Each year over 795,000 Americans will suffer a stroke resulting in death or significant disability. While considerable functional gains are often made, significant assistance in daily life is still required in approximately one-third of stroke survivors [
Hyperbaric oxygen can be defined as the breathing of 100% oxygen at a pressure higher than atmospheric pressure. Initially, hyperbaric oxygen therapy (HBOT) was used to treat decompression sickness in divers; however, over the years its far-reaching potential was recognized, and it has been approved for a variety of purposes including wound repair, carbon monoxide poisoning, anemia, thermal burns, delayed radiation injuries, osteomyelitis, and actinomycosis (for review see [
Supporting a role for the use of HBOT in stroke patients is a wealth of experimental studies in a number of different animal models [
Despite the seemingly overwhelming potential of HBOT as defined by basic research and the underlying mechanistic rationale, clinical investigations have largely not produced the expected results. While research has provided some favorable evidence for HBOT in both acute strokes and poststroke [
Seven subjects (4 females) were enrolled in this study; 6 completed the study. Subjects eligible for this study were male or female and any age between 18 and 80 years who had suffered an ischemic stroke at least 12-month ago (to minimize the chance for spontaneous recovery) and exhibited some functional impairments. Of the participants 50% were 1 year after stroke when they enrolled in the study and the other 50% were 2 years after stroke (Table
Subject demographics.
| | | | |
---|---|---|---|---|
| Male | 79 | 2 years | Ischemic R MCA |
| Male | 61 | 1 year | Ischemic R MCA |
| Female | 57 | 1 year | Ischemic R MCA |
| Female | 57 | 2 years | Ischemic R MCA |
| Female | 59 | 2 years | Ischemic R MCA |
| Female | 31 | 1 year | Ischemic R MCA |
To carry out this study we used a within-subject design in which each subject provided his or her own baseline (i.e., pretreatment) comparison. A stable baseline was confirmed by the absence of functional recovery over the previous 3-months. Each subject was tested on a battery of outcome measures (cognitive, physical, speech, and quality of life measures) twice during a 3-month period to determine the baseline and 4 more times throughout the 9-month study to assess the effect of HBOT. If a steady baseline was established (based on the variables measured), the subject was eligible to begin the first round of HBOT treatment. HBOT consisted of 20 treatments of 100% O2 at 2.0 ATA for 60 minutes each day Monday through Friday for a total of 4 weeks. After this first treatment period ended and following four weeks without HBOT (labeled as Off below and in Figures
Effect of HBOT on memory: we observed a significant effect of HBOT on verbal and nonverbal memory using the CVLT (a & b), which measures verbal memory, and the WMS (c & d), which measures nonverbal memory. Graphs (a) and (c) represent the difference between baseline and treatment and (b) and (d) show all individual data points.
Effect of HBOT on Gait and UE mobility: we observed significant improvement in physical abilities as measured with the Upper Extremity Fugl Meyer (UEFM, (a & b) and gait velocity (c & d). Graphs (a) and (c) represent the difference between baseline and treatment and (b) and (d) show all individual data points.
Effect of HBOT on Quality of Life: participants reported significant improvement in sleep (a & b) and overall global recovery (c & d) following HBOT. Graphs (a) and (c) represent the difference between baseline and treatment and (b) and (d) show all individual data points.
The experimental endpoints for this study included speech measures, neuropsychological measures, physical measures, quality of life measures, and physiological biomarkers.
To assess changes in aphasia, communication ability, language, and verbal fluency the Boston Naming Test (BNT) and Reading Comprehension Battery for Aphasia (RCBA) with latencies were used. The Porch Index of Communication Ability (PICA) scoring was used for all tests, which provides a multidimensional scoring system for communication ability that describes accuracy, responsiveness, completeness, promptness, and efficiency of each response.
Cognitive and behavioral impairments (e.g., memory, attention/concentration, verbal fluency, and depression) were assessed using the Mini-Mental Status Exam (MMSE), California Verbal Learning test (CVLT-II), Grooved Pegboard test (GP), Trails A and B, Controlled Oral Word Association test (COWAT), Semantic Fluency (SF, animals), Wechsler Abbreviated Scale of Intelligence (WASI) block design, Wechsler Memory Scale (WMS) Visual reproduction, and Delis Kaplan Executive Function System (DKEFS).
Physical abilities including, gait, balance, and upper extremity function were assessed using the Upper Extremity Fugl Myer (UEFM), Berg Balance test, and GaitRite computerized system. Gait velocity, step length, and step time were measured with the GaitRite system.
