Immediate Effects of Spinal Manipulative Therapy on the Performance of Elite Soccer Players. A Pilot Randomized Controlled Trial With an Internally Validated Placebo.

Asymptomatic spinal biomechanical alterations may generate impaired central nervous system proprioceptive input, and motor control, and central processing decits, which could impair sports performance. Several of the demonstrated neurophysiological effects of spinal manipulative therapy (SMT) could theoretically inuence sports performance. Further, some studies have demonstrated that SMT' may improve aspects of sports performance. These include improved full-swing in golfers, judoka grip strength, taekwondo muscle strength, and ball kicking speed in soccer. The purpose of this study was to assess the effect on the performance of elite athletes of one session of SMT to correct asymptomatic, biomechanical dysfunction, using professional soccer players and sprint and change of direction (COD) tests. Also, the proposed placebo intervention was tested for ecacy in achieving blinding. A parallel-randomized, controlled, clinical trial was performed with 20 elite soccer players who were randomly assigned to one of two groups by a coin ip: one receiving SMT (n=10) and the other a placebo intervention (n=10). All players were from the same team, and were injury free and naive to SMT. Data were collected at the team training facility. Measured outcome tests (30-meter sprint run with a 10-meter split and COD test) were performed at the same time by all subjects immediately prior to and after interventions. Photocell devices were used for data acquisition. Participants and those assessing outcomes were blinded to group assignment. Care givers were not able to be blinded.


Background
Asymptomatic spinal biomechanical alterations may generate impaired central nervous system proprioceptive input, and motor control, and central processing de cits, which could impair sports performance. Several of the demonstrated neurophysiological effects of spinal manipulative therapy (SMT) could theoretically in uence sports performance. Further, some studies have demonstrated that SMT' may improve aspects of sports performance. These include improved full-swing in golfers, judoka grip strength, taekwondo muscle strength, and ball kicking speed in soccer. The purpose of this study was to assess the effect on the performance of elite athletes of one session of SMT to correct asymptomatic, biomechanical dysfunction, using professional soccer players and sprint and change of direction (COD) tests. Also, the proposed placebo intervention was tested for e cacy in achieving blinding.

Methods
A parallel-randomized, controlled, clinical trial was performed with 20 elite soccer players who were randomly assigned to one of two groups by a coin ip: one receiving SMT (n=10) and the other a placebo intervention (n=10). All players were from the same team, and were injury free and naive to SMT. Data were collected at the team training facility. Measured outcome tests (30-meter sprint run with a 10-meter split and COD test) were performed at the same time by all subjects immediately prior to and after interventions. Photocell devices were used for data acquisition. Participants and those assessing outcomes were blinded to group assignment. Care givers were not able to be blinded.

Results
20 participants were analyzed, 10 in each group. There were no changes to the sprint (10m and 30m) and COD test results immediately following either of the interventions. All subjects (SMT and placebo) answered YES to a question after the intervention asking if they were treated by SMT. No adverse events were reported.

Conclusion
Spinal manipulative therapy to correct asymptomatic, biomechanical dysfunction in the spine and pelvis did not have any immediate effect on the performance of elite soccer players as measured by 10-and 30-meters sprint times and COD sprint times. Additionally, the proposed placebo strategy was successful in blinding these athletes.

Background
An integral part of competitive sport for athletes is the pursuit of performance enhancement. [1] Enhanced performance is sought from many contributing factors, including physiological, dietary, physical, social and economic ones. [1][2][3] There is some evidence that common forms of spinal biomechanical dysfunction, which are often asymptomatic, may generate impaired central nervous system (CNS) proprioceptive input, and motor control and central processing de cits. [4,5] This could result in performance loss. [4][5][6][7] Such vertebral, biomechanical dysfunctions seem to be responsible for maladaptive CNS neuroplasticity. [6,8] When addressed with spinal manipulative therapy (SMT), there may be improvement in joint proprioception and motor response. [6][7][8][9] SMT has no consistent de nition in the literature. [10,11] For present purposes it is de ned as any manual technique which incorporates a controlled, high-velocity, low-amplitude dynamic thrust applied to a joint, taking it beyond the passive range of motion but without exceeding anatomic limits at the paraphysiological space. [10,11] The choice of direction of this controlled thrust is based upon physical examination ndings of speci c joint dysfunctions. [11] This treatment method for spinal biomechanical functional disorders is widely used by chiropractors [10,[12][13][14] and other health professionals, including orthopedists [15], physiotherapists [16], and osteopaths [17].
Further, there are studies which have demonstrated that SMT' may in uence and improve aspects of sports performance. These aspects include increased full-swing for golfers [40], increased judoka grip strength [25], and increased ball kicking speed for soccer players [41]. However there are few studies in this eld, and there is a lack of data regarding the in uence of SMT on sports performance from laboratory studies, eld tests, or real events. [42] In soccer, athletic trainers and coaches use various eld tests to assess the tness and capacity of their players. [43,44] These include sprint (10 and 30 m) and change of direction (COD) tests, which are widely used to evaluate ability to sprint and to change direction quickly. [45,46] Their usage ranges from guidance for individual training routines to decisions on collective competition tactics, as when used to choose the most suitable player for a particular position for a speci c match. [45,46] Employing these test methods, individual training programs can be developed to improve a player's ability to accelerate (10 m), increase maximal velocity during sprints (30 m), and abruptly change direction while running at speed. Additionally, these tests can provide guidance on improved running technique and use of energy systems. [45,46] This study hypothesized that athletes given SMT to correct asymptomatic spinal biomechanical dysfunction would have better physical performance in eld tests when compared with others in an appropriate placebo group. Its aim was to study whether such SMT may in uence the performance of elite, soccer players as seen in sprint and COD tests. The proposed placebo intervention was also tested for blindness e cacy.

