Efficacy and time course of acute intermittent hypoxia effects in the upper extremities of people with cervical spinal cord injury

Abstract

Spinal cord injuries (SCI) disrupt neural pathways between the brain and spinal cord, causing impairment of motor function and loss of independent mobility. Spontaneous plasticity in spared neural pathways improves function but is often insufficient to restore normal function. One unique approach to augment plasticity in spinal synaptic pathways is acute intermittent hypoxia (AIH), meaning brief exposure to mild bouts of low oxygen, interspersed with normoxia. While the administration of AIH elicits rapid plasticity and enhances volitional somatic motor output in the lower-limbs of people with incomplete SCI, it is not known if AIH-induced neuroplasticity is equally prevalent in spinal motor pathways regulating upper-extremity motor-function. In addition, how long the motor effects are retained following AIH has not yet been established. The goal of this research was to investigate changes in hand strength and upper-limb function elicited by episodic hypoxia, and to establish how long these effects were sustained in persons with incomplete cervical SCI. We conducted a randomized, blinded, placebo-controlled and cross-over design study consisting of a single AIH or sham AIH session in 14 individuals with chronic, incomplete cervical SCI. In a subset of six participants, we also performed a second protocol to determine the cumulative effects of repetitive AIH (i.e., two consecutive days). In both protocols, hand dynamometry and clinical performance tests were performed pre- and post-exposure. We found that a single AIH session enhanced bilateral grip and pinch strength, and that this effect peaked ~3 hours post-intervention. The strength change was substantially higher after AIH versus sham normoxia. These findings demonstrate the potential of AIH to improve upper-extremity function in persons with chronic SCI, although follow-up studies are needed to investigate optimal dosage and duration of effect.

INTRODUCTION

Acute intermittent hypoxia (AIH) is an emerging technique for facilitating neural plasticity in individuals with neurological deficits. AIH consists of brief, rapid reductions of oxygen concentration in the inspired air, which results in stimulation of serotonergic neurons and serotonin release near motor neurons, new synthesis of brain-derived neurotrophic factor (BDNF), and activation of cellular cascades within motor neurons that potentiate motoneuron activity (Baker-Herman et al., 2004; Satriotomo et al., 2012). AIH holds considerable promise for inducing plasticity in persons with spinal cord injury (SCI), when implemented alone or paired with rehabilitation protocols. For example, a single AIH presentation of 15 hypoxic episodes (1 minute each) in persons with chronic incomplete SCI dramatically increases voluntary force generation at the ankle joint, beginning within minutes after administration of AIH (Lynch et al., 2017; Sandhu et al., 2019; Trumbower et al., 2012). AIH-induced plasticity in limb function is enhanced by combining daily AIH with task-specific training (e.g., walking practice), suggesting that AIH can be used as a primer for traditional rehabilitation, restoring function in spinal neural networks. In addition, recent studies show that daily AIH combined with task-specific training causes substantially larger effects than either intervention alone after SCI (Hayes et al., 2014; Navarrete-Opazo et al., 2017; Trumbower et al., 2017; Prosser-Loose et al., 2015).

The effectiveness of various AIH protocols for motor recovery have been studied in both pre-clinical (rodent model) and clinical studies. In rodent models, moderate intensity AIH protocols (3 episodes, 5 min at 11% O₂, 5-min intervals) elicit prolonged elevations in respiratory motor output known as long-term facilitation (LTF) (Golder and Mitchell, 2005; Hayashi et al., 1993; Lovett-Barr et al., 2012). Respiratory output in these preparations is typically recorded for up to 60–90 minutes post-hypoxia. In humans, the effects of AIH presentations have typically been studied for no more than 30–60 minutes after the last hypoxic exposure. In healthy individuals, 10 bouts (3 min) of isocapnic hypoxia (8% O₂) interspersed with 5 min of normoxia results in upper airway dilatator LTF, which persists for 40 minutes post-AIH (Rowley et al., 2007). In humans with chronic SCI, AIH (15, 1-min episodes of 9% O₂, 1-min intervals) increased voluntary ankle plantar flexion torque for up to 4 hours after a single AIH session (6/10 individuals; Trumbower et al., 2012). Recent evidence showed that a single AIH session potentiates corticospinal excitability in the hand muscles in uninjured humans (Christiansen et al., 2018), and in humans with incomplete cervical SCI (Christiansen et al., 2021). However, despite this growing body of evidence demonstrating that AIH improves motor function and corticospinal excitability, we do not yet know whether AIH augments upper extremity strength and function in persons with quadriplegia similar to the lower extremity (Trumbower et al., 2017). In addition, understanding how long the effects of AIH on motor function are sustained (the time course) remains unknown — a critical area of research needed to plan optimal combinatorial interventions when using AIH as a plasticity primer (Hayes et al., 2014; Navarrete-Opazo et al., 2017).

