A single bout of coordination training does not lead to EIH in young healthy men – a RCT

Introduction

Physical activity is able to induce a so-called exercise-induced hypoalgesia (EIH), which is described as a decrease in pain sensitivity after exercise [1], [2]. Research has indicated that a single bout of endurance as well as strengthening exercises, depending on dosage and intensity, are associated with a decrease in pain ratings for experimentally induced pain as well as an increase in pain thresholds in healthy people [2]. While analgesic effects could be observed for the above-mentioned exercises, nothing is known about acute effects of coordinative exercises on the somatosensory system in healthy people.

Ackerley et al. (2019) demonstrated, that nerve function can be affected by external stimulation. The activation of Aß-mechanoreceptors via vibration modulated muscle sympathetic nerve activity and reduced pain in a healthy population [3]. Moreover, it is known that there exists a multimodal interaction of sensory modalities [4]. One stimuli can modulate or overlap the perception of another stimuli. For example, rubbing the skin after bumping your arm can temporarily alleviate the pain. Transferring this, coordinative activity, including proprioceptive stimuli, predominantly assigned to the group of Aß-afferents and generated from various receptors in skin, joints and muscles could modulate the perception of (noxious and innocuous) stimuli from other sensory modalities. This could also be supported by a few studies, observing analgesic effects of coordination or sensory-motor training in patients with chronic pain, induced by knee osteoarthritis or chronic low back pain [5], [6]. But, depending on the type of pain, results in populations with chronic pain states were heterogeneous. Furthermore, somatosensory function is already altered in this populations [7], [8] and might be more or maybe less adaptable by external stimuli. In addition, only long-term training interventions were investigated and pain was quantified only by subjective rating scales, so far.

On the basis of these aspects and to better understand physiological principles of potential alterations in pain perception, in a first step the influence of a single bout of coordination training (CT) on somatosensory function of a healthy population should be investigated by using Quantitative Sensory Testing (QST). The QST is a semi-objective tool for pain measurement and frequently used to detect pain sensitivity and pain mechanisms. By means of QST, thermal as well as mechanical stimuli are applied to measure sensory function [9]. Based on the above-mentioned hints, the authors hypothesised a decrease in pain sensitivity, what means an increase in detection and pain thresholds, induced by a single bout of CT in healthy men.

Material and methods

Study population and experimental design

Thirty-five healthy young men between 18 and 35 years were included in this study. To avoid side effects based on gender differences in pain perception only men were recruited for answering the purpose of this study [9]. Exclusion criteria were acute or chronic pain conditions, regular pain medication and any medication within the last 24 h before examinations as well as any kind of diseases influencing perceptive sensation, cognitive and pain perception (e.g. neuropathies, dementia, depression, multiple sclerosis etc.). In addition, participants were excluded in any case of surgeries 12 months before study inclusion. To ensure that all requirements were met, participants were screened using a general anamnesis questionnaire, the German Pain Questionnaire [10] as well as the painDETECT questionnaire [11]. Participants were recruited via notices in local institutions and modern media. The presented study was conducted in accordance with the principles of good clinical and ethical practice and received approval from the local ethic committee of the University of Wuppertal (Germany). Along with the Declaration of Helsinki, all participants gave written informed consent after being informed about the study protocol.

This study was performed as a randomised crossover design including two different interventions within a period of seven days: a CT and a resting session as control condition. Both interventions lasted 45-min and all participants absolved both interventions. Via block randomisation, participants were divided to start with the CT or the control condition. Trial profile is presented in Figure 1. During control condition participants stayed in the examination room with the choice to sit on a chair or lay on an examination bench, whereas they were not allowed to sleep during this condition. Participants were examined before and after each intervention by means of QST [9]. All QST-measurements were absolved by the same experienced researcher. Before starting with the first examination, each participant provided information about his subjectively perceived dominant leg. Measurement point at dominant leg was the dorsum of the foot [9], because all exercise were chosen to enhance the proprioceptive input on the ankle joint.

Figure 1:

Trial profile of the randomised crossover design including coordination training as well as resting as control condition (n=number).

Coordination training

Adapted to the four-staged model of Diemer and Sutor (2007), all exercises were arranged from low to high coordinative demands as well as from low to high sensory input (see Table 1) [12].The CT included nine standardised exercises to generate high sensory input for the ankle joints (see Figure 2). All nine exercises were performed barefoot in standardised order. Exercises were performed on a yoga mat (Table 1 – number: 1, 2, 3, 6), a balance pad (Table 1 – number: 4, 5, 7, 9) or a swinging platform (Table 1 – number: 8). Three different balance pads (air-filled and plain surface; air-filled with spiky knobs; foam rubber with round knobs) were used to generate varying sensory input for the feet. The exercises were performed in three sets (each set lasting over 30 s) with 30 s intervals for relaxation between each set. To instruct the participants into next exercise, 2 min breaks were interposed between the exercises. All exercises were selected with little as endurance and strengthening input as possible in order to avoid side effects of these activity types [2]. In order to standardise instructions and duration of exercise, the performance was filmed in the run-up to the study and showed to the participant on a screen during the intervention. In addition to the instruction video, a researcher supervised the whole training and gave feedback to the participant in case of incorrect performance.

