Examining the role of mood in pain‐limited treadmill walking duration in young healthy individuals

The purpose of the study was to examine the effects of acute mood modulation on treadmill walking duration during experimental pain application.


| INTRODUCTION
Regular physical activity is a highly promoted management approach for improving pain and disability outcomes in acute and chronic pain conditions (Booth et al., 2017;National Clinical Guideline Centre UK, 2014).Moreover, regular physical activity has been shown to slow the degeneration in musculoskeletal, cognitive and cardiorespiratory health in people with chronic pain (Ambrose & Golightly, 2015;Geneen et al., 2017).Besides from its integral role in pain management, active engagement in regular physical activity offers acute improvements in mood, enhanced self-image as well as protection against mental health disorders (Ambrose & Golightly, 2015).
While exercise-based rehabilitation is recommended by health authorities and is associated with well-recognized health benefits, pain and pain-related cognitive and behavioural adaptations limit exercise tolerance and adherence to physical activity programmes, jeopardizing participation and rehabilitation success (Geisser et al., 2003).Accompanying these pain-related cognitive adaptations are changes in mood, with a well-defined reciprocal relationship between the two (Kamping et al., 2013;Ossipov et al., 2014).For instance, psychological distress stemming from persistent pain has been shown to increase the susceptibility of an individual to be diagnosed with a mood disorder (Roditi & Robinson, 2011).Conversely, mood disorders, such as anxiety and depression, have been shown to further enhance the intensity of the pain experience, impair function and prolong the course of the illness (Ambrose & Golightly, 2015;Berna et al., 2010).In an experimental setting, the relationship between mood and pain has been tested via the Motivation Priming Hypothesis (Lang et al., 1997), which predicts that negative emotional priming leads to greater expression of pain compared with positive emotional priming (Finan & Garland, 2015;Torta et al., 2017;Veilleux et al., 2019).This supports the idea that pain networks can be modulated by emotion and attention (Kamping et al., 2013;Meagher et al., 2001).
While acute changes in mood appear to alter pain perception, the subsequent relationship with exercise performance remains unexplored.For many, exercise performance is limited by pain (Friedrich et al., 1998;Jack et al., 2010), and it remains to be seen if changing mood can alter both pain perception and walking duration.The current study chose to use treadmill walking as the mode of exercise as it is commonly used in a range of exercisebased rehabilitation programmes.Hence, the aim of the current study was to investigate whether modulating an individual's mood can alter pain-limited treadmill walking duration.When given a fixed noxious stimulus, we hypothesized that viewing low valence (unpleasant) images would result in a more rapid increase in pain scores and reduced treadmill walking duration when compared to viewing higher valence (pleasant) images.

| Design
This was a repeated measure, within-subject study design.Participants completed four sessions in a randomized order approximately 1 week apart, which included one familiarization session and three experimental sessions.

| Participants
Thirty healthy participants between the ages of 18-30 years (14 females) were consecutively recruited from the general and university communities between December 2020 and April 2021.Griffith University's Human Research Ethics Committee (GU Ref No: 2020/586) approved the study and participants provided written informed consent prior to enrolment.Participants were healthy and undertook average to moderate levels of activity.Individuals who were currently prescribed medication and/or being treated for depression or anxiety, and individuals with a current health condition that may affect exercise capacity or pain tolerance were excluded from the study.Individuals with a lower limb musculoskeletal injury which may affect the ability to walk on the treadmill and those prescribed regular pain medication were excluded.and adherence in walking-based rehabilitation has shown to be jeopardized by pain and pain-related cognitive and behavioural adaptations.This study examined the effect of a shift in mood on pain perception and treadmill walking tolerance.We found that with a worse mood, individuals were less tolerant of pain and walked on the treadmill for a shorter duration.These results suggest that factors which improve mood should be combined with walking-based training to improve tolerance.

| Procedure
Participants attended four sessions approximately 1 week apart.The first (familiarization) session involved collection of demographic and baseline outcome measures, optimizing the treadmill walking speed and the target cuff pressure used to induce the noxious pain stimulus (see below).Participants were also familiarized with the pain and mood rating scales and the International Affective Picture System (IAPS).
Participants then completed three experimental sessions, approximately 1 week apart, in which they walked at the same speed each session on the treadmill, with the pre-determined, target cuff pressure (i.e.noxious stimulus) while viewing either positive, negative or neutral valence images delivered via the IAPS.A single mood change (negative, neutral, positive valence) was induced during each experimental session and the valence order randomized for each participant using a computer-generated randomization sequence.Participants were blind to the order of the sessions (positive, negative and neutral).During the three experimental visits the participant performed two walking tests while viewing the same mood condition images.There was a break between each of the walking tests of at least 30 min.
During each treadmill walking trial, the participant listened to a pre-recorded audio reminding them to focus on the images and the extent to which the images influenced their mood, as well as prompting to rate any changes in their pain.