In light of these knowledge gaps, the purpose of this study was to: 1) test the hypothesis that episodic hypoxia increases hand strength and function; and 2) establish the time course of AIH effects on hand function in individuals with chronic SCI. Identifying the time course of AIH therapy could advance development of optimal interventions using AIH as a plasticity primer.

MATERIALS AND METHODS

Participants.

Fourteen individuals with SCI (mean age 46.9±11.9 years, 2 females; 12 males; Table 1) participated in the study. All participants gave informed consent to the experimental procedures, which were approved by the local ethics committee at Northwestern University and performed in accordance with the Declaration of Helsinki (IRB protocol #202027). Participants with SCI had a chronic injury (≥6 months) and were classified using the International Standards for Neurological Classification of Spinal Cord Injury examination as having a C4-C7 level SCI and by the American Spinal Injury Association Impairment Scale as C (n=7) or D (n=7). Recruitment was done through a research volunteer database at the Shirley Ryan AbilityLab (SRALab). Participants needed to have a preserved ability to voluntarily produce a visible grasping movement in at least one hand. Two individuals were not able to perform pinch grip on the left side (participant #16, Table 1) and on the right side (participant #17, Table 1). Only grip data was used from these individuals. No adverse clinical events were recorded in this study.

Table 1.

Demographic characteristics of study participants. AISA, American Spinal Injury Association Impairment Scale. W, white; AA, African American.

Study ID	Sex	Age	Race	ASIA Classification	Injury Level
01*	M	52	W	D	C5–7
02*	M	50	AA	C	C4–7
03	M	48	W	C	C7
04*	M	50	AA	D	C4–7
05*	M	61	W	C	C5–6
06	M	53	W	C	C6–7
07*	M	50	AA	D	C4–5
08	M	33	W	C	C7
09	M	64	Other	D	C5
10	F	40	W	C	C5
13	M	29	W	D	C5
16	F	23	W	D	C5–7
17	M	58	AA	C	C5
18*	M	46	W	D	C7

^*Denotes individuals who participated in the second protocol.

Exclusion criteria were a history of diagnosed sleep apnea, pregnancy, currently on anti-spasticity medications, and any other neurologic, orthopedic or cognitive conditions that could affect performance of the outcome measures. Following screening and informed consent, participants received instructions to maintain normal activity and not alter their medications. All experiments were performed at the SRALab in Chicago, IL, USA.

Study Design.

We conducted a randomized, blinded, placebo-controlled and cross-over study consisting of one session each of AIH and sham AIH (i.e., normoxia). Prior to the first visit, a member of the study staff who was not involved with data collection randomized subjects with a computer-based random number generator. The two interventions were separated by at least one week to minimize carryover effects. Outcome measures (see below) were recorded prior to, and for up to 5 hours post-AIH, and on the next day.

To determine the cumulative effects of successive AIH administration, we performed a second protocol in six individuals from the first study (marked by * in Table 1). In this protocol, participants visited the laboratory on three consecutive days. They received AIH on the first two days. Outcome measures were recorded prior to, and for up to 5 hours post-AIH on both days. These participants visited the laboratory on a third day for outcome measures. No sham AIH was given in this protocol. Study designs are illustrated in Figure 1.

Figure 1. Study Experimental Design.