Health status following a stroke was assessed for this study using the Stroke Impact Scale, which measures 8 different domains: strength, hand function, ADL/IADL, mobility, communication, emotion, memory/thinking, and participation [
Potential biomarkers for treatment and recovery were assessed using ELISA’s for astrogliosis (GFAP, Aplco Immunoassays, Salem, NH), astrocytic damage (S100
First, we established the stable baseline by comparing Baseline 1 and Baseline 2. Only the SIS global and sleep measures (part of Quality of Life) revealed reliable improvement across the two baseline assessments. Next, Cohen’s d effect sizes were calculated for each measure by combining the two baseline data points and the two treatment data points: Cohen’s d = (
To examine the effect of HBOT following an ischemic stroke we utilized a within design where each subject provided his or her own baseline. A total of 7 subjects were enrolled in this study. One patient withdrew due to ear pain associated with the HBOT, but this was the only adverse event noted during this study. One participant did not return for the 3-month follow-up visit. All subjects experienced an ischemic stroke in the right hemisphere of the brain (Table
To investigate the potential effect of HBOT on aphasia, communication ability, language, and verbal fluency the RCBA and BNT were used. We did not identify significant deficits at baseline in these speech and language domains; therefore, as we would not expect differences when comparing treatment to baseline no further analysis was completed.
An extensive neuropsychological battery was used to measure changes in cognitive impairments (e.g., memory, attention/concentration, and verbal fluency). Using the MMSE we did not observe impairments in general cognition for any participant (average score before treatment = 28.6 out of 30 possible points). Therefore, we did not complete further analysis on the MMSE. Overall, mild to moderate impairments were observed in the various cognitive domains measured. The baseline was stable for all the cognitive assessments. We observed a significant effect of treatment on verbal and nonverbal memory. Specifically, we observed a significant effect of HBOT on the CVLT, which measures verbal memory, and the WMS, which measures nonverbal memory (Figure
Effect of HBOT on functional impairments: means and effect size.
| | | | |
---|---|---|---|---|
| ||||
CVLT-II | 49.6 | 59 | 0.01 | 1.56 |
GP | 39.4 | 44.5 | 0.13 | 0.38 |
Trails A | 34.1 | 39.2 | 0.10 | 0.79 |
Trails B | 37.9 | 40.8 | 0.51 | 0.28 |
COWAT | 38.1 | 41.7 | 0.09 | 0.81 |
SF - animals | 42.9 | 44.1 | 0.76 | 0.13 |
WASI | 44.1 | 44.6 | 0.78 | 0.11 |
WMS | 10.4 | 12.2 | 0.03 | 1.11 |
DKEFS | 9.7 | 11.1 | 0.11 | 0.76 |
| ||||
| ||||
UEFM | 37.7 | 42.0 | 0.03 | 0.85 |
Berg | 48.7 | 50.7 | 0.26 | 0.5 |
Gait Velocity | 68.7 | 83.5 | 0.01 | 1.5 |
Step length | 3.79 | 2.36 | 0.33 | -0.39 |
Step time | 0.05 | 0.04 | 0.36 | -0.36 |
| ||||
| ||||
BDI | 14.5 | 10.0 | 0.09 | -0.82 |
SIS Global | 52.1 | 61.7 | 0.04 | 0.73 |
SIS Strength | 49.0 | 49.5 | 0.75 | 0.04 |
SIS Memory | 85.9 | 90.9 | 0.59 | 0.34 |
SIS Emotional | 74.4 | 82.6 | 0.14 | 0.54 |
Communication | 84.8 | 86.8 | 0.65 | 0.28 |
SIS ADL’s | 72.3 | 72.5 | 0.95 | 0.02 |
SIS mobility | 71.2 | 78.3 | 0.32 | 0.38 |
Hand Function | 58.9 | 60.6 | 0.57 | 0.23 |
Participation | 50.6 | 64.6 | 0.11 | 0.54 |
Physical Comp. | 56.0 | 61.1 | 0.09 | 0.81 |
Sleep | 41.2 | 48.5 | 0.04 | 1.17 |
Satisfaction | 46.2 | 53.8 | 0.16 | 0.68 |
| ||||
| ||||
NSE | 2.2 | 2.9 | 0.005 | 2.1 |
GFAP | 10.1 | 17.3 | 0.13 | 0.74 |
IL-6 | 6.3 | 3.9 | 0.05 | -1.0 |
TNF-a | 7.0 | 4.7 | 0.01 | -1.5 |
Effect of HBOT on functional impairments: repeated measures ANOVA.
