Abstract
Background:
We systematically reviewed randomized controlled trials (RCTs) of complementary and alternative interventions for fatigue after traumatic brain injury (TBI).
Methods:
We searched multiple online sources including ClinicalTrials.gov, the Cochrane Library database, MEDLINE, CINAHL, Embase, the Web of Science, AMED, PsychINFO, Toxline, ProQuest Digital Dissertations, PEDro, PsycBite, and the World Health Organization (WHO) trial registry, in addition to hand searching of grey literature. The methodological quality of each included study was assessed using the Jadad scale, and the quality of evidence was evaluated using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) system. A descriptive review was performed.
Results:
Ten RCTs of interventions for post-TBI fatigue (PTBIF) that included 10 types of complementary and alternative interventions were assessed in our study. There were four types of physical interventions including aquatic physical activity, fitness-center-based exercise, Tai Chi, and aerobic training. The three types of cognitive and behavioral interventions (CBIs) were cognitive behavioral therapy (CBT), mindfulness-based stress reduction (MBSR), and computerized working-memory training. The Flexyx Neurotherapy System (FNS) and cranial electrotherapy were the two types of biofeedback therapy, and finally, one type of light therapy was included. Although the four types of intervention included aquatic physical activity, MBSR, computerized working-memory training and blue-light therapy showed unequivocally effective results, the quality of evidence was low/very low according to the GRADE system.
Conclusions:
The present systematic review of existing RCTs suggests that aquatic physical activity, MBSR, computerized working-memory training, and blue-light therapy may be beneficial treatments for PTBIF. Due to the many flaws and limitations in these studies, further controlled trials using these interventions for PTBIF are necessary.
Keywords
Introduction
Fatigue is a common phenomenon following traumatic brain injury (TBI), with a reported prevalence ranging from 21% to 80% [Ouellet and Morin, 2006; Bushnik et al. 2007; Dijkers and Bushnik, 2008; Cantor et al. 2012; Ponsford et al. 2012], regardless of TBI severity [Ouellet and Morin, 2006; Ponsford et al. 2012]. Post-TBI fatigue (PTBIF) refers to fatigue that occurs secondary to TBI, which is generally viewed as a manifestation of ‘central fatigue’. Associated PTBIF symptoms include mental or physical exhaustion and inability to perform voluntary activities, and can be accompanied by cognitive dysfunction, sensory overstimulation, pain, and sleepiness [Cantor et al. 2013]. PTBIF appears to be persistent, affects most TBI patients daily, negatively impacts quality of life, and decreases life satisfaction [Olver et al. 1996; Cantor et al. 2008, 2012; Bay and De-Leon, 2010]. Given the ubiquitous presence of PTBIF, treatment or management of fatigue is important to improve the patient’s quality of life after TBI. However, the effectiveness of currently available treatments is limited.
Although pharmacological interventions such as piracetam, creatine, monoaminergic stabilizer OSU6162, and methylphenidate can alleviate fatigue, adverse effects limit their usage and further research is needed to clarify their effects [Hakkarainen and Hakamies, 1978; Sakellaris et al. 2008; Johansson et al. 2012b, 2014]. Therefore, many researchers have attempted to identify complementary and alternative interventions to relieve PTBIF [Bateman et al. 2001; Hodgson et al. 2005; Gemmell and Leathem, 2006; Hassett et al. 2009; Johansson et al. 2012a; Björkdahl et al. 2013; Sinclair et al. 2014]. In this study, we aimed to systematically review randomized controlled trials (RCTs) that evaluated treatment of PTBIF using complementary and alternative medicine (CAM) to provide practical recommendations for this syndrome.
Materials and methods
Search strategy and selection criteria
We performed online searches using multiple sources, including ClinicalTrials.gov, the Cochrane Library database, MEDLINE, CINAHL, Embase, the Web of Science, AMED, PsychINFO, Toxline, ProQuest Digital Dissertations, PEDro, PsycBite, and the World Health Organization (WHO) trial registry, in addition to hand searching of grey literature. The last search was performed on May 10, 2016. Search terms included TBI, head injury, head trauma, and brain injury; or brain trauma, concussion, fatigue, and tiredness. The reference lists of all retrieved studies and relevant reviews were searched manually to identify additional trials missed by the electronic literature search. G.X. and M.W. initially screened and included all articles based on the title and abstract. Full-text articles were obtained for all eligible studies and were assessed independently by G.X. and Y.L. against the inclusion and exclusion checklist. Disagreements were resolved by discussion until consensus was reached; if this failed, a third party, D.C., was consulted. All RCTs investigating the effect of complementary and alternative interventions on fatigue management in patients with TBI were included, regardless of intervention type. The inclusion criteria were: (1) studies that were RCTs or randomized crossover trials; (2) articles written in English and published or informally published; and (3) studies that compared interventions with a placebo intervention, no treatment, or other types of interventions. The exclusion criteria were: (1) nonrandomized studies; (2) studies that lacked outcome measures for fatigue; and (3) studies testing pharmacological interventions, as well as botanical or herbal interventions.