Panel A: In the first protocol, outcome measures (marked by downward arrows) were collected prior to administration of AIH or sham AIH (baseline), every 30 minutes for up to 5 hours post-intervention (T₀ – T₅) and at 24 hours post-intervention (T₂₄). The AIH protocol consisted of 15, 60-second alternating episodes of inspiring hypoxic air (~9% O₂) with ambient room air (20.9% O₂). Sham AIH consisted of episodes of normoxia. Panel B: In the second protocol, AIH was administered on two consecutive days, and outcome measures were obtained prior to and after AIH on both days (marked by downward arrows), as well as on day 3.

AIH Administration.

Participants were fitted with a latex-free full non-rebreather mask using a custom neoprene head strap. A single AIH sequence was administered using a hypoxia generator (Model HYP-123, Hypoxico Inc., New York, NY, USA), constituting 60 seconds of 9% O₂ (Fraction of inspired air [FIO₂]: 0.09), alternating with 60 seconds of 21% O₂ (FIO₂: 0.21). The delivery of hypoxia and normoxic air mixtures was repeated 15 times per session, for a total of 30 minutes; oxygen saturations (SpO₂) were between 82–88% during hypoxic exposures. Sham AIH consisted of alternating exposures to normoxic air. Heart rate and oxygen saturation (SpO₂) were recorded continuously during the AIH administration, and blood pressure measurements were made prior to and after each administration.

Outcome Measures.

During each session, hand dynamometry and clinical performance tests were performed. The testing order was constant for all participants and conditions: grip strength and pinch strength, Nine Hole Peg Test (9-HPT), and Box and Block Test (BBT). For hand dynamometry, participants began with the pretest measurements of grip and pinch strength, which were followed by 30 minutes of AIH or sham AIH. A follow-up assessment was done immediately after intervention (T₀) and then every 30 minutes for the next 5 hours (i.e., T_0.5 – T₅). Individuals visited the laboratory for another measurement on the next day, the 24-hour time-point (T₂₄). Functional tests were performed prior to and at 0, 1, 4, and at 24 hours post-intervention (i.e. T₀, T₁, T₄ and T₂₄).

Hand Dynamometry.

Participants were seated with the shoulder adducted and neutrally rotated, elbow positioned at 90°; the forearm and wrist were in a neutral position. The forearm was in neutral rotation, and was not supported by arm rest or research personnel. Grip strength for each hand was measured using a hand dynamometer (Jamar Hydraulic Hand Dynamometer, Model: 5030J1). Measurement of maximal voluntary force was conducted by asking the subject to contract the hand against the handgrip. Each contraction lasted for approximately 3 to 5 seconds, and was repeated 3 times on each hand after a rest period of at least 1 minute. The average of 3 trials was used for each assessment. Maximal key pinch strength was tested using a digital dynamometer (MicroFet4, Hoggan Health Industries, West Jordan, UT). Participants held the dynamometer between the lateral aspect of the middle phalanx of the index finger and thumb pad. Similar to hand-grip assessment, they were instructed to squeeze as hard as they could for 3 s. Maximal pinch grip force was measured on each trial, and the average of 3 consecutive maximal pinch force measurements was calculated for each individual.

Functional Tests.

Clinical hand performance was tested using the 9-HPT and BBT. The 9-HPT test is a measure of finger dexterity and involves the placement of pegs from a container into holes on a board, and subsequent removal, as quickly as possible. The total number of seconds until task completion was recorded, measured by when the subject touched the first peg to when the last peg was placed back into the container. The BBT is a measure of manual dexterity that requires repeatedly moving 1-inch blocks from one side of a box to another in 60 seconds. For both tests, an average of two tests at each time-point was calculated for assessment.

Statistical Analyses.

Performance metrics were averaged across all observations in a given time-point for each participant. A liner-mixed model repeated-measures analysis of variance (RM ANOVA) was used to examine the effect of intervention (AIH, sham AIH) and time point (BL – T₂₄) on grip strength, pinch strength, and hand function. The Tukey post hoc test was used to identify significance in individual comparisons. For the second protocol, one-way RM ANOVA was used to examine the effect of time (BL – T₄₈) on grip and pinch strength. The repeated measures linear mixed model takes into account correlation across time for the same subjects and interventions. Values are reported as mean ± 1 standard error of the mean (SEM). A 2-sided P < 0.05 was considered statistically significant.

