Walking increases pain tolerance in humans: an experimental cross-over study

2.1 Participants

Pain-free subjects aged 18–50 years were recruited through advertisement at the University of Southern Denmark, local sports associations and social media from August to November 2017. Inclusion criteria were 18–50 years of age, pain-free during the last week, and Danish speaking. Exclusion criteria were neurological, cardiovascular or psychological disorders, pregnancy, acute or chronic pain conditions, use of any pain medication, and not being able to read or understand Danish. Furthermore, all subjects were asked to refrain from physical exercise on days of participation. Menstrual cycle, smoking, other medications than analgesics as well as exercise history was not considered in the exclusion criteria. The study was carried out according to the Declaration of Helsinki, approved by the ethical committee of Southern Denmark (S-20170102), and all subjects signed an informed written consent before participating.

Estimated number of subjects was based on a priori sample size calculation. To detect a significant difference in change in pressure pain tolerance after walking compared to rest with a medium effect size [5], a power of 0.80, and an alphaof ≤0.05, 34 participants were required in this repeated-measures within subject design.

2.2 Procedure

All subjects participated in two sessions at approximately the same time of the day with a 1-week interval between sessions. The purpose of the 1-week time interval between sessions was to avoid substantial changes in physical conditioning within subjects, and to reduce potential muscle soreness or fatigue from session 1 potentially influencing the pain sensitivity or response to walking in session 2. However, 4 subjects rescheduled their second appointment (unrelated to pain or illness) and completed the second session within 4 weeks of the first session.

At the beginning of each session a thorough introduction to the procedures both verbally and via demonstrations were given and all subjects were familiarized to the pressure pain sensitivity assessments which was used as EIH is more robust on pressure pain [5], [9]. Next, all participants performed a 6 min quiet rest condition and a standardized 6 min walking exercise in randomized and counterbalanced order in each of the two identical sessions, hence half of the subjects performed the rest condition first and half of the subjects performed the walking session first in each of the two sessions (Fig. 1). A 10 min “wash-out” period was applied between the two conditions to prevent potential hypoalgesic carry-over effects from walking or repeated assessments of pain sensitivity. This period was chosen as most studies on EIH have shown short-lasting effects of less than 10 min [2], [21]. Before and after both conditions pressure pain thresholds (PPTs) at the dominant thigh and non-dominant shoulder were assessed with handheld algometry and cuff pressure pain thresholds (cPPT) and cuff pain tolerance (cPTT) at the lower leg were assessed with computer-controlled cuff algometry (Fig. 1). During assessment of the pain sensitivity participants were seated with their arms resting in the lap and without footrest. Each session lasted approximately 60 min and all assessments were performed by the same assessor.

Fig. 1:

Illustration of the experimental procedure performed in both sessions. The order of rest and walking was randomized and counterbalanced between subjects. Pressure pain thresholds were assessed at the thigh and shoulder with handheld algometry (PPTs), and at the lower leg with computer-controlled pressure pain threshold (cPPT). Pain tolerance was assessed at the lower leg with computer-controlled cuff algometry (cPTT).

2.3 Walking (exercise condition) and Quiet Rest (control condition)

The 6-min walk test (6MWT) is a standardized walking protocol used for assessing functional aerobic capacity and is normally used for elderly people [22]. In this study the 6MWT was used as the exercise condition with the purpose of exposing the participants to a standardized walking exercise. The 6MWT was performed in accordance with the standardized protocol [23], and a 20 m course between two cones was used. Standard instructions were given prior to the walking condition in both testing sessions. Subjects were encouraged to walk as far as possible within 6 min and were told that they were not allowed to run. At the end of each minute the subjects were cheered with standardized phrases; “Keep up the good work” or “You are doing well”. After 6 min of walking the perceived exertion was assessed using Borg’s Rating of Perceived Exertion (RPE) scale [24]. All walking conditions were instructed and supervised by the same physiotherapist who are trained and experienced in instructing the 6MWT.

In the quiet rest condition, subjects were instructed to relax in a seated position in a comfortable armchair for 6 min in a quiet room keeping a temperature of approximately 21°C.

2.4 Assessment of pain sensitivity

2.5 Statistical analysis

Data in this study were reported as means and standard deviation (SD) in the text and as mean and standard error (SE) in figures unless otherwise specified. The distribution of pain sensitivity measures across sessions did not deviate significantly from normality which was assessed by visual inspection of residual plots.