Outcome measures
The outcome of this study was any symptom of fatigue or tiredness, as evaluated by a range of current valid and reliable indices such as the Fatigue Severity Scale (FSS) [Krupp et al. 1989], Fatigue Impact Scale (FIS) [Fisk et al. 1994], Mental Fatigue Scale (MFS) [Johansson et al. 2009], and Profile of Mood States (POMS) [Shahid et al. 2012]. While there are many measures and instruments of fatigue with acceptable psychometric properties, there is not yet a ‘gold standard’ measure of fatigue due to clinical overlap between fatigue, depression, sleepiness, and other conditions.
Evaluation of quality of evidence
The methodological quality of each included study was assessed using the Jadad scale, based on the description of randomization, blinding, and patient attrition. Higher scores indicate better quality [Jadad et al. 1996]. The quality of evidence was evaluated using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) system [Balshem et al. 2011].
Data extraction
Data were extracted and independently crosschecked using a standard data extraction form. For each study, the types of interventions and detailed patient characteristics were extracted, as well as relevant fatigue outcome data.
Data analysis
For continuous outcomes, a weighted mean difference was calculated. In the presence of heterogeneity, we explored potential clinical, methodological and statistical sources. Because interstudy heterogeneity precluded a meta-analysis, a narrative synthesis of all of the included studies was employed.
Results
Search results
The study selection process is shown in Figure 1. A total of 1718 records were identified from searches. Ultimately, 10 unique RCTs that met the inclusion criteria were included in the review (Figure 1).
The study selection process for the systematic review.
Characteristics of the included randomized controlled trials
Details of the included trials are summarized in Table 1. The interventions and outcome measures varied between studies, indicating apparent clinical heterogeneity. There were 10 different types of interventions in the 10 included trials. In addition, because of the heterogeneity in population characteristics, intervention types, outcome measures and durations of intervention, it was not considered appropriate to perform a meta-analysis to provide a pooled estimate of the outcome measure. Therefore, narrative analysis rather than meta-analysis was employed. Among these interventions, there were four types of physical interventions: fitness-center-based exercise [Hassett et al. 2009], Tai Chi [Gemmell and Leathem, 2006], aquatic physical activity [Dijkers and Bushnik, 2008], and aerobic training [Bateman et al. 2001]. The three types of cognitive and behavioral interventions (CBIs) were cognitive behavioral therapy (CBT) [Hodgson et al. 2005], mindfulness-based stress reduction (MBSR) [Johansson et al. 2012a], and computerized working-memory training [Björkdahl et al. 2013]. The two types of biofeedback therapy were the Flexyx Neurotherapy System (FNS) [Schoenberger et al. 2001] and cranial electrotherapy [Smith et al. 1994], and blue-light therapy was the only included light therapy [Sinclair et al. 2014].
Characteristics of the studies and quality of the evidence included in the systematic review.
Abbreviations: ANOVA, analysis of variance; CHI, closed-head-injured; CBT, cognitive behavioural therapy; CG, control group; ES, effect sizes; FIS, Fatigue Impact Scale; FSS, Fatigue Severity Scale; FNS, Flexyx Neurotherapy System; IG, intervention group; Mn, mean; MFS, Mental Fatigue Scale; MBSR, mindfulness-based stress reduction; MFI, Multidimensional Fatigue Inventory; POMS, Profile of Mood States; RCT, randomized controlled trial; SD, standard deviation; SAH, subarachnoid hemorrhage; TBI, traumatic brain injury; VAMS, Visual Analog Mood Scale; WLC, wait-list control.
