S, Panza GS, Mateika JH. A comprehensive review of respiratory,
writing – review and editing, supervision; KC: investiga- autonomic and cardiovascular responses to intermittent hypoxia in
tion, methodology, validation GB: investigation, method- humans. Exp Neurol 2021;341:113709
2124 WELCH ET AL.

13. Welch JF, Nair J, Argento PJ, et al. Acute intermittent hypercapnic-hypoxia 39. Tymko MM, Young D, Vergel D, et al. The effect of hypoxemia on muscle
elicits central neural respiratory motor plasticity in humans. J Physiol sympathetic nerve activity and cardiovascular function: a systematic
2022;600(10):2515–2533 review and meta-analysis. Am J Physiol Regul Integr Comp Physiol
14. Welch JF, Perim RR, Argento PJ, et al. Effect of acute intermittent hypoxia 2023;325(5)R474–R489
on cortico-diaphragmatic conduction in healthy humans. Exp Neurol 40. Foster GE, McKenzie DC, Milsom WK, et al. Effects of two protocols of
2021;339:113651 intermittent hypoxia on human ventilatory, cardiovascular and cerebral
15. Finn HT, Bogdanovski O, Hudson AL, et al. The effect of acute intermittent responses to hypoxia. J Physiol 2005;567(Pt 2):689–699
hypoxia on human limb motoneurone output. Exp Physiol 2022;107(6): 41. Foster GE, Hanly PJ, Ahmed SB, et al. Intermittent hypoxia increases
615–630 arterial blood pressure in humans through a Renin-Angiotensin system-
16. Williams AM, Gee CM, Voss C, et al. Cardiac consequences of spinal cord dependent mechanism. Hypertension 2010;56(3):369–377
injury: systematic review and meta-analysis. Heart 2019;105(3):217–225 42. Panza GS, Puri S, Lin H, et al. Daily exposure to mild intermittent
17. Berlowitz DJ, Wadsworth B, Ross J. Respiratory problems and manage- hypoxia reduces blood pressure in male patients with obstructive sleep
ment in people with spinal cord injury. Breathe 2016;12(4):328–340 apnea and hypertension. Am J Respir Crit Care Med 2022;205(8):949–
18. Graham BL, Steenbruggen I, Miller MR, et al. Standardization of spirom- 958
etry 2019 update. An official American Thoracic Society and European 43. Panza GS, Puri S, Lin H, et al. Divergent ventilatory and blood pressure
Respiratory Society technical statement. Am J Respir Crit Care Med responses are evident following repeated daily exposure to mild
2019;200(8):e70-e88 intermittent hypoxia in males with OSA and hyptension. Front Physiol
19. Hinkle DE, Wiersma W, Jurs SG. Applied Statistics for the Behavioral 2022;13:897978
Sciences. Houghton Mifflin: Boston, MA; 2003 44. Jordan AS, Catcheside PG, O’Donoghue FJ, et al. Long-term facilitation
20. Hayashi F, Coles SK, Bach KB, et al. Time-dependent phrenic nerve of ventilation is not present during wakefulness in healthy men or
responses to carotid afferent activation: intact vs. decerebrate rats. women. J Appl Physiol 2002;93(6):2129–2136
Am J Physiol 1993;265(4 pt 2):R811–R819 45. McEvoy RD, Popovich RM, Saunders NA, et al. Effects of sustained and
21. Morris KF, Arata A, Shannon R, et al. Long-term facilitation of phrenic repetitive isocapnic hypoxia on ventilation and genioglossal and dia-
nerve activity in cats: responses and short time scale correlations of phragmatic EMGs. J Appl Physiol 1996;81(2):866–875
medullary neurones. J Physiol 1996;490(pt. 2):463–480 46. Stewart D, Berthon-Jones M, McNamara SG, et al. Ventilatory responses to
22. Baker-Herman TL, Fuller DD, Bavis RW, et al. BDNF is necessary and suf- repetitive hypoxia in normal males. Am J Respir Crit Care Med 1994;
ficient for spinal respiratory plasticity following intermittent hypoxia. 149:A260
Nat Neurosci 2004;7(1):48–55 47. Harris DP, Balasubramaniam A, Badr MS, et al. Long-term facilitation of
23. Baker-Herman TL, Mitchell GS. Phrenic long-term facilitation requires ventilation and genioglossus muscle activity is evident in the presence
spinal serotonin receptor activation and protein synthesis. J Neurosci of elevated levels of carbon dioxide in awake humans. Am J Physiol
2002;22(14):6239–6246 Regul Integr Comp Physiol 2006;291(4):R1111–R1119
24. Fuller DD, Zabka AG, Baker TL, et al. Phrenic long-term facilitation requires 48. Wadhwa H, Gradinaru C, Gates GJ, et al. Impact of intermittent hypoxia on
5-HT receptor activation during but not following episodic hypoxia. long-term facilitation of minute ventilation and heart rate variability in
J Appl Physiol 2001;90(5):2001–2006 men and women: do sex differences exist? J Appl Physiol 2008;
25. Krogh A, Lindhard J. The regulation of respiration and circulation 104(1625-1633