Trial design
A parallel randomized, controlled, clinical trial was performed with an allocation ratio of 1:1 between the two intervention groups (SMT or placebo). Ethics and research committee (IRB) approval for the study was received from the Instituto Mantenedor de Educação Superior, Brazil, and the study was prospectively registered at the Institute's database under the number 3993. The study protocol was also retrospectively registered at the ISTRCTN registry under the identi cation number ISRCTN29691307. This trial adheres to CONSORT reporting guidelines.

Participants
All team players at the under-20 years category (U20) were invited to participate and did so (population study). As a result, twenty Fluminense Football Club elite, professional, soccer players were included. The players' regular physical training schedule comprised ve sessions per week, with the duration of each session typically being 120-180 minutes. Each session included soccer training, strength training and conditioning.

Eligibility criteria
All subjects included were male, volunteers, and elite soccer players from the same team. Inclusion criteria were ability to understand and signature of an informed consent form, participation in training or competition at least 5 days a week, and no previous experience of SMT.
With respect to non-inclusion criteria, the athletes were assessed by a physician quali ed in both medicine and chiropractic, and with postgraduate sports medicine quali cations in both disciplines. First, this assessment was for a number of common contraindications to SMT, any of which would lead to non-inclusion in the study. [47] These included acute fracture, acute infections, neurological de cits, signs of joint instability or pathological ligament laxity, the absence of biomechanical spinal joint dysfunction amenable to SMT, and other contraindications to SMT given in guidelines from the World Health Organization. [47] Second, the assessment was for another non-inclusion criterion, which was the presence of any acute musculoskeletal lesion that might prevent the subject from full participation in the study tests. A nal non-inclusion criterion was previous experience of SMT treatment.
Settings and data collection Interventions All tests were conducted and data collected after subjects performed a standard warm-up consisting of 15 minutes of low-intensity running and striding, followed by three sub-maximal sprints. Then, in a line, they performed the sprint test sequentially. After completion of this by all subjects, they lined up again to perform the COD test sequentially. All subjects then attended the intervention o ce. After the entire group completed the intervention stage, which took about 3 hours, the warm-up was performed again and the subjects were retested. This ow of procedures is summarized in Fig. 4 (Results Section).
In the intervention o ce subjects visited one of the two treating physicians available, both quali ed as chiropractors, and having ongoing postgraduate training in sports chiropractic and at least three years' experience of chiropractic practice. At the o ce the physician randomly assigned them to the SMT or placebo group on a coin ip.
All subjects in both groups had received the same general information about the treatment under investigation and its possible adverse effects. They spent a similar time at the treatment room with the treating physician, namely between 10 and 15 minutes. Each subject attended a single intervention session.

Spinal manipulative therapy
For those in the SMT group, correction of spinal biomechanical dysfunction was performed as needed, which was determined by dynamic and static evaluation of all spinal (cervical, thoracic, and lumbar) and pelvic (sacro-iliac) joints. Physical examination ndings used as indicators of joint dysfunction included abnormal or blocked joint play, restricted inter-segmental motion, soft-tissue tenderness over relevant joints, and paraspinal asymmetric muscle tension. These ndings are commonly used by physicians as indicators of spinal biomechanical dysfunction. [11] Treatment was given employing high-velocity, low-amplitude (HVLA) Diversi ed Technique. [11] All subjects in this SMT group received some treatment, though the number of treatments and spinal regions targeted varied because treatment was based upon clinical ndings. Cervical SMT was performed with the subject lying supine, thoracic SMT with the subject lying prone, and lumbo-pelvic SMT with the subject in a side-lying position. Treatment was delivered using hands only, and on a Thuli Table (Thuli Table Inc, Dodgeville, WI).