Quality evaluation
A summary of quality assessment scores for the included trials is shown in Table 2. Of the 10 RCTs, only three studies were high-quality RCTs (3–5 points) based on the Jadad scale [Smith et al. 1994; Bateman et al. 2001; Hodgson et al. 2005], and the other studies were low-quality RCTs (0–2 points) [Schoenberger et al. 2001; Gemmell and Leathem, 2006; Driver and Ede, 2009; Hassett et al. 2009; Johansson et al. 2012a; Björkdahl et al. 2013; Sinclair et al. 2014]. The quality of this evidence was judged to be low/very low using the GRADE system (Table 1). Risk of bias in study design, imprecision and indirectness was the most common reason for low grades, keeping in mind there was only one RCT trial for each intervention.
Jadad quality assessment scores for the included trials.
CBT, cognitive behavioural therapy; FNS, flexyx neurotherapy system; MBSR, mindfulness-based stress reduction;
0 = very poor; 5 = rigorous.
Reported effects of intervention
Physical interventions
Many different types of physical interventions are used in clinical settings. Physical exercise programs provide cognitive and functional benefits [Wischenka et al. 2016]. Driver and Ede conducted an 8-week aquatic physical activity intervention for patients after TBI and used a vocational rehabilitation class as a control. After assessing pre- and post-treatment symptoms using the fatigue subscale of the POMS, they found a significant improvement in the intervention group [IG, effect size (ES) = 1.00], while there was no significant change in the control group (CG, ES = 0.08) [Driver and Ede, 2009].
Fitness training is an intervention that can potentially reverse deconditioning. Hassett and colleagues compared cardiorespiratory fitness and psychosocial functioning effects between a supervised fitness-center-based exercise program and an unsupervised home-based exercise program in subjects with TBI in a multicenter study. There was no between-group difference in psychosocial functioning at the end of the intervention or at follow up, and no difference in fatigue was found between the two groups (p = 0.070 at 3 months; p = 0.178 at 6 months) [Hassett et al. 2009].
Tai Chi is a gentle stress-free exercise form that is characterized by soft flowing movements that can be practiced for health improvement. The effects of a 6-week Tai Chi course on individuals with TBI were investigated by Gemmell and Leathem. No significant difference was found in the intervention group in tiredness as assessed using the Visual Analog Mood Scale (VAMS) [Arruda et al. 1999], but data from the control group were not provided. The authors reported that no within-subject improvement in fatigue was found [Gemmell and Leathem, 2006].
Aerobic exercise has various health benefits, including improving cardiorespiratory fitness and psychological well-being. The impact of fitness training on inpatients with brain injuries was examined by Bateman and colleagues. They conducted an RCT with 157 total participants, including inpatients with TBI, stroke, subarachnoid hemorrhage, and other brain injuries. The patients attended a 12-week aerobic training program or relaxation training as the control condition. Assessments were performed using a fatigue scale before and after the 12-week training program and during a follow-up assessment 12 weeks post-training. Their results indicated that patient fatigue questionnaire scores did not significantly improve after the aerobic training program [Bateman et al. 2001].
Cognitive and behavioral interventions
CBIs are psychotherapy approaches that teach patients the cognitive and behavioral competencies needed to function adaptively in their interpersonal and intrapersonal worlds [Heimberg, 2002]. The common CBIs used in clinical settings include cognitive restructuring, relaxation training, social skills training, MBSR, and computerized working-memory training. Hodgson and colleagues evaluated the efficacy of CBT for acquired brain injury. They selected a set of patients, the majority of whom had suffered a TBI. The intervention group received hour-long CBT sessions each time, once per week for 9–14 weeks, and the control group was waitlisted. CBT may potentially reduce fatigue, as there was a medium ES (0.4) [Hodgson et al. 2005].
An MBSR is an educational program to improve attention and cognitive flexibility, increase brain neuronal connectivity, and help individuals to better cope with their difficulties. It has been used to treat patients with a wide range of conditions, such as stress, depression, pain, and fatigue [Johansson et al. 2012a]. The effects of MBSR on PTBIF with 29 participants with stroke and TBI were evaluated recently. Fifteen individuals participated in an MBSR program for 8 weeks, while the other 14 individuals served as controls and received no active treatment, but they were offered MBSR treatment during the following 8 weeks. Participants who completed the MBSR program had a decline in self-assessed MFS score (p = 0.004), while control group scores remained unchanged over the 8 weeks (p = 0.89). There was a significant difference in MFS score between the two groups after the 8-week MBSR program (p = 0.008). The control group completed the MBSR program at a later stage and also showed similar and significant declines in MFS scores after 8 weeks of MBSR (p = 0.002) [Johansson et al. 2012a].
