Systematic Review of Lumbar Elastic Tape on Trunk Mobility: A Debatable Issue

Objectives To systematically review the literature to analyze the effect of lumbar elastic tape application on trunk mobility, surpassing the minimal detectable change of the used outcome measurement tool, and to analyze the additional effect of applied tension and direction of elastic tape application in low back pain and participants without low back pain. Data Sources Four databases were used: PubMed, Web of Science, Physiotherapy Evidence Database (PEDro), and Google Scholar. Study Selection The inclusion criteria were randomized and clinical controlled trials evaluating the effectiveness of lumbar elastic tape application on trunk mobility. Data Extraction Two researchers executed the search and a third author was consulted to resolve disagreements. The methodological quality was scored using the PEDro scale, with studies scoring ≤5 being excluded. Data Synthesis Eight out of 6799 studies were included; 5 studied individuals with low back pain, and 3 studied participants without low back pain. Two studies scored low on the PEDro scale and were excluded. None of the reported significant changes in trunk mobility due to elastic tape application exceeded the indicated minimal detectable change. No conclusions can be drawn from the direction and applied tension of elastic tape application. Conclusions Based on the results of this systematic review, there is no evidence supporting the effect of lumbar elastic tape application. We recommend consensus in the use of more reliable and valid instruments in future studies.

In patients with low back pain (LBP), staying active is indicated as the most important advice in rehabilitation. 1 However, back pain reduces the ability to move freely. 2 This adaptation is caused by altered trunk neuromuscular response in reducing the risk of noxious lumbar tissue stresses. 3 These adaptations eventually lead to painful muscle contractions and loss in trunk mobility 4 and are primary targets of physiotherapeutic interventions. 5 Physical therapists regularly use elastic tape application (ETA) in treating and preventing musculoskeletal disorders. 6,7 ETA is the application of therapeutic tape developed by Kenzo Kase in the early 1970s and has gained popularity in recent years. ETA is the gluing of an elastic cotton strip to the skin with an acrylic adhesive while the skin is stretched.
The superficial fascia within the skin is adjunct with the deep fascia and muscles owing to different types of connective tissues. 8 The kinematic behavior of skin and fascia are unique per layer. 9 Fascial kinematic studies showed that the skin, superficial fascia, and deep fascia will move in the cranial direction during trunk flexion and that the perimuscular fascia and muscle will move in the caudal direction during trunk flexion; the opposite occurs for trunk extension. [9][10][11] Previous studies investigating possible working mechanisms and the effect of ETA have confirmed that ETA causes heterogeneous deformations of the dermis, hypodermis, and deep fascia underneath the ETA, [11][12][13] increasing the bloodand lymph flow underneath the ETA and peripheral areas, 14 enhancing proprioceptive sensation, 15 decreasing subjective pain, increasing joint range of motion, 16 and enhancing muscle activity. 15,17 Moreover, it is suggested that the direction of ETA on the skin, over the muscle of interest, has specific effects on muscle activity. [18][19][20] Based on these effects, using these rationales as a basis for their clinical intervention, physical therapists often use ETA to improve mobility, enhance sport and functional performance, and treat musculoskeletal complaints, including patients with LBP. [21][22][23] A meta-analysis studying the effectiveness of ETA on pain and disability in a large variety of musculoskeletal complaints demonstrated moderate evidence that ETA reduces pain and very low evidence for an improvement of disability. 24 Recently, another meta-analysis on the effectiveness of ETA in musculoskeletal disorders on pain and disability, which only included studies comparing ETA with a sham condition, showed that pain does not immediately change posttreatment. However, pain does reduce during follow-up (range, 4-12wk). 25 Although it seems logical in patients with LBP to study the effect of ETA on pain and disability, we expected the largest effect to occur in the gain in trunk mobility (TRM) based on the above-described working mechanisms. Some studies have published the TRM outcome; however, this outcome was not included in the meta-analyses studies. Furthermore, in the studies published demonstrating a positive effect, the minimal detectable change (MDC) of the measurement tool used to measure mobility was not considered. The purpose of this study was to systematically review the literature to analyze the effect of ETA on TRM surpassing the MDC of the outcome measurement tool, and if effective, to analyze the additional effect of applied tension and direction of ETA in patients with LBP with respect to participants without LBP.

