The influence of isometric exercise on endogenous pain modulation: comparing exercise-induced hypoalgesia and offset analgesia in young, active adults

2.1 Participants

Participants were recruited via word of mouth from a university campus on the Gold Coast in Queensland, Australia, over a 1-month period between August and September 2016. Written informed consent was obtained, and participants screened with the Physical Activity Readiness Questionnaire (PAR-Q) [34] to ensure participant safety during the exercise component of the testing. Participants were eligible to participate if pain-free and aged over 18 years with a level of English language sufficient to provide informed consent and complete a questionnaire regarding their physical activity behaviours. Exclusion criteria included the following: history of current or chronic pain condition; current use of prescription medications for pain, depression or elevated blood pressure; diagnosis of a neurological disorder (e.g. multiple sclerosis), inflammatory condition (e.g. rheumatoid arthritis), cardiovascular (e.g. hypertension) or metabolic disorder (e.g. diabetes mellitus); psychopathology (e.g. depression); pregnancy; unwillingness or inability to perform the prescribed isometric exercise, or inability to reliably report pain ratings.

2.2 Procedure

A flow diagram of the study participation procedure is outlined in Fig. 1. Demographic information and questionnaires were completed prior to the recording of baseline measures (heart rate, blood pressure, height, weight and limb dominance). Baseline PPTs and suprathreshold responses to pressure pain (PP₅₀) were recorded, followed by baseline suprathreshold response to heat pain (HP₅₀) and OffA assessment. Although OffA effects have been shown to completely dissipate after 20 s [15], [21], 5 min rest period was provided following OffA and prior to performance of exercises to prevent possible carry-over effects.

Fig. 1:

Flow diagram of study participation procedure.

The isometric knee extension exercise was then performed. Heart rate, blood pressure and RPE were monitored throughout the exercise. Immediately following the exercise, PPT and PP₅₀ were assessed, followed by OffA. After a 15 min resting period, PPT, PP₅₀ and OffA were re-measured. One investigator (SH) was responsible for measurements performed. The investigator was blinded to study aims and PPT outcomes, with another investigator (SF) responsible for the recordings, who was also blind to study aims.

2.3 Outcome measures

2.4 Data analysis

Normality of the variables was tested with the Shapiro-Wilk test in conjunction with visual inspection of histograms and box plots. Appropriate descriptive statistics were calculated. Friedman’s test was used to test differences between the non-normally distributed repeated heat pain threshold (HPT) measures over time (baseline, exercise, rest). HPTs were collected as part of the determination of OffA and served as a manipulation check to determine if sensitisation was occurring over time as a result of study methodology. Repeated measures (RM) ANOVA investigated the effect of time (0 s, 120 s, 240 s) during exercise on normally distributed blood pressure, heart rate and RPE (0 s, 120 s and 300 s) levels. The effect of sex was measured using independent t-tests.

For the primary aim investigating the relationship of EIH and OffA, Pearson correlation analysis was used to investigate the magnitude of percentage change in pain relief during the OffA protocol performed following exercise, against the percentage change in pain relief measured using PP₅₀ (post-exercise minus pre-exercise). Correlation analyses for differences in pain relief measured via the OffA protocol and difference in PPTs (post-exercise minus pre-exercise) were also calculated. The analyses were performed for both the local (tibialis anterior) and remote (cervical spine) areas of interest.

For the secondary aim, repeated measures ANCOVA (RM ANCOVA) were performed to investigate the effect of time and bodily region on the outcomes of PPT or PP₅₀ during and following exercise. Time (baseline, post-exercise and post-rest), region (cervical spine or tibialis anterior) and the interactive effects of time * region were included in the model. RM ANCOVA was also used to investigate the effect of time (baseline, post-exercise and post-rest) on the magnitude of OffA effect. Sex was included as a covariate in all models. Homogeneity of variance was evaluated via Mauchly’s test of sphericity. When significant, Greenhouse-Geisser estimates were used. Pairwise comparisons with Bonferroni corrections were performed to investigate significant differences. Paired t-tests were also conducted to determine whether participants exhibited OffA (change in pain rating from T1 to T2 vs. change in pain rating from T2 to T3). Effect sizes were calculated by subtracting the measures (PPT or PP₅₀) collected at baseline from those collected post-exercise.

Data were analysed with SPSS Statistics, Version 23 (IBM, USA). The significance level for all tests was set at p<0.05.

3.1 Participants

Thirty-six healthy, pain-free participants [mean (±SD) age: 23.6 (±6.6) years; 17 females] volunteered to participate, and completed the study. Males demonstrated significantly higher BMI [t₃₄=−2.35, p=0.025; males=24.0 kg/m² (±2.8), females=21.8 (±2.8) kg/m²]. Seventy-eight percent of participants presented with a “high” physical activity level (>3,000 MET min/week).