| | | | |
---|---|---|---|---|
| ||||
CVLT-II | F(1,5)=0.13, p=0.74 | F(1,5)=15.7, p=0.01 | F(1,4)=0.13, p=0.74 | F(1,4)=2.78, p=0.19 |
Trails A | F(1,5)=0.01, p=0.77 | F(1,5)=3.9, p=0.10 | F(1,4)=1.27, p=0.34 | F(1,4)=2.95, p=0.18 |
COWAT | F(1,5)=2.59, p=0.17 | F(1,5)=4.0, p=0.09 | F(1,4)=0.31, p=0.61 | F(1,4)=0.003, p=0.96 |
WMS | F(1,5)=0.29, p=0.61 | F(1,5)=7.74, p=0.03 | F(1,4)=1.0, p=0.39 | F(1,4)=1.0, p=0.39 |
DKEFS | F(1,5)=0.11, p=0.78 | F(1,5)=3.6, p=0.11 | F(1,4)=0.07, p=0.81 | F(1,4)=0.11, p=0.76 |
| ||||
| ||||
UEFM | F(1,5)=0.65, p=0.46 | F(1,5)=9.6, p=0.03 | F(1,4)=6.5, p=0.06 | F(1,4)=0.24, p=0.64 |
Gait Velocity | F(1,4)=3.7, p=0.12 | F(1,4)=24.9, p=0.01 | F(1,3)=0.13, p=0.75 | F(1,3)=0.25, p=0.65 |
| ||||
| ||||
BDI | F(1,5)=0.16, p=0.71 | F(1,5)=6.7, p=0.09 | F(1,4)=1.4, p=0.32 | F(1,4)=0.66, p=0.48 |
SIS Global | F(1,5)=8.4, p=0.03 | F(1,5)=7.2, p=0.04 | F(1,4)=2.0, p=0.23 | F(1,4)=1.67, p=0.27 |
SIS Emotional | F(1,5)=0.56, p=0.28 | F(1,5)=3.0, p=0.14 | F(1,4)=0.83, p=0.41 | F(1,4)=0.51, p=0.51 |
SIS Participation | F(1,5)=0.005, p=0.9 | F(1,5)=3.7, p=0.11 | F(1,4)=0.56, p=0.50 | F(1,4)=0.08, p=0.80 |
SIS Physical | F(1,5)=2.8, p=0.17 | F(1,5)=4.2, p=0.09 | F(1,4)=0.96, p=0.38 | F(1,4)=1.0, p=0.36 |
Sleep | F(1,5)=6.5, p=0.06 | F(1,5)=8.2, p=0.04 | F(1,4)=1.5, p=0.28 | F(1,4)=0.001, p=0.97 |
Satisfaction | F(1,5)=0.06, p=0.82 | F(1,5)=2.7, p=0.16 | F(1,4)=6.8, p=0.06 | F(1,4)=0.11, p=0.75 |
| ||||
| ||||
NSE | F(1,5)=0.01, p=0.85 | F(1,5)=22.8, p=0.005 | F(1,4)=16.0, p=0.01 | F(1,4)=0.24, p=0.64 |
GFAP | F(1,5)=0.06, p=0.81 | F(1,5)=3.27, p=0.13 | F(1,4)=2.97, p=0.16 | F(1,4)=0.13, p=0.74 |
IL-6 | F(1,5)=0.02, p=0.89 | F(1,5)=6.0, p=0.05 | F(1,4)=5.3, p=0.08 | F(1,4)=1.1, p=0.36 |
TNF-a | F(1,5)=0.04, p=0.85 | F(1,5)=14.2, p=0.013 | F(1,4)=18.4, p=0.01 | F(1,4)=0.34, p=0.59 |
Gait, balance, and upper extremity movement were compared before and after HBOT treatment. We observed significant improvement in physical abilities as measured with the Upper Extremity Fugl Meyer (UEFM) and gait velocity. Gait velocity, as reported by the %-normalized to the general population increased nearly 20%, this effect was maintained at follow-up (Figure
Using the PROMIS QOL measure, participants reported significant improvement in sleep following HBOT and at the 3-month follow-up (Figure
To examine physiological biomarkers for treatment and recovery neural, glial, and inflammatory markers were measured. The strengths of these relationships are displayed with the effect size, which range from .69 to 1.5. A significant treatment effect was observed for 3 of the physiological biomarkers measured, NSE, TNF-alpha, and IL-6 (Tables
The purpose of this study was to investigate the role of HBOT as a therapeutic intervention for stroke patients. A stroke may result in a variety of functional deficits including physical, cognitive, and behavioral impairments. Using a within-subject design, we measured the impact of HBOT across a number of functional domains including speech, language, cognition, physical function, emotional / behavior impairments, and quality of life. In this preliminary study, our approach was to identify effects that were strong enough to emerge with a very small sample and, then, to examine the nature of those effects. For example, whether or how quickly they fade and whether they can be attributed to practice. This approach is likely to underestimate the potential for hyperbaric treatment due to the low statistical power for identifying effects strong enough for examination. However, the consistency of improvement noted over repeated assessments spread out over months argues against any account based on statistical fluke. Significant improvements following HBOT were observed with cognition (including, memory, and processing speed), gait velocity, upper extremity mobility, sleep, and overall recovery, as measured with the SIS. These treatment effects were maintained when examined at 3 months following treatment with the potential exception of the UEFM. We also observed a significant change in neural and inflammatory biomarker expression levels in response to HBOT. The pattern observed for the biomarkers was different than all the functional measures suggesting transient physiological responses but sustained functional change.