Protocol and registration
This systematic review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyzes guidelines 26 and was prospectively registered in the Open Science Framework (https://osf.io/2jhzg). The analyses were based on cumulative data from previously published studies, so no ethical approval was required.

Search strategy, inclusion, and exclusion of studies
A systematic literature research was performed independently by 2 reviewers (R.A. and L.B.). In the case of disagreement, a third independent reviewer was consulted. Publications were identified by searching multiple literature databases, including PUBMED, Web of Science, Physiotherapy Evidence Database (PEDro), and Google Scholar. To exclude gray literature, we decided that only the first 400 Google Scholar results would be screened. 27 A sensitive search string was used (table 1). The search was performed between September and December 2019 on "all fields"; the filter "humans" was applied if possible.

Study selection
In advance, narrative research was performed by 2 authors (R.A., K.N.) into the psychometric quality of various mobility measurement tools (table 2) to only include studies using reliable and valid instruments to measure TRM in terms of trunk range of motion (ROM) and to register the minimal detectable difference per outcome measure. Based on the quality of the instruments, studies were excluded using a mobile inclinometer application not reaching sufficient psychometric quality. Smartphones have a great range in sensor and software quality, which influence the reliability and validity of the smartphone range of motion applications. 39 Inclusion criteria were randomized controlled trials (RCTs) and controlled clinical trials (CCT) on the effects of ETA on TRM in participants without LBP and patients with LBP. Only trials with a control group were considered eligible (eg, ETA vs placebo/sham ETA and ETA vs no-taping intervention or usual care). ETA had to be applied to the individual's back with no restriction to the manor, technique, direction, or applied tension. Studies were included if TRM was a (primary) outcome measure with methods used according to table 2. No restrictions were applied to the search strategy regarding the date of publication. Studies written in Dutch, English, or German were included.

Data extraction
The first (R.A.) and third (K.N.) authors performed a systematic literature search independent from each other. The titles, abstracts, and full texts (indicated) of all articles were screened for inclusion by the reviewers. Data were extracted independently by the first (R.A.) and second (L. B.) reviewers from full-text articles. Data extraction included details on the ETA protocol used (eg, taping method, tape application direction, amount of stretch applied to the tape) and the effect on the TRM.

Standard measurement error
All calculations for measurement error parameters were performed with MATLAB. a Per instrument, the MDC was retrieved as an error parameter to analyze the meaningfulness of the significant improvements found in the systematically included study-results. If the MDC error parameter was absent, the SEM consistency was calculated with use of the interobserver reliability ð ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi 1 À ICC SD p SD Þ or the Pearson correlation validity ð ffiffiffiffiffiffiffiffiffiffi ffi 1 À r SD p Þ. Subsequently, it was used to calculate the MDC to determine the magnitude of change that would exceed the threshold of measurement error at 95% confidence interval ð1:96 Â SEM Â ffiffiffi 2 p Þ (see table 2).

Quality assessment
The methodological quality of the included RCTs and CCTs was assessed using the 11-item PEDro scale. The RCTs and CCTs had to compare at least 2 interventions, and one intervention had to lay within the scope of this systematic review. The PEDro scale has been described as a valid and reliable tool for the investigation of the internal validity of RCTs and has shown sufficient reliability for use in systematic reviews. [40][41][42] Full texts that met the eligibility criteria were independently assessed using the PEDro scale for methodological quality by the first and second reviewer (R.A., L.B.). In case of disagreement, a third independent reviewer was consulted for the final score. Studies scoring ≤5 on the PEDro scale were excluded owing to their low methodological quality that indicates they are less likely to yield meaningful results. 43