3.2 Exercise performance

At baseline, males presented with higher systolic blood pressure (t₃₄=−3.36, p=0.002), with a mean systolic blood pressure of 132 mmHg (±12) compared to 118 mmHg (±14) for females. There were no significant sex differences in baseline resting heart rate or diastolic blood pressure (p>0.27 for both). One way RM ANOVA demonstrated significant increases in heart rate, systolic, and diastolic blood pressure over time (0 s, 120 s, 240 s) during performance of the exercise task (Table 1). Post hoc pairwise comparisons with Bonferroni correction indicated that systolic and diastolic blood pressures were greater at 120 s when compared to baseline (p≤0.02 for both), whilst heart rate and diastolic blood pressure were significantly greater at 240 s when compared to baseline (p≤0.02 for both). Perceived exertion significantly increased over time (Table 1).

3.3 Manipulation check

There were no significant differences in HPTs over time (baseline, exercise, rest) (χ²=2.97, p=0.23) indicating that individuals were not becoming sensitised over time.

3.4 Exercise-induced hypoalgesia

3.5 Offset analgesia

OffA results are demonstrated in Table 3. OffA was elicited at each time point as demonstrated by significant paired t-tests at baseline (t₃₅=14.8, p<0.001); post-exercise (t₃₅=14.4, p<0.001), and post-rest (t₃₅=14.5, p<0.001), indicating that the magnitude of pain relief following the 1 °C reduction in temperature significantly exceeded the increase in pain reported during the initial 1 °C temperature increase. The magnitude of OffA did not significantly differ over time (F_2,68=2.83, p=0.066; partial ᶇ²=0.077). Sex did not interact with time to effect the magnitude of OffA (F_2,68=1.09, p=0.034; partial ᶇ²=0.031).

3.6 EIH and OffA correlation

There was no correlation between magnitude of percentage change of OffA following exercise and any measure of EIH (tibialis anterior PP₅₀: r=0.28, p=0.11; cervical spine PP₅₀: r=0.054, p=0.76; tibialis anterior PPT: r=0.003, p=0.99; cervical spine PPT: r=−0.03, p=0.88).

4.1 Correlation between EIH and OffA

EIH and OffA have both been reported to demonstrate temporal filtering of nociception [1], [17], [21] and underpinned by non-opioid mechanisms [8], [9], [18], [19], [20], [21]; indicating that these phenomena may be related. However, our results did not demonstrate the presence of such a relationship. We do not believe this is due to methodological concerns, as both EIH and OffA were appropriately induced in the majority of participants and we were careful to anchor the respective EIH and OffA measurements to individualised 50% pain ratings. However, the measurement of EIH and OffA did involve different test stimuli over different body regions that are associated with different peripheral receptors and central processes. We used PPT as a test measure for EIH, whilst OffA measures changes in HPTs. Factor analysis of responses to thermal and pressure pain modalities has demonstrated that response to each testing modality is distinct [44], and despite our attempt to individualise both mechanical and thermal thresholds to numerical pain ratings of 50%, this difference in peripheral receptor stimulation may have resulted in activation of different neural pathways [45], [46], and thus may also explain the lack of correlation between OffA and EIH. Future studies may wish to employ a thermal paradigm for testing EIH to investigate this further.

The other possibility is that these data may suggest that different mechanisms underpin these two endogenous analgesia paradigms. Recent fMRI studies have possibly demonstrated different neurophysiological brain processes underpinning these two phenomena. OffA has been associated with reduced activity of the primary somatosensory cortex [1], [47], [48] and increased activation of the periaqueductal grey (PAG), medulla, insula, dorsolateral prefrontal cortex, pons and cerebellum [1], [47]. Similar brain findings have been demonstrated with walking exercise, but more strenuous exercise (e.g. running) has been associated with reduced activity in the PAG, pregenual anterior cingulate cortex and the middle insula [49]. Given that the exercise performed by participants in the present study resulted in high levels of exertion, it is plausible that participants may have engaged brain processes similar to those associated with running [49]; which differ from those associated with OffA (although it should be noted that participants performed 2 h of running in the study by Scheef et al., compared to 5 min of isometric exercise performed in our study). The difference in neuroimaging findings between walking and running was associated with opiodergic mechanisms expressed in the running exercise [49], which differs from the non-opiodergic mechanisms thought to underpin OffA [20], providing another plausible explanation for the lack of correlation in findings between the two phenomena. Future research could investigate the mechanisms associated with different exercise intensities and OffA. The lack of relationship between EIH and OffA in our study would suggest that future research investigating endogenous analgesia efficacy, such as in aged or patient populations, should include testing of both paradigms.