RESULTS

Grip strength.

Study participants had a grip strength of 49.3 ± 5.5 lbs. and 44 ± 4.8 lbs. on the left and right side, respectively. Average grip strength increased in both right and left hand after AIH administration compared with baseline. Peak effect was observed at 3.5 hours on left side (10% ± 4.1%, p = 0.038 vs. baseline), and 3 hours on the right side (10% ± 4.5%, p = 0.037 vs. baseline). Change in grip strength after AIH was statistically larger than sham AIH at all time-points from T₀ to T₂₄ on the left side (all p ≤ 0.046, vs. sham AIH; Fig. 3A), and at all time-points from T_0.5 to T₂₄ on the right side (all p ≤ 0.023 vs. sham AIH; Fig 3B). We also observed a moderate inverse correlation between baseline strength and the change in strength after AIH (R = 0.49, P = 0.03, Fig. 4A). Thus, individuals with low baseline strength showed a greater increase in strength following AIH.

Figure 3. Effect of AIH on grip and pinch strength.

Mean changes in maximal grip strength (Panel A and B) and pinch strength (Panel C and D) after a single session of AIH or sham AIH in N = 14 individuals with cervical SCI. Data points represent percent change from baseline (BL). AIH or sham AIH was given immediately after BL. Outcomes assessments were done at BL, immediately after intervention (T₀), every 30 minutes for the next 5 hours (i.e., T_0.5 – T₅), and finally the next day at a 24-hour time-point (T₂₄). Solid circles represent AIH and white circles represent sham AIH. *p < 0.046 vs. sham AIH

Figure 4. Relationship between baseline strength and AIH-induced plasticity.

Relationship between baseline strength (X-axis) and percent change in strength from baseline to peak strength after AIH (Y-axis). The left and right data were grouped together, and nonlinear regression analysis was done to show that individuals with a lower strength at baseline have a higher increase in strength following AIH.

Pinch strength.

Study participants had a pinch strength of 9 ± 1 lbs. and 8.8 ± 1.2 lbs. on the left and ride side, respectively. Similar to grip strength, the average pinch strength increased bilaterally post-AIH versus baseline however, it was not statistically different relative to baseline. Peak effect was observed at 2 hours (T₂) on the left side (6% ± 4.7%, p > 0.05), and at 30 mins (T_0.5) on the right side (12% ± 6%, p > 0.05). Change in pinch strength was however statistically larger than sham AIH at time-points T_0.5 – T₂, T₅ and at T₂₄ on the right side only (all p < 0.033 vs. sham AIH; Fig. 3D). We also observed a strong correlation between baseline strength and change in strength after AIH (R = 0.63, P = 0.0006, Fig. 4B). Thus, individuals with low baseline pinch strength showed a higher increase in strength following AIH.

Effect of AIH on hand function.

Mean box and block test (BBT) score changed from 44.4 ± 2.8 at baseline to 47.5 ± 3.1 at one hour (T₁; p = 0.029 vs baseline), and 51.3 ± 3.8 at 24 hours (T₂₄; p = 0.001 vs. baseline) post-AIH. Based on the findings of Chen et al. (Chen et al., 2009), a 5.5 point change in the BBT is necessary to be a clinical relevant change. Sham AIH had no significant effect on BBT (46.7 ± 3.6 at baseline to 48.3 ± 3.6 at one hour, and 49.3 ± 4 at 24 hours post sham; p > 0.05 at all time-points). No significant differences were observed in the 9-HPT score after AIH or sham AIH administration.