Björkdahl and colleagues investigated whether computerized working-memory training after brain injury had a significant effect on daily life functions. Both the intervention group and control group underwent a 5-week standard rehabilitation. In addition, the intervention group also received working-memory training. The FIS score in the intervention group improved significantly after working-memory training compared with pretraining, but the score in the control group did not significantly improve [Björkdahl et al. 2013].
Biofeedback therapy
Conventional electroencephalographic biofeedback has the potential to improve the cognitive symptoms and problematic behaviors. Schoenberger and colleagues evaluated the potential efficacy of the FNS, a type of electroencephalographic biofeedback, as treatment for TBI. In between-group comparisons, the intervention group exhibited significantly improved General Fatigue and Mental Fatigue subscale scores versus the control group, though the total Multidimensional Fatigue Inventory (MFI) score did not change significantly [Schoenberger et al. 2001].
Cranial electrotherapy stimulation has been used to alleviate anxiety, depression and insomnia in clinical settings. Smith and colleagues assessed the effects of cranial electrotherapy stimulation (1.5 mA, 100 Hz, 45 min/day, 4 days/week for 3 weeks) on patients who suffered closed head injuries. Compared with both a sham treatment group and a no-treatment group, the treatment group showed significantly greater improvement on the Fatigue–Inertia scale of the POMS [Smith et al. 1994].
Light therapy
Light therapy can improve depressed mood and fatigue in patients with cancer. Sinclair and colleagues investigated the efficacy of 4 weeks (45 min/day) of blue-light therapy for fatigue reduction in patients with TBI and compared it with yellow light as a placebo and no treatment. Treatment with blue-light therapy significantly reduced fatigue by FSS score during the treatment phase. These changes were not observed in the groups that underwent yellow light therapy or no treatment. However, the authors indicated that improvements in these measures did not persist following cessation of the treatment at week 8 [Sinclair et al. 2014].
Side effects
Two of 10 included studies reported no side effects [Smith et al. 1994; Bateman et al. 2001], and three studies reported minor side effects, such as musculoskeletal pain for fitness training exercise and headache for blue-light therapy [Schoenberger et al. 2001; Hassett et al. 2009; Sinclair et al. 2014]. All adverse events resolved spontaneously and did not result in discontinuation of therapy. The other five studies lacked information regarding the side effects of CAM interventions [Hodgson et al. 2005; Gemmell and Leathem, 2006; Driver and Ede, 2009; Johansson et al. 2012a; Björkdahl et al. 2013].
Discussion
The objectives of this review were to systematically characterize and evaluate CAM studies for PTBIF. We have revealed the various CAM interventions explored to date for the treatment of PTBIF. There is evidence for the effectiveness of physical activity, MBSR, computerized working-memory training, and blue-light therapy for the treatment of PTBIF. However, these interventions must be used with caution in clinical practice because of the high risk of bias in most studies and the small number of studies for each intervention type. The quality of this evidence was judged to be low/very low using the GRADE system based on the risk of bias, imprecision and indirectness, and more important, there was only one RCT trial for each intervention.
CAM therapies are attractive because they use an integrative approach to healing and usually cause fewer side effects than drug treatment. In the 10 studies included in this review, two studies reported no side effects and three studies reported minor side effects, whereas five studies lacked information concerning side effects. In another systematic review there were only five studies among 26 included RCTs on CAM treatment for chronic fatigue syndrome that assessed the side effects of CAM treatment. The conclusion from those five studies also indicated that no severe side effects were found for CAM treatment. However, the side effects of CAM treatment should be assessed in future CAM trials, even though they are minimal [Alraek et al. 2011].
Despite the high occurrence and enduring nature of fatigue complaints after TBI and the trials using complementary and alternative interventions that show promising preliminary findings, the most effective strategies for PTBIF treatment are not yet established. A systematic review of the literature on fatigue management currently being undertaken suggests there are few high-quality studies on effective PTBIF interventions [Hicks et al. 2007]. While Cantor and colleagues previously conducted a systematic review of interventions for fatigue after TBI, they included both RCTs and non-RCTs [Cantor et al. 2014]. In the current study, we included only RCTs, with three new RCTs in addition to the seven RCTs in Cantor’s systematic review [Bateman et al. 2001; Johansson et al. 2012a; Björkdahl et al. 2013]. To our knowledge, this is the first systematic review of RCTs on complementary and alternative interventions for PTBIF. This study provides information regarding the knowledge of CAM interventions by evaluating the efficacy of CAM modalities when treating PTBIF.