Although we observed significant improvements in cognition and gait velocity, there are limitations to interpretation. For example, while there was an increase in gait velocity from baseline to treatment other gait kinematics such as stride length and step length which compare symmetry of the left to right side did not show significant differences; the participants were just walking at a faster speed maintaining their biomechanical deficits. Furthermore, there were no significant differences on the Berg Balance test, which may indicate the increases in gait velocity may not be due to improvements of motor control of the paretic limb. In general the testers suggested that participants might have been trying to perform better at each testing interval, as they were not blinded to treatment. However, there is evidence suggesting a minimal practice effect with the CVLT and WMS [
The use of HBO as a treatment following stroke was first raised 40–50 years ago [
Adding to the studies investigating the effects of HBOT following stroke, with their mixed results, there has been a strong recent interest in the effectiveness of HBOT following a TBI, due to the increase in brain injuries sustained during recent military combat conflicts. The Department of Defense is implementing many facilities whose purpose is to use hyperbaric oxygen therapy to help veterans recover from TBI. A number of case studies have supported the use of HBOT following a TBI, suggesting beneficial effects even years after injury [
Other confounding variables including the type and sensitivity of outcome measures or domains assessed, and the timing of HBOT may all play a role in the incongruous and inconsistent findings in the literature. Due to the nature of ischemic injury one may conclude that HBOT would be most effective during acute injury when neurons can be rescued in addition to modulating plasticity and synaptic changes in new or existing neurons to compensate. Due to normal recovery during the initial 6–12 months following injury it is difficult to assign responsibility to one intervention. Despite the preponderance of evidence for an acute timeframe, Boussi et al. suggest neuroplasticity is possible in patients as far as 5 years after traumatic brain injury [
We appreciate that there are several limitations to the design of this study, some of which have been discussed above. While the within-subject design eliminates some of these issues previously discussed with RCT, the small sample size and lack of blinding and controls limit the generalizability of the results. Prior research has demonstrated the significant subject and observer/researcher bias inherent in this type of research, specifically with HBOT, and thus interpretation of the results its overall contribution to the scientific are narrow. Potentially the greatest limitation is only including subjects who have experienced a plateau in their recovery. This time frame places our study at risk of missing a critical therapeutic window to rescue cells before they are no longer viable, suggesting that the effects we hope to observe from HBOT treatment would be due to other mechanisms, i.e., not classical neuroprotection pathways. However, this time frame is necessary for this study design as the baseline must be stable to compare treatment effects. This is where other observational studies and even controlled clinical trials have fallen short and why the results from those studies are ultimately ineffectual. However, despite the limitations that can be found in most experimental design, the growing body of literature provides new and reliable data helping us to better understand the effects of HBOT on impairments resulting from ischemic strokes.
This study investigated the impact of HBOT as a therapeutic intervention following stroke across a number of functional domains including speech, language, cognition, physical function, and quality of life. We found a beneficial effect of HBOT on memory, processing speed, gait velocity, upper extremity mobility, sleep, and overall recovery. We also observed significant transient changes in neural and inflammatory biomarkers in response to HBOT that may result in the sustained functional changes that were observed. Despite these encouraging results further research is needed to more clearly define the mechanism and potential role of HBOT following stroke.
The data from this study is not an archived dataset. It was a small case series study.
This work was presented at the American Congress for Rehabilitative Medicine and thus the abstract has been previously published in the Archives of Physical Medicine and Rehabilitation.
The authors declare that they have no conflicts of interest.
This research was funded solely by the Casa Colina Foundation. The authors would like to thank Dr. Loverso, the Casa Colina Board of Directors, and the Casa Colina Foundation for supporting this research. We would also like to thank Kerry Gott, MD, Laura Seibert, PhD; Cindy Sendor, MA,CCC-SLP, Jose Fuentes, PhD, Felice Loverso, PhD, Lauren Meeks, BS., Elizabeth Cisneros, PhD, Adeel Popalzai, DO, Cathy Timple PT, DPT, and Laura Espinoza, MSW, for their technical support, input, and clinical guidance.