Description of studies
In total, 6799 titles were found in the initial search ( fig 1). Only the first 400 Google Scholar results were included (5140 were excluded). 27 Upon eliminating duplicates, the remaining 1613 titles and abstracts were reviewed, excluding another 1336 studies and determining eligibility for the remaining 277 articles. Of these 277 articles, 267 articles were omitted owing to a lack of relevant information (eg, no ROM outcome, no ETA research, and/or ETA in combination with another intervention), resulting in 10 studies that met the requirements for methodological evaluation. [44][45][46][47][48][49][50][51][52][53] After the methodological evaluation, 2 studies did not meet our quality criteria of a score of ≤5 on the PEDro scale and had to be excluded, 52,53 which resulted in 8 articles that met the criteria for inclusion. There was a good literature search agreement between the authors who reviewed the literature (k=0.736; P>.001). 54 scored "excellent" in 6 studies. [44][45][46]48,50 Moreover, 2 studies were scored as "good," 47,49,51 1 study was scored as "fair," 53 and 1 study was scored as "poor." 52 We excluded the last 2 studies from this review owing to their low quality because they might provide unmeaningful information concerning ETA effects on the TRM (table 3).

Differences in study design
All studies used the ETA in the experimental group and compared it with a control group (sham ETA intervention or nonintervention). However, there was a difference in study design. Some studies studied the effectiveness of ETA on the TRM in situ (condition 1), but others evaluated the effect after ETA was removed (condition 2). Two of the 8 studies evaluated condition 1, 48,49 3 of the 8 studies evaluated condition 2, 44,51,56 and 3 studies evaluated the effect of both conditions. 45,47,50 The characteristics of the 8 studies included in the systematic review are shown in table 4.

Flexion ROM
There is conflicting evidence with regard to the effectiveness of ETA on TRM. Regarding condition 1 with ETA in situ, none of the control groups showed a significant change in flexion ROM (FROM). Two included studies did not find a significant improvement in FROM in patients with LBP. 48,50 In contrast, 3 studies found a significant improvement in FROM in patients with LBP 45,47,48 and 2 studies found a significant improvement in FROM in individuals without LBP 47,49 regarding condition 1 ETA. Two of these studies concluded that ETA affected the FROM positively measured with the Finger Floor Distance test (FFD). 47,48 Preece and White 48 concluded that this FFD improved immediately in patients with LBP after the ETA was applied. In addition, Lemos et al 47 concluded the same in individuals without LBP after the ETA was worn for 48 hours. Both studies did not exceed the threshold of FFD measurement error at a 95% confidence interval using the MDC 95 of 9.8 cm. 28 Based on the MDC 95 , the significant change reported in these studies did not surpass the MDC. 47,48 One study reported a positive effect of ETA on the FROM with an inclinometer (fleximeter) after the ETA was worn for 1 week. In this study, a significant evolution was found between baseline and the 1-week follow-up. 45 The inclinometer MDC 90 indicated that the observed change in this study was made within the intraobserver error (MD=5 §SD2<MDC 90 =7). The significant change in this study cannot be considered as a true change. 32 Shin and Heo 49 found a significant FROM evolution in individuals without LBP measured with a BROM II device, reporting that it was beyond its MDC 95 (MD=5.74 § SD1.33>MDC 95 =2.83). 34 Also, conflicting evidence is present in condition 2. Regarding the evaluated effectiveness of ETA on FROM in condition 2, 3 studies failed to demonstrate significant FROM change in patients with LBP, 50,51,56 and 1 study in individuals without LBP after ETA was removed. 47 By contrast, 3 studies reported a significant change in condition 2 for patients with LBP 44,45,51 and 1 in individuals without LBP. 47 Two of these studies evaluated the effectiveness of ETA using the Schober flexion test. In both studies, the FROM increased positively from baseline to week 2. 44,51 The demonstrated significant difference in both studies did not exceed the MDC 95 measurement error of 1.8 cm. 28 The third study reported a significant effect in patients with LBP on FROM measured with an inclinometer, 45

Extension ROM
No study found a significant difference regarding the extension ROM.    improve and is not better than X-sham tape.
Tape 2: paravertebral muscles ETA: 2 I-stripes were used (5 cm wide and~30 cm long) and were applied paravertebral at the height of the erector spinae started from~L5/S1 to the area below the scapula.
Within time (mean difference §SD): Control set-up: The GRr received sham tape, which was applied in an Xshape that crossed around T12.   After 7 days of ETA in situ, there was no significant treatment effect found on the flexion ROM after the ETA was removed. In conclusion, ETA in CON2 has no treatment effect in terms of improved flexion and extension trunk mobility and is not better than rigid paravertebral tape application.
Control set-up: The CGr received 2 "I" shaped rigid tape strips, which were applied using the same protocol as in the CGr.     The tape was reduced to 0% before it was adhered to the skin over the muscles.
Between group pre-post difference:

Tape direction and applied tension
Six out of 8 studies evaluated the caudal to cranial paraspinal placement of the ETA effect and no other placement. 44,[47][48][49][50][51] The ETA tension varied between approximately 0%-50%. The effects of differences in tensions on TRM effect was evaluated in 4 studies. 44,47,48,50 The ETA tension difference did not exceed the MDC 95 and was deemed irrelevant.