In the second protocol, AIH was administered on 2 consecutive days in 6 individuals, and outcome measures were obtained prior to and after AIH on both days, as well as on day 3. On the left side, baseline grip strength (pre-AIH) was 52.9 ± 8 lbs. on day 1, 54 ± 10 lbs. on day 2, and 61 ± 10 lbs. on day 3 (p = 0.003 on day 3 vs. day 1). On the right side, baseline grip strength (pre-AIH) was 50.4 ± 11 lbs. on day 1, 53 ± 12 lbs. on day 2, and 51 ± 11 lbs. on day 3. The baseline pinch strength was 11 ± 2 lbs. on day 1, 12 ± 2 lbs. on day 2, and 12 ± 2 lbs. on day 3 on the left side; and 12.6 ± 8 lbs. on day 1, 12.8 ± 3 lbs. on day 2, and 14 ± 3 lbs. on day 3 on the right side (p > 0.05 at both time-points vs. day 1). Individuals showed a relatively higher response to AIH on day 1 vs. day 2. Grip strength changed by 10% ± 7% on left and 6% ± 5 % on the right on day 1, and −5% ± 4% on the left and −3% ± 4% on the right on day 2 (p > 0.05 at both time-points vs. baseline). Pinch strength changed by 23% ± 17% on the left, and 13% ± 10% on the right on day 1, and −7% ± 6% on the left and −1% ± 8% on the right side on day 2 (p > 0.05 at both time-points vs. baseline). Figure 5 shows mean baseline grip strength on days 1 – 3, as well as individual values at baseline and at 2 hours after AIH (T₂) on days 1 and 2.

Figure 5. Effect of successive AIH on Grip and Pinch Strength.

Mean grip strength in the left and right hand (Panels A and B) and pinch strength in the left and right hand (Panels C and D) during the successive AIH protocol is presented. Horizontal mark (–) denotes mean strength at baseline (Pre), and at 2 hours after AIH (Post) on days 1 and 2. Not that no AIH was given on the third day. Change in strength for each individual (N=6) for day 1 and 2 are also depicted. *p < 0.003 vs. mean grip strength pre-AIH on day 1.

DISCUSSION

Impairment of arm and hand function is a significant problem after cervical SCI. AIH therapy is a promising technique to elicit spinal plasticity and drive functional recovery. Here we show that AIH improves hand strength in persons with chronic incomplete cervical SCI. Specifically, we show that a single AIH session enhances bilateral grip and pinch strength, an effect that peaks ~3 hours post-AIH. Paired AIH and upper limb/hand training is an attractive possibility to improve hand function in people with cervical SCI.

AIH triggers episodic spinal serotonin release, triggering new BDNF protein synthesis and activation of its high-affinity receptor, tropomyosin-related kinase B (Baker-Herman et al., 2004; Baker-Herman and Mitchell, 2002; Dale et al., 2017; Satriotomo et al., 2012). Subsequent downstream signaling cascades ultimately enhance the strength of synaptic inputs onto spinal motoneurons (Baker and Mitchell, 2000; Golder and Mitchell, 2005; Lovett-Barr et al., 2012; Vinit et al., 2009). AIH-initiated plasticity has been well demonstrated in both respiratory and non-respiratory neural pathways of rodents (Lovett-Barr et al., 2012). In persons with SCI, AIH has been demonstrated to cause broad-ranging changes, including improvements in respiratory function, plantar flexion torque, and locomotor function (Hayes et al., 2014; Lynch et al., 2017; Sandhu et al., 2019; Tester et al., 2014; Trumbower et al., 2012). Moreover, daily or cumulative AIH in chronic incomplete SCI is safe, and when combined with intensive rehabilitation training further enhances and prolongs walking recovery in individuals with SCI (Hayes et al., 2014; Navarrete-Opazo et al., 2017).

The main finding of the present study is that AIH improves hand strength and function in persons with cervical SCI. The corticospinal tract is a major descending motor pathway contributing to the control of voluntary movement in mammals (Lemon, 2008). Our results are consistent with previous evidence showing that a single AIH session potentiates corticospinal excitability in hand muscles in control subjects (Christiansen et al., 2018) and in people with incomplete cervical SCI (Christiansen et al., 2021). It is noteworthy that the magnitude of AIH-improved hand strength is lower than observed in lower extremities in previous studies. Trumbower et al. (2012) showed an increase in ankle plantarflexion torque by ~82% after a single AIH session; other studies have shown a lower magnitude of ankle torque change (~20% - 30%) within 30–60 mins post-AIH (Lynch et al., 2017; Sandhu et al., 2019). Our findings show that upper extremity effects are more limited (i.e. ~ 10% from baseline). The basis for differences in AIH effects on limb strength in humans with chronic SCI is unknown.