4.2 OffA and the effects of exercise

Our study showed that OffA processes were effectively induced, with the magnitude of response similar to those recently reported in a systematic review [50]. The effect was also robust over time, with no modulation by exercise demonstrated. Exercise has previously been shown to modulate central processes, as demonstrated by improvements in resting state functional connectivity measured via fMRI in individuals with fibromyalgia following 3 months of strength training [51]; whilst changes following acute bouts of isometric exercise have also been observed using paradigms such as temporal summation [35], [52]. However, until now, the effects of exercise have not been explored in OffA. It is currently unclear what modulates OffA. There is no evidence available to suggest that OffA can be modulated pharmacologically, either through opioid or NMDA-receptor involvement [8], [9], [18], [19], [20], [21]. It has also been demonstrated that no significant alteration in OffA magnitude results from experimentally induced sensitisation [53]. It has been suggested that OffA may be associated with peripheral mechanisms [26], which our study may indirectly support. As our participants were healthy and did not present with pain, our data may provide some support to this conclusion, given that exercise did not modulate OffA. Studies involving clinical populations, especially those with opportunities to investigate individuals prior to and following nociceptive modulation (e.g. surgery) may shed further light on the mechanisms underlying OffA.

4.3 EIH

Although 95% of individuals reported 6–27% greater tibialis anterior PPTs following exercise, EIH was only present locally (at the site of exercise) and at slightly lower magnitudes than expected for this exercise duration and intensity. Suprathreshold pain ratings to pressure stimuli also only improved over tibialis anterior. These results are in contrast with the findings of a recent systematic review [17], which reported that most studies in healthy individuals demonstrated EIH systemically; albeit to a lesser extent than that observed locally. Large effect sizes have previously been observed for long duration (>5 min) isometric contractions [14], [17]. However, despite replicating this protocol, our effect sizes were slightly smaller (moderate to large), and participants did not demonstrate EIH remotely in the cervical spine region, contrasting from our previous study results whereby a 3 min isometric wall squat induced cervical spine hypoalgesia [54]. This may be a result of sex bias. Lannersten and Kosek’s [14] results were demonstrated in a female-only study, which differed from the mixed sex sample we investigated. The stronger EIH demonstrated by females [23], [24] may explain the disparate findings, although our study failed to demonstrate this sex relationship. It has been observed that large EIH effect size drop offs are observed at high contraction intensities [17], [41], which may have occurred in our study, given that individuals’ RPEs averaged 8/10, corresponding to “very severe” levels of exertion. As a result, the 5 min isometric knee extension exercise at 25% MVC, although initially a low intensity exercise prescription, may not have been the ideal condition to elicit EIH. Our recent study results, whereby a 3 min isometric wall squat resulted in cervical spine hypoalgesia may support this, given that median RPE’s of 6/10 were reported [54]. It is also possible that the increased pain sensitivity demonstrated in the cervical spine region over time may have resulted from increased sensitisation to the repeat mechanical stimuli applied, although our manipulation check indicated that sensitisation did not occur for HPTs.

The difference in EIH magnitude we observed could also be attributed to the sample. The group was young, fit and healthy and did not present a broad cross section of the community, as indicated by the high physical activity levels. It might be that the 5 min of isometric contraction in these highly active individuals, albeit resulting in high levels of exertion, was not sufficient to elicit processes associated with greater magnitudes of EIH, as has also been demonstrated in running exercise when compared to walking [49]. Additionally, higher levels of physical activity have been demonstrated to improve pain tolerance [40], [55], which is consistent with our data, whereby many participants reported PPTs close to and even reaching the ceiling values. Future research is warranted to determine the optimum protocol for eliciting EIH in individuals with high levels of physical activity. In summary, although the majority of participants demonstrated large effect sizes following an isometric exercise protocol, reducing the duration of acute exercise to 3 min may be more successful in inducing systemic hypoalgesia. Research testing procedures should also limit the number of repeat mechanical stimuli applied.

4.4 Limitations and future research directions

The main limitation of this study involved the lack of a rest period prior to performance of exercise to determine the efficacy of the EIH testing paradigm. Future studies would benefit from inclusion of a rest period, and this is currently underway. Another limitation involved the participation of high numbers of highly active participants, and likely associated excellent general health, which may not be representative of the general nor chronic pain communities. Further comparison with sedentary and/or patient groups may demonstrate a different relationship. Other exercise types, durations and intensities have been demonstrated to induce EIH [17], such as aerobic exercise [56], [57] and dynamic resistance exercise [58], [59] and may be more closely related to OffA and as such, this requires further investigation.