3.1 Comparison of the effects of walking and quiet rest on pain thresholds and pain tolerance

3.1.2 Pain tolerance

A significant main effect of condition (i.e. walking vs. rest) was observed for absolute (Fig. 2; ANOVA: F(1,31)=6.01, p=0.02) and relative change (7.1±14.5% in session 1 and 10.2±11.8% in session 2) compared with rest (1.0±8.1% in session 1 and 5.6±11.2% in session 2 in cPTT, with post-hoc test showing a larger increase in cPTT after walking (mean difference across sessions 2.6 kPa [95% CI: 0.45–4.9 kPa]).

Fig. 2:

Absolute mean change (n=32)+standard error (SE) in pressure pain tolerance (cPTT) assessed with computer-controlled cuff algometry after 6 min quiet rest and 6 min walking in session 1 and session 2: Significant difference between conditions (*, p<0.05), and significant difference between sessions (†, p<0.05).

3.2 Exercise parameters between sessions

All subjects completed the walking conditions in both session 1 and session 2. The perceived intensity of walking assessed with rating of perceived exertion (RPE) and walking distance did not differ significantly between session 1 (distance: 632.5±75.2 meters, RPE: 10.9±1.9) and session 2 (distance: 642.1±80.2 meters, RPE: 11.0±2.4) (p>0.11). Moreover, RPE showed excellent relative reliability with an ICC value of 0.95 [95%CI: 0.90–0.97].

No significant correlations were found between EIH responses and walking distance or RPE in session 1 or session 2 (r=−0.31 to 0.32, p>0.07).

3.3 Test-retest reliability of PPTs, cPPT, cPTT and EIH responses

4.1 The effect of walking on pressure pain sensitivity

Walking induced a significant EIH response in pressure pain tolerance but not pressure pain thresholds when compared to quiet rest. The effect of walking on pressure pain tolerance has not previously been investigated, but the results are in line with a recent study demonstrating an increase in pressure pain tolerance and not pain threshold after a submaximal isometric exercise (muscle contraction without joint movement) [5]. This suggests that low intensity exercise has a specific effect on pain sensitivity above the pain threshold in healthy subjects, by increasing pain tolerance, but not significantly affecting pain thresholds. The increase in pain tolerance is also in line with previous studies showing an increase in pressure pain tolerance after submaximal intensity aerobic exercise [28] and mixed types of excises at moderate intensity [29]. However, the results are in contrast to the results from a study investigating the effect of the 6-min walk test on pressure pain thresholds previously presented in an abstract [30]. The lack of a control condition in the previous study may explain the contrasting results. Furthermore, pressure pain threshold was only assessed with handheld pressure algometry, whereas the current study also investigated the pressure pain threshold using cuff algometry. Computer-controlled cuff algometry provides pressure to a larger area compared to handheld pressure algometry, and therefore is less affected by local variations in pain sensitivity [31]. The automated cuff algometry is moreover independent of the assessor, which reduces the possibility of measurement error.

The lack of increase in pressure pain thresholds observed in this current study after walking when compared to quiet rest, is also in contrast to previous studies demonstrating a systematic increase in pressure pain thresholds after high intensity aerobic exercises [2], [13], [32]. However, it is in agreement with other studies investigating submaximal intensity isometric exercise [5], and mixed types of exercises at moderate intensity [29]. Furthermore, it has been demonstrated that there is a dose-response effect between exercise intensity and pain threshold increase in healthy young adults [2], [32]. Consequently, the chosen low exercise intensity may explain the equivocal results.

Still, the mechanisms underlying EIH are not entirely clear. Release of endogenous opioids has been hypothesized as a primary mechanism responsible for widespread EIH, however walking does not result in large increases in endogenous opioids [33]. The “pain inhibits pain” phenomenon or the CPM has also been suggested as a potential mechanism responsible for the widespread hypoalgesic response after exercise. Unfortunately, pain due to walking was not assessed in this study, but based on the exercise intensity and the RPE rating it is unlikely that the 6MWT induced significant pain in young pain-free subjects. Exercise causes changes in the cardiovascular response, which have been suggested as a possible mechanism due to baroreceptor activation [34]. The concentration of several opioid and non-opioid substances that can influence nociceptor sensitivity is increased in the blood during exercise. Jones and colleagues recently showed that blocking the blood flow to a non-exercising body site reduced the EIH response in the body site where blood flow was blocked compared with the EIH response in non-blocked body sites after 5 min intense bicycling [35], suggesting that systemic peripheral factors may contribute to the EIH effects seen in this study. Although still speculative, EIH may also be related to distraction through attention towards other somatic sensations (e.g. increased breathing, heart rate and sweating) induced by fast walking, however this needs further investigation.