Placebo
The same physicians that performed the SMT performed interventions in the placebo group. They used a Thuli Table (Thuli Table Inc, Dodgeville, WI) and its drop mechanisms, addressing three different areas -the cervical, thoracic, and lumbo-pelvic regions of the spine.
With the subject lying prone, the physician activated a mechanism that lifts a speci c section of the table, either the cervical, thoracic, or lumbo-pelvic sections. The section dropped back to its neutral position on application of pressure to the margin of the elevated section by the physician, done without physical contact with the subject. When the raised section fell back to a neutral position its impact with the main body of the table generated a loud clap. This procedure was performed 3 times for each spine-related area (cervical, thoracic, and lumbo-pelvic), providing a total of 9 maneuvers for each subject in the placebo group. A similar placebo intervention has been used in a previous study. [25] Outcomes The outcome tests (30-meter sprint run with a 10-meter split, and the COD test) were performed after a standard and simultaneous warm-up by all subjects, and immediately prior to the interventions and after approximately 3 hours. The subjects were placed in a line and consecutively performed the tests using the same track line and measurement devices. The researcher responsible for the testing was not aware of the group allocation of each subject. Only the two care providers were aware of group allocations. They had no access to each other during data acquisition.
The sprint test was carried out on a 30 meters straight-line track, using a system containing three photocell devices (Microgate, Bolzano, Italy). These devices were set one at the beginning, one at 10 meters (split time), and one at 30 meters mark (Fig. 1). The starting position was with the preferred foot just before the starting line. Each subject decided when to start, meaning that data were not in uenced by differences in reaction time. [48] The rst device or barrier of photocells was located immediately after the starting line, and time was recorded when the subject's body crossed this and then the next two photocell barriers. During the entire execution of the test, athletes were verbally encouraged. This procedure was performed two times, with a rest interval of 5 minutes in between. The lowest value (faster time) was used for analysis. [44] The change of direction test was conducted over a length of 20 meters. Subjects were required to run in a zigzag pattern while crossing four, 5-meters distance barriers, each one located at a 100° internal angle from the next one (Fig. 2). [49] Time was measured by a system of two photocell barriers (Microgate, Bolzano, Italy), one at the beginning of the track and another at the end. Starting position was the same as used for the 30-meters run. During the entire execution of the test subjects were again verbally encouraged. The test was performed twice, with a rest interval of 5 minutes in between. The lowest value (faster time) was used for analysis. This test provides rapid acceleration and deceleration movements, similar to the real demands of a match.

Sample size
All team players at the under-20 years category (U20) were invited to participate and did so. As a full population study, no sample size calculation was performed.

Allocation concealment
Group allocation was performed just before the intervention, by means of a coin-ip by the physician at the intervention o ce, before physical examination. No one else had access to information on allocation. There was no contact between the physicians and those assessing the outcomes (sprint and COD).

Blinding
Those assessing outcomes and performing statistical analysis were blinded to each subject's group allocation. Care providers, given the nature of the interventions, could not be blinded. Subjects were blinded to interventions. To assess the blindness of subject, right after interventions all completed a form asking them to answer one question: "Have you been treated by spinal manipulative therapy procedures?".

Statistical methods
Shapiro-Wilk's and Levene's tests were used to test the data for normality and homogeneity of variance. Two-way, mixed-model ANOVA was used to assess for differences between the effects of a single session of SMT or placebo on sprint and COD test times.
Intervention (SMT and placebo) and time (pre and post-intervention measures) were used as factors. The partial eta squared (η p 2 ) was calculated to determine the effect size for the interactions. Cohen's d for the paired test was used to determine the effect size when there was a main time effect. T-test or Chi-square assessed baseline characteristics. Alpha value was established as 5%. Statistical testing was carried out using SPSS Version 20 (IBM Corp., Armonk, NY).

Results
All team players (n = 20) were randomly assigned in one of the two groups (SMT or Placebo). All subjects received the intended intervention. Data from all subjects were included for analysis. Figure 3 summarizes the study ow. Also, for tests and interventions sequence, subjects progressed through the steps shown in Fig. 4. All athletes performed these steps during the same time. Following randomization each of the SMT and placebo groups had 10 subjects. There were no losses in either group. All individuals were analyzed for baseline characteristics, which are shown in Table 1. Comparisons between groups at 10-meters, 30-meters and on the COD tests are shown below in Table 2. Table 2 The mean (± SD) of time (s), P-value, effect size (ES) and 95% con dence interval of pre and post SMT and placebo groups for 10 m sprint, 30 m sprint and COD test.

Placebo validity internal test
All subjects in both groups (n = 20) answered YES to the question "Have you been treated by spinal manipulative therapy procedures?" asked after the intervention by an independent researcher who was not aware of group allocations. Accordingly, the placebo intervention employed in the trial was effective in blinding these athletes, as showed in Table 3.