Discussion
This systematic review demonstrates that ETA does not affect TRM in individuals with and without LBP. Those studies that reported a significant positive effect of ETA on TRM did not consider the MDC of the outcome instrument. 44,[45][46][47][48]51 None of the studies included in this systematic review surpassed the MDC in their results, and none of the measurement tools were precise and accurate.
The quality of the included RCTs, assessed by PEDro score, was generally high, despite the small sample size in all studies. However, the methodological quality scored by PEDro is based on the design and does not account for the psychometric quality of the mobility measurement instruments used. Hence, in this systematic review, we used the MDC to correct for this.
No conclusion can be drawn regarding the effects of ETA direction because none of the included studies investigated the effect of the ETA direction on TRM. This is surprising because the opposite direction of ETA application has been described to have an effect. [17][18][19][20]57 However, most (6 of 8) of the studies applied the ETA from caudal to cranial over the lumbar paravertebral muscles and did not compare the effect of this direction with that of the opposite (cranial to caudal) ETA direction. Moreover, the resistance in elastic tape is determined by the Youngs's modulus of the tape and tension that is applied to the tape. 58 The studies that were included in this review used variable types of tape brands and, subsequently, the applied tension differed between studies. According to a biomechanical engineering study, each brand of elastic tape creates a different level of tension at equivalent strain conditions and as such, the ETA tension could be interindividual depending on body characteristics and skin flexibility. 59 Hence, the variability in tension may be an explanation for the lack of a meaningful effect.
In general, the elastic tape applications used in the studies are not similar to those applied in clinical practice since longitudinal paravertebral, diagonal thoracolumbar fascia, and horizontal elastic tape applications are used. [60][61][62] Moreover, different ETA directions are tested before ETA in choosing the best ETA direction. 63

Study limitations
The strength of this study is that narrative research was done in advance into the psychometric quality of various mobility measurement instruments to analyze the effectiveness of ETA. The validity, reliability, and MDC were retrieved from these studies. If the MDC was absent, the SEM and MDC 95 were calculated. Subsequently, the strictest MDC error-parameter was used to evaluate the meaningfulness of the reported significant changes in TRM influenced by ETA. We excluded studies in which a mobile phone inclinometer application was used to evaluate the effect because of the insufficient reliability and the various downloadable applications. Among RCTs and CCTs, only trials with a control group were considered eligible. However, we excluded RCTs and CCTs that used ETA in a multimodal interventional setting because of the possibility that the ETA could interact with the other interventions used.
There are also limitations in this systematic review. In general, the included studies are not homogeneous concerning the study population. Moreover, different evaluation time points were used to evaluate the TRM effect and there are unstandardized or unknown standard measure procedures, failed sham elastic tape methods, unstandardized elastic tape application methods, and inconsistencies between studies concerning the elastic tape sessions and the duration of the elastic tape in situ. Limitations to consider on the part of the authors responsible for the synthesis of this systematic review were the inclusion of moderate impact strength articles, the methodological quality assessment tool used with the posed limitations, the search methods, and keywords used.

Conclusions
Based on the results of this systematic review, there is no evidence supporting the effect of ETA. There is no evidence that ETA improves TRM. There is no evidence regarding application tension or direction of ETA influencing TRM. It is necessary to conduct further high-quality methodological research on the effect of ETA application on TRM to ascertain whether an effect on mobility is present. This indicates the use of more valid and reliable TRM measure instruments within a standardized measure protocol. When evaluating lumbar ETA effects on TRM and considering the psychometric quality, we cannot confirm its effects on TRM. Based on this systematic review, current lumbar ETA interventions should be questioned.