4.2 Test-retest reliability

Relative test-retest reliability of pressure pain sensitivity assessed with both handheld pressure algometry and computer-controlled cuff algometry demonstrated excellent reliability when repeated within the same session, as well as good to excellent reliability when repeated across sessions separated by a week. This is in line with previous studies reporting ICCs over 0.7 in both healthy subjects [5], [13], [36] and patients with chronic pain [8]. However, a systematic increase was observed in PPTs and cPTT within session 2, as well as in cPPT in session 1. These findings are in contrast with previous studies showing no within-session systematic change for handheld pressure algometry [13] and computer-controlled cuff algometry [5]. Lack of randomization of the condition order in previous studies and difference in the duration of the resting condition may explain the contrasting results. In the previous studies showing no within-session systematic difference, the duration of the resting condition was 15 min. The contrast in findings between resting for 15 min and resting for 6 min may suggest a lack of time to neutralize the habituation effect of being exposed to repeated pain sensitivity tests in this current study. Nevertheless, the excellent ICCs after rest suggests that all included subjects change approximately the same amount due to the repeated pain tests. Furthermore, no difference was seen in absolute change of pain sensitivity after rest between EIH responders and non-responders. This indicates that the potential hypoalgesic effect of repeated pain testing was equal within the participants between conditions in both sessions and randomization compensated for potential bias of conditioning order.

Although a significant increase in pain tolerance was produced after walking in both sessions, the relative test-retest reliability for the hypoalgesic response only demonstrated a fair to good reliability (ICC=0.61). Additionally, the agreement in classification of EIH responders and non-responders between the two sessions was low. These results suggest that despite a significant increase in pain tolerance after 6 min walking on a group level, there is a substantial difference in the hypoalgesic response within subjects between two sessions separated by 1 week. Unknown physical or psychological factors, which was not accounted for in this study may explain the individual variations. Moreover, a possible training effect can have contributed, as the test persons were familiar with the tests and walking condition in session 2. Interestingly, a systematic difference in the increase in pain tolerance between sessions was present, with a significantly larger effect of walking in session 2 compared to session 1. However, about the same number of subjects were classified as EIH responders in session 2 (16 subjects) compared with session 1 (15 subjects). In addition, there was no difference in ratings of perceived effect (RPE) or walking distance, which indicates no difference in exercise intensity on a group level in session 2, as well as no significant correlation between walking distance, RPE and the hypoalgesic response.

The relative test-retest reliability of the hypoalgesic response after walking in this current study shows comparable ICCs to what has been demonstrated previously in studies investigating the test-retest reliability of the EIH responses after high intensity aerobic exercise [13], moderate aerobic exercise [37] and isometric wall squat [14]. Moreover, the relative reliability of EIH across different sessions is comparable to the reliability of the CPM response [38] that may be partly responsible for the EIH response [2], [16], [17]. It has been suggested that a high focus on standardizations of test protocols and action to prevent assessor bias would increase reliability in test-retest investigations [39]. These factors should therefore be of high priority in future EIH reliability studies.

4.3 Limitations

This study has limitations. First, the assessor was not blinded to condition when testing. Second, the use of only two testing sessions as a plateau of the EIH response (when assessed with cPTT) was not reached. This was underlined by a significantly larger EIH response in session 2 compared to session 1. Assessment of the EIH responses in several sessions might have increased the reliability of the EIH response. Third, computer-controlled cuff algometry is not easily accessible in all settings and the results may thus differ from studies using other non-automated devices. Fourth, pressure pain tolerance was only assessed at the calf muscle and pain tolerance in other body parts may differ. Lastly, the study population consisted of healthy young adults, and it is uncertain whether comparable results would be found in different groups of individuals, i.e. elderly people or chronic pain patients. Future studies should investigate whether walking induces similar effects in subjects with chronic pain.