The etiology of PTBIF is complex, as it is a multidimensional syndrome that includes physical, psychological, motivational, situational, and activity-related components [Lachapelle and Finlayson, 1998; Ouellet and Morin, 2006; Cantor et al. 2008]. Fatigue management guidelines compiled by Mock [Mock, 2001] include coping strategies, cause-specific interventions, pharmacologic interventions, and nonpharmacologic interventions such as exercise, nutrition, sleep therapy, and restorative therapy.
For physical and cognitive limitations, individuals with TBI must be monitored, and their lifestyles must be adjusted to minimize PTBIF. Regarding physical activity interventions, Borgaro and colleagues [Borgaro et al. 2005] suggested that engaging patients in physical activity increased endurance, improved restful sleep, and may help reduce fatigue. Aerobic exercise and other forms of physical activity reduce fatigue levels in individuals with cancer, multiple sclerosis, and other conditions [Krupp et al. 2010; Mitchell, 2010], but whether or not aerobic exercise benefits individuals with PTBIF warrants further studies. Among all physical interventions, only aquatic physical activity was effective for PTBIF [Driver and Ede, 2009], while fitness-center-based exercise [Hassett et al. 2009], Tai Chi [Gemmell and Leathem, 2006] and aerobic training [Bateman et al. 2001] had no impact on PTBIF. However, the evidence is weak because of the small and underpowered sample sizes.
Two studies indicated that CBI might be effective in reducing PTBIF [Johansson et al. 2012a; Björkdahl et al. 2013]. However, another study by Hodgson and colleagues questioned these results [Hodgson et al. 2005], which were considered ‘effective’ by Cantor and colleagues [Cantor et al. 2014] in the systematic review. TBI-related cognitive impairment and behavioral deficiencies can lead to increased PTBIF, as patients may lack the capacity and ability to expend the effort necessary to perform previously manageable tasks (i.e. the coping hypothesis). Through adopting a collaborative team approach to address patient concerns, cognitive and behavioral skills are beneficial to individuals with PTBIF [Heimberg, 2002]. Since research in other patient populations (multiple sclerosis, cancer and chronic fatigue syndrome) also suggests that CBI is effective in the management and reduction of fatigue [Montgomery et al. 2009; Krupp et al. 2010; Wiborg et al. 2010], CBI approaches are worth further pursuing.
The effects of the two related electro–biofeedback therapies were questionable, as they had elevated risks of bias [Smith et al. 1994; Schoenberger et al. 2001]. One study on light therapy presented better results in managing PTBIF, but further study is required to confirm its efficacy [Sinclair et al. 2014].
From the three additional RCTs [Bateman et al. 2001; Johansson et al. 2012a; Björkdahl et al. 2013] not included in Cantor’s systematic review, MBSR [Johansson et al. 2012a] and computerized working-memory training [Björkdahl et al. 2013] interventions were effective, but aerobic training was ineffective [Bateman et al. 2001]. Although the numbers of subjects with TBI included in these studies were small, they should not be ignored because of their positive results concerning PTBIF. In addition, some other CAM interventions have been demonstrated to be effective for fatigue, such as massage, tuina, qigong for chronic fatigue syndrome [Alraek et al. 2011], as well as yoga and acupuncture for cancer-related fatigue [Finnegan-John et al. 2013], but these interventions have not been examined for PTBIF. Therefore, it is worth determining whether these interventions can reduce PTBIF, since PTBIF management remains unstandardized and unsatisfied.
There are some limitations to the present systematic review. First, study quality varied among the included studies. Second, heterogeneity may exist because of differences in intervention style, study parameters, and outcome measurements, which impede comparisons between studies. Discrepancies also exist in the inclusion and exclusion criteria. Third, though a systematic search of multiple databases was undertaken, some unpublished grey literature might have been missed. Thus, potential publication bias and selection bias could not be eliminated. Therefore, these findings should be interpreted cautiously.
In conclusion, among complementary and alternative interventions, physical activity, MBSR, computerized working-memory training, and blue-light therapy may be beneficial in PTBIF treatment. However, due to the limited number of RCTs for each intervention in addition to methodological problems and high risks of bias in the most included studies, the quality of evidence was judged to be low/very low using the GRADE system. Further RCTs with larger sample sizes and more scientific rigor in particular for all of these interventions are necessary to determine the efficacy of these treatments in PTBIF.
Footnotes
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and publication of this article: Shaanxi Province Natural Science Basic Research Foundation of China (2016JM3015) and National Natural Science Foundation of China (81671097).
Conflict of interest statement
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.