during the initial stages of muscular work. J Physiol 1913;47(1-2):112–136 49. Vermeulen TD, Benbaruj J, Brown CV, et al. Peripheral chemoreflex con-
26. Devinney MJ, Nichols NL, Mitchell GS. Sustained hypoxia elicits compet- tribution to ventilatory long-term facilitation induced by acute inter-
ing spinal mechanisms of phrenic motor facilitation. J Neurosci 2016; mittent hypercapnic hypoxia in males and females. J Physiol 2020;
36(30):7877–7885 598(20):4713–4730
27. Mitchell GS, Baker TL. Respiratory neuroplasticity: Mechanisms and 50. Dempsey JA, Veasey SC, Morgan BJ, et al. Pathophysiology of sleep
translational impliations of phrenic motor plasticity. Handb Clin Neurol apnea. Physiol Rev 2010;90(1):47–112; doi:90/1/47 [pii] 10. 1152/
2022;188:409–432 physrev.00043.2008
28. Marciante AB, Mitchell GS. Increased spinal adenosine impairs phrenic 51. Cutler MJ, Swift NM, Keller DM, et al. Hypoxia-mediated prolonged
long-term facilitation in aging rats. J Appl Physiol 2023;134(6):1537–1548 elevation of sympathetic nerve activity after periods of intermittent
29. Mateika JH, Panza GS, Alex R, et al. The impact of intermittent or sustained hypoxic apnea. J Appl Physiol 2004;96(2):754–761
carbon dioxide on intermittent hypoxia initiated respiratory plasticity. 52. Morgan BJ, Crabtree DC, Palta M, et al. Combined hypoxia and hyperca-
What is the effect of these combined stimuli on apnea severity? Respir ponia evoked long-lasting sympathetic activation in humans. J Appl
Physiol Neurobiol 2018;256:58–66 Physiol 1995;79(1):205–213
30. Welch JF. Intermittent hypercapnic hypoxia: a model to study human 53. Xie T, Skatrud JB, Puleo DS, et al. Exposure to hypoxia produces long-
respiratory motor plasticity? J Physiol 2021;599(7):1951–1953 lasting sympathetic activation in humans. J Appl Physiol 2001;91(4):
31. Panza GS, Sutor TW, Gee CM, et al. Is sleep disordered breathing con- 1555–1562
founding rehabilitation outcomes in spinal cord injury research? Arch 54. Chacaroun S, Borowik A, Morrison SA, et al. Physiological responses to
Phys Med Rehabil 2022;103(5):1034–1045 two hypoxic conditioning strategies in healthy subjects. Front Physiol
32. Alex RM, Panza GS, Hakim H, et al. Exposure to mild intermittent hypoxia 2016;7:675
increases loop gain and the arousal threshold in participants with ob- 55. Lynch M, Duffell L, Sandhu M, et al. Effect of acute intermittent hypoxia
structive sleep apnoea. J Physiol 2019;597(14):3697–3711 on motor function in individuals with chronic spinal cord injury fol-
33. Teasell RW, Arnold JM, Krassioukov A, et al. Cardiovascular consequences lowing ibuprofen pretreatment: a pilot study. J Spinal Cord Med 2017;
of loss of supraspinal control of the sympathetic nervous system after 40(3):295–303
spinal cord injury. Arch Phys Med Rehabil 2000;81(4):506–516 56. Stuckless TJR, Vermeulen TD, Brown CV, et al. Acute intermittent hyper-
34. Fossey MPM, Balthazaar SJT, Squair JW, et al. Spinal cord injury impairs capnic hypoxia and sympathetic neurovascular transduction in men.
cardiac function due to impaired bulbospinal sympathetic control. Nat J Physiol 2020;598(3):473–487
Commun 2022;13(1):1382 57. Perim RR, Vinit S, Mitchell GS. Cervical spinal hemisection effects on spinal
35. Figoni SF. Exercise responses and quadriplegia. Med Sci Sports Exerc tissue oxygenation and long-term facilitation of phrenic, renal and
1993;25(4):433–441 splanchnic sympathetic nerve activity. Exp Neurol 2023;368:114478
36. Davis GM. Exercise capacity of individuals with paraplegia. Med Sci Sports 58. Center NSCIS. Traumatic Spinal Cord Injury Facts and Figures at a Glance.
Exerc 1993;25(4):423–432 University of Alabama at Birmingham: Birmingham, AL; 2023
37. Claydon VE, Steeves JD, Krassioukov A. Orthostatic hypotension following 59. Nair J, Welch JF, Marciante AB, et al. APOE4, age, and sex regulate
spinal cord injury: understanding clinical pathophysiology. Spinal Cord respiratory plasticity elicited by acute intermittent hypercapnic-
2006;44(6):341–351 hypoxia. Function 2023;4(5):zqad026
38. Sarafis ZK, Squair JW, Barak OF, et al. Common carotid artery responses 60. Sandhu MS, Gray E, Kocherginsky M, et al. Prednisolone pretreatment
to the cold-pressor test are impaired in individuals with cervical enhances intermittent-hypoxia induced plasticity in persons with
spinal cord injury. Am J Physiol Heart Circ Physiol 2022;323(6):H1311– chronic incomplete spinal cord injury. Neurorehabil Neural Repair
H1322 2019;(11):911–921