We observed a substantial and consistent difference between AIH and sham AIH, a difference that may be more substantial than first appears. Because of unexpected reductions in grip strength in sham experiments, possibly due to fatiguing properties of forearm muscles or disruption of wrist stabilizing forces, differences between AIH and sham AIH were greater than increases from baseline in the same subject. In previous studies, individuals with SCI have a higher fatigability during exercise due to several factors, such as a changes in motoneuron excitability, muscle contractile properties, or reductions in sensory inputs (Lin et al., 2012; Thomas et al., 1997; Thomas and Zijdewind, 2006). We speculate that grip and pinch force dropped over time due to fatigue in the sham AIH protocol, but that AIH-induced plasticity offset fatigue by increasing alpha-motor neuron excitability. A similar effect could also partly explain the lower magnitude of AIH effects in this study versus the lower extremities, which are less susceptible to peripheral fatigue (Vernillo et al., 2018). Gripping and pinching require multiple-joint muscle movements, and likely require more wrist stabilization versus ankle plantar flexion, which is a single-joint movement.

The BBT score was significantly higher on the day following AIH administration (a mean change of 5.9 points between baseline and 24 hours). To place this in perspective, a change of 5.5 points in the BBT score is considered a clinically relevant change (Chen et al., 2009). Thus, one might expect improvements in skilled movements or gross motor dexterity following AIH. In a recent pilot study, daily AIH combined with hand opening practice for 5 days improved BBT score in individuals with chronic SCI (5 subjects; Trumbower et al., 2017). Together, these results indicate that there may be AIH effects in multiple parts of the central nervous system beyond those linked to motor function and contraction strength alone.

The time course of response to successive AIH sequences showed unexpected results. Specifically, there was a modest but consistent response to the first AIH sequence, but a much less consistent effect after a second AIH sequence delivered 24 hours later. This was observed in both grip and pinch strength. While we are unclear as to the reason for this change, it may be linked in part to higher baseline strength on the second day, suggesting a ceiling effect. The baseline on the third day was even higher versus day 1 in most outcome measures, suggesting that the impact of AIH is accumulating with repetitive exposures. We also observed a higher baseline 24 hours in the first protocol after both AIH and sham AIH treatment. However, grip strength after AIH was significantly higher versus sham AIH in the first protocol. Thus, this higher baseline on day 2 could result from a combination of learning effects due to repeated administration of strength tests, and AIH-linked motor plasticity. Prolonged effects of a single AIH session warrant further investigation.

CONCLUSIONS

This study builds on several previous studies demonstrating that AIH improves motor function in individuals with chronic SCI. Here, we show that a single AIH session can improve upper limb function, as well as pinch and grasping strength, in persons with cervical SCI, although the magnitude of effects in the upper extremity appears more limited versus leg function. An important observation was that individuals develop increased voluntary grip strength and hand function one day after AIH treatment. More studies concerning the persistence and timing of AIH effects are needed to establish how long improvements are sustained.

Figure 2. AIH administration.

A representative example illustrating percent of inspired oxygen during AIH administration (top panel) and corresponding change in oxygen saturation (SaO₂, bottom panel).

Highlights.

• Acute intermittent hypoxia (AIH) enhances grip and pinch strength in persons with incomplete spinal cord injury

• The effect of AIH peaks around 3 hours post-intervention

• AIH significantly improves upper limb function as assessed by clinical measures (i.e., Box and Block Test).

Glossary

Abbreviations:

AIH

Acute Intermittent Hypoxia

ASIA

American Spinal Injury Association

BDNF

Brain-derived neurotrophic factor

LTF

Long-term facilitation

SCI

Spinal cord injury

SpO₂

Oxygen saturation