Discussion
This is the rst study to test the impact on the performance of elite soccer players of SMT to correct spinal biomechanical dysfunction, with performance based on sprint and COD tests. Team staff commonly use these tests to predict physical performance in professional soccer. [43,44] The present pilot study reports no change in sprint or COD outcomes after SMT.
Although there is a rationale to support the hypothesis that SMT to address biomechanical dysfunction may have a positive in uence on sports performance, [42] there is a lack of compelling evidence in support, a lack that has been observed and investigated by other middle-distance runners. Even though the protocols and populations were different in these two studies, their combined results suggest that SMT to address asymptomatic spinal dysfunction seems not to in uence sprint times in the athletic population.
One plausible explanation relies on the complexity of sprinting, which is in uenced by several variables that may mask any potential bene t from SMT. Such variables include muscle activation patterns, motor control, movement coordination, and anaerobic resistance.
One possibility is that there are positive effects of SMT, but that these do not have su cient impact to overcome other factors involved in sprinting and be seen in the nal outcome measured, sprint time.
Similar SMT investigations to ours, though using different outcomes, have shown increased performance after SMT for biomechanical dysfunction.  [7] reported ankle plantar exor muscles increased strength after SMT lasted more than 30 but no longer than 60 minutes. Also, the range of motion and balance variables associated with the full-swing gol ng movement may be in uenced by SMT in a different way to the main variables in sprinting.
The one previous study that analyzed the effects of SMT on soccer players was a non-randomized study performed by Deutschmann et al. (2015). [41] They found an increased ball-kicking speed after lumbar or sacroiliac SMT or a combination of both (average increase of 3.52 to 6.57 km/h, depending on group allocation, P < 0.05). Data was acquired immediately before and after the interventions (SMT or sham). Their results, reporting a more positive response to SMT than in our study, may suggest that SMT has a different impact on tasks with different motor control, strength and complexity. We could not nd any other study where SMT effects were evaluated for a COD test.
Our study is the rst study of its type, and two critical methodological aspects are worth noting with some emphasis. One is that all subjects were naive to the treatment being given (SMT), and the other is that the placebo intervention was internally validated. Despite the signi cant mechanical and physical differences between the chosen placebo maneuver and the SMT delivered in the study, subjects in the placebo group believed that the intervention they were receiving was the active one being studied. This may have been because these subjects were naive to SMT, and because of the combined effect of the drop mechanism moving their body segments and causing the loud sound described.
Also, regarding study design strength and weaknesses, we believe collecting data from all athletes at the same time was noteworthy because it avoided between-subjects interference from weather, such as effects of wind, sun exposure and temperature. This is especially signi cant when collection of data is in open eld, as was the case in our study. All of the subjects (n = 20) completed each test within 15 minutes (2 trials). On the other hand, this design did not allow assessment of the effects of SMT immediately after it was performed, as could be done with assessments on an individual basis.

Limitations
Conducting clinical investigations with high-performance athletes can be very challenging. This is especially true because of the inherent di culties of having a large sample size and of conducting a trial while the athletes are training or competing, and with minimum or no impact on their routine. [1] Two different physicians treated the study participants, and the treatment provided was based upon their skills and clinical judgment to identify and correct biomechanical joint dysfunctions. This subjective aspect of the treatment may represent a limitation. However both had formal quali cations as chiropractors, were graduates of accredited university schools of chiropractic and had at least 3 years of chiropractic, clinical experience.
Execution of the tests by the subjects was done consecutively, but the order of execution prior to and after the interventions were not recorded or required to be the same. This means the interval between outcomes measurement performed after the interventions may have varied up to approximately 3 hours between subjects. Even thought the subjects were compared with themselves, and it is thought that this limitation should not have in uenced outcomes, this should be taken in consideration.
The research protocol did not address lifestyle factors that might lead to performance bias, such as nutrition and patterns of sleep.
That was because it was judged that the potential in uence of these was minimal to zero because of the study design, with its single treatment session and immediate pre-post measurements of subjects. All subjects were fed at the team facilities and offered the same meal times and food. All subjects were using their regular sports gear.
Although the subjects were instructed not to discuss the treatment received with their teammates, some may have and this was beyond our control. However any such discussion is unlikely to have in uenced the outcomes because, as shown by the internal placebo validation test, all participants believed they were treated by SMT.

Conclusion
The correction of asymptomatic, spinal biomechanical dysfunctions by means of spinal manipulative therapy (SMT) did not have an immediate effect on the performance of elite soccer players as measured by common sprint (10-and 30-meters) and change of direction (COD) test sprint times. There were no appreciable changes observed for subjects in either the SMT treatment group or in the placebo group.
An additional nding in this study was that the proposed placebo strategy was successful in blinding these subjects, athletes who were naive to SMT treatment. .  Study ow diagram.

Figure 4
Tests and interventions' participants ow diagram.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. CONSORTChecklistSOCCERPERFCMT.docx