Table 1:

Exercises of the coordination exercise training categorized to the level of coordinative demand and sensory input.

No.	Exercise	Level
1.	Standing in one leg stand with the dominant leg on an unstable platform.	A, B-I
2.	Fast movement of ankle joints with minimal forward motion (ankle drills).	B-II
3.	Positioning both feet at a precise distance apart (heel to toe) and simultaneous rotation of the head.	A, B-I
4.	Standing in one leg stand with the dominant leg on an air-filled balance pad including short touch-downs of the toes in front and behind the balance pad.	B-I
5.	Standing in one leg stand with the dominant leg on an air-filled balance pad with spiky knobs. Positioning the upper body as well as the free leg in horizontal position and back.	A, B-II
6.	Standing in one leg stand with the dominant leg and lifted knee of the free leg as well as outstretched arms overhead. Moving the outstretched arms in front of the hip while stretching the free leg behind the body and back.	B-II
7.	Standing in one leg stand with dominant foot on a balance pad with smooth knobs and bouncing a ball simultaneously.	B-I, C
8.	Side steps on an unstable, swinging platform.	B-II, C
9.	Squats with rotation of the upper body while standing on an air-filled balance pad with spiky knobs.	B-II, C

1. Level A=depth perception; Level B-I=static stability with variation of the sensory input; Level B-II=dynamic stability; Level C=reactive stability [12].

Figure 2:

Presentation of single exercises of the coordination exercise training. For precise description of each exercise see Table 1. Numbers presented in the pictures stood for exercise numbers simultaneous to Table 1.

Statistics

Statistical analyses were performed in accordance to the protocol of the German Research Network on Neuropathic Pain [9] using IBM SPSS 22^© for Windows (USA). QST-parameters which were not normal distributed (CDT, WDT, TSL, MDT, MPT, WUR, PPT), were transformed into a normal distribution via log-transformation. To avoid a loss of zero ratings, a constant of 0.1 was added prior to log-transformation for MPS and WUR. In cases of pain rating valued with zero for one part of WUR, extremely high WUR results in the analysis. Based on this, higher scores with more than twice of standard deviation were excluded in the analyses.

By performing a variance analysis with repeated measurements (ANOVA), the factors group (CT, rest) *time (pre, post) and *stimuli (QST-parameter) were analysed. The Greenhouse–Geisser adjustment was used to correct for violations of sphericity. Because no significant effects could be observed in the main analysis, a post-hoc test was not performed. Though, in a secondary analysis, the differences from pre to post measurement (delta-values) for both conditions were analysed. Based on not normal distributed raw-data a Wilcoxon-test was used for this analysis.

Based on an effect size of 0.1, representing a small effect size (Cohen’s f) and a two-sided alpha level of 0.05, we calculated a sample size of 33 participants would provide 80% power (G*Power 3.1, Germany). Descriptive data as well as QST-data (raw and log-transformed) are presented as mean and standard deviation (mean ± sd). Level of significance was set to 5% (p≤0.05).

To illustrate the participants’ somatosensory profiles, pre and post data of each condition were analysed by Z-transformation using the following expression [9]:Z-score=(individual value_{post condition} − mean_{pre condition})/sd_{pre condition}

The calculated Z-score reflected the participants’ sensitivity for an applied stimulus. An increase in Z-scores (>0) indicates a gain of function, whereas a decrease of Z-scores (<0) indicates a loss of function. Zero-axis demonstrates mean values of the related pre-measure [9].

Results

Thirty-five young healthy men (mean age 27 ± 3 years) were included in the study and all included participants completed both conditions. Anthropometric data of the cohort is shown in Table 2. The main analysis of the QST data in respect of the factors group, time and stimuli showed no significant main effects. Neither for the factor group*time (p=0.51), nor for the factor group*time*stimuli (p=0.32) significant alterations could be detected. Thus, all stimuli remained constant in the course of the CT as well as the control condition (e.g. mean ± sd of pre CT/post CT: CPT in °C: 9.3 ± 8.8/8.5 ± 9.1; PPT in N: 85.8 ± 27.2/90.2 ± 27.1). Comparison of the pre to post-differences for both conditions showed no statistically significant alterations of the QST parameters of the participants, also. Course of the data of both conditions are presented as somatosensory profile of the participants in Figure 3 and as well as raw and log-transformed data in Table 3.

Figure 3:

Participants Quantitative Sensory Testing (QST) profile after proprioception exercise training and control condition. CDT=cold detection threshold; WDT=warmth detection threshold; TSL=thermal sensory limen; CPT=cold pain threshold; HPT=heat pain threshold; MDT=mechanical detection threshold; MPT=mechanical pain threshold; MPS=mechanical pain sensitivity; WUR=wind-up-ratio; VDT=vibration detection threshold; PPT=pressure pain threshold. Zero-axis value=mean value of the related pre-measures. The zero-axis represents the mean-value of related pre-measures of intervention or respectively control condition, n=35.

Discussion

Aim of this study was to investigate, if a single bout of CT leads to an EIH in healthy men. The results presented did not show any significant alterations in pain sensitivity of the participants following CT in comparison to the control condition. Thus, we have to disprove our hypothesis of reduced pain sensitivity and therefore an EIH following a single bout of CT.

In addition to the main analysis, a secondary analysis was performed with a Wilcoxon-test, because an ANOVA is statistically robust for so many parameters as intended for the QST. By means of this test, the differences of each parameter from pre to post measurements (delta-values) were analysed between both conditions to focus more on each sensory modality. Nevertheless, the individual comparison did not show any statistically significant alteration of each parameter. The single bout of CT in this setting did not seem to be the right way to induce an EIH in a healthy population. But, why Habring et al. (2012) indicated that physical activity, accompanying with proprioceptive training may act as an effective activator for endogenous pain inhibition, assuming that a stimulation of proprioceptors, predominantly assigned to the group of Aß-afferents leads to an activation of inhibitory GABAergic interneurons in the spinal cord [15]. In addition, the transmission of impulses, relaying by peripheral receptors to the dorsal horn cells via C- and Aß-afferents, can be slowed or overlapped by any simultaneous sensory input in larger myelinated nerve fibres [16]. Another frequently described mechanism of EIH is the activation of the endogenous opioid system [17]. A stimulation of Aδ- and C-afferents can activate the endogenous opioid system and therefore lead to alterations in pain processing [18]. So, one reason for missing EIH after CT might be that the sensory input and the activation of the peripheral receptors were not high enough to generate the described mechanisms in the healthy system of the participants. Looking on aerobic and strengthening exercises Naugle et al. (2012) reported a dose-response relationship of training intensity and duration and its hypoalgesic effect. So, the activation of cardiovascular and metabolic metabolisms seem to be important factors to induce an EIH [19]. For aerobic exercises largest effect sizes could be observed for high intensity performances (75% of maximal oxygen uptake and a minimum duration of >10 min). For strengthening exercises the greatest hypoalgesic effects were reported when performing long-duration muscle contractions with low intensity (isometric 5–9 min). The authors concluded that during long-term muscle contractions, active motor units become fatigued and higher threshold motor units were progressively recruited to maintain the required force [2]. In the presented study each exercise was performed for a duration of 90 s (3 × 30 s with a rest period of 30 s) and we chose low intensity exercises to exclude side effects induced by aerobic and strengthening exercises because more intensity comes along with more work load and therefore more cardiovascular stress. In conclusion, training intensity might be too low to generate EIH. A limitation of this study was, that cardiovascular intensity as well as the influence on the metabolic system were not monitored during the CT. In addition, general physical activity level as well as coordinative abilities of the included participants were not documented in this study, so that the challenge of CT and the trainings intensity might had varied between the participants. The included participants represented an average group of young, healthy and moderately trained people, representative for normal population with. Furthermore, it is not known if baseline physical abilities are associated with alterations in pain perception, in general. In future studies, general physical activity level of the participants as well as trainings intensity should be monitored and adapted, whereby cardiovascular and metabolic effects due to the training should be controlled.

Another limitation of this study was, that results cannot be transferred on another population than healthy, young men. Even if the results indicate that a healthy somatosensory system could not be influenced by the CT of this study, pain perception in woman differ from pain perception in men, especially in younger age cohorts and moreover, pain perception is changing during course of life [9]. In further studies these aspects should be considered.

As mentioned, somatosensory function and therefore pain perception is also adapted in people with chronic pain states and the results can probably not be transferred from people with a physiological somatosensory system to a chronic pain population with pathological changes of the somatosensory system. Looking on aerobic as well as strengthening exercises, studies showed that EIH in this population is, depending on the type of chronic pain condition, highly variable and in contrast to a healthy population. In people with chronic pain states less training intensity is needed to generate an EIH than in people without pain [20]. What could mean that the low-intensity CT of the presented study could still lead to hypoalgesic effects in a chronic pain population. It should be the next step to analyse that.

Conclusion

In this setting, a single bout of CT did not lead to an EIH in young healthy men. Further research is needed, to examine the influence of a single bout of CT as well as long-term CT on pain sensitivity in people with different types of chronic pain states. In addition, the effects of CT combined with strengthening exercises and a higher cardiovascular input, similar to training interventions in clinical practice of rehabilitation should be examined. Detailed knowledge regarding the effects of different training interventions on pain modulation is needed to completely understand the mechanism of EIH.