| Treadmill walking and noxious stimulus
Participants were asked to walk on a treadmill (zero grade) at a self-selected brisk pace.The walking speed was recorded and then used in the experimental sessions.Ischaemic pain was induced during treadmill walking using a 30 cm pressure cuff (E20 Rapid Cuff Inflator, D.E.Hokanson, Inc and Hokanson AG101 Cuff Inflator Air Source, D.E.Hokanson, Inc), which was placed around the upper thigh of right leg.The pressure used to induce the noxious stimulus, the target cuff pressure, was determined during the familiarization session.During this session, the pressure cuff was inflated to a baseline of 30 mm Hg and the participant asked to start walking on the treadmill at the pre-determined speed.After 2 min of walking the participant was asked to report their ischaemic pain severity on a Numerical Rating Scale (NRS) from 0 to 10 (NRS, 0 = no pain, 10 = worst pain imaginable).The cuff pressure was then incrementally increased every 30 s, and pain severity was recorded prior to each incremental increase.Once the participant reported pain levels of 2-3/10, the pressure was recorded, and this target cuff pressure was then used in subsequent experimental sessions.
In order to minimize the risk of adverse events associated with inducing ischaemic-type pain through the pressure cuff, while still achieving our aim of quantifying the outcomes of pain and treadmill walking duration, we chose a target pain severity of 7/10 and a walking duration limit of 10 min.To ensure that participants reached the pain severity target of 7/10 within the 10-min time limit, a test trial was conducted during the familiarization (i.e., first) session.Participants were asked to walk on the treadmill at the pre-determined speed, while viewing neutral images on the IAPS.The pressure cuff was incrementally inflated to the pre-determined target cuff pressure over the first 4 min of walking (Figure 1).Participants were asked to continuously report their level of pain on the NRS and to continue walking for a maximum of 10 min.Once the reported pain intensity reached 7/10, the familiarization trial ceased.If a minimum pain intensity of 7/10 was not achieved within 10 min of walking, then either the walking speed or the cuff inflation pressure were adjusted, and the trial was repeated until a 7/10 pain score was achieved within the 10 min.Once achieved, the individualized speed and target cuff pressure for each participant were recorded and then used for each subsequent experimental session.Participants were told that the walking trial would last for up to 10 min, but they were blinded to the F I G U R E 1 Timeline representation of the experimental protocol for the treadmill walking trial.

min 4 min 10 min
Pressure cuff inflaƟon begins to increase from 30 mm Hg in set increments Pressure cuff has reached the target inflaƟon pressure

START
Test concludes when a 7/10 is reported on the NRS or at 10 min ParƟcipant begins walking at set speed and with pressure cuff inflated at 30 mm Hg pain intensity cut-off of 7/10.A schematic representation of the experimental protocol for the treadmill walking trial can be seen in Figure 1.

| Mood intervention
The International Affective Picture System (IAPS) (Lang, 2005) is a series of images, designed to invoke a mood change (negative, neutral or positive valence).Note participants were not shown any erotic images to minimize the potential confounding effects of arousal (de Wied & Verbaten, 2001).The IAPS has demonstrated validity and reliability in inducing mood change in healthy, cross-cultural, and clinical populations (Huang & Chiang, 2014;Lang et al., 1997).The images were displayed on a monitor screen set at eye level.Each image was displayed for 6 s followed by a 4-s interval (no image) before progressing to the next image.

| Measured variables
Participant demographic characteristics (gender, age, height and body mass) were collected at the time of enrolment into the study.Pre-intervention outcome measures of mood (potential effect modifier) and pressure pain threshold (potential confounding variable) were collected at the beginning of each session.Mood was evaluated using the Brunel Mood Score (BRUMS, Zhang et al., 2014) and pain pressure threshold (PPT) was measured using a digital pressure algometer (Somedic AB, Sweden), which has been shown to be a reliable (Bisset et al., 2015).PPT was measured using a 1 cm 2 probe tip over the tibialis anterior muscle (i.e.superior anterolateral border of tibia) of the participant's dominant leg.
Pressure was applied at 40 kPa/sec and participants were asked to press a handheld switch the moment they perceived the first onset of pain.PPT was tested three times at 30-s intervals on the tibialis anterior muscle and the average was then used in further analyses.The primary outcome was the duration of treadmill walking which was defined as the time taken (in seconds) to reach a participant pre-defined maximal pain endpoint of 7. The time to reach pain scores of 2, 3, 4, 5, 6 were also recorded during treadmill walking.The following dependent variables were also measured during the treadmill walking trial: (1) Mood: measured digitally every minute using the Self-Assessment Manikin (SAM) affective rating system, the 10-point pictorial rating scale (Lang, 1980), a previously validated measure of the sense of pleasure; (2) Pain: Measured continuously using the NRS on a digital sliding scale (Spike, Cambridge electronic Design).Mood and pain numerical values were recorded continuously using a custom-written data collection software (Labview, National Instruments Corporations).

| DATA ANALYSIS
Baseline demographic characteristics were reported as frequencies (categorical) and mean ± standard deviation.Pre-intervention measures (PPT and BRUMS) and outcome measures of mood, pain and treadmill walking duration (time taken to reach the end point) were compared between the three mood conditions (negative, neutral and positive) using a one-way analysis of variance (ANOVA).Additionally, individual NRS scores during each mood condition were analysed using a two-way ANOVA, with the mood manipulation conditions and time as factors.Order effect was checked by comparing the mean time taken to reach a pain score of 7 per experimental session using a one-way ANOVA.When the primary effect was significant, Bonferroni post hoc analyses were conducted to investigate pairwise comparisons and further uncover differences between the mood conditions.Effect sizes (Cohen's d) were calculated where significant differences in mood conditions were detected (Lakens, 2013).Data were presented as mean ± standard deviation.Missing data were accounted for using the Last Observation Carried Forward method (Shao & Zhong, 2003).All statistical analyses were carried out using a standard statistical software package (IBM.SPSS statistics 26.0); a p-value of 0.05 was used to establish statistical significance.Note a convenience sample size of 30 participants was chosen to participate in the study.Given the novel experimental design it was not possible to undertake a priori calculations of the required sample size required to detect significant (α < 0.05) differences in the primary outcome (treadmill walking duration) between mood conditions with fixed power (1-β).Note that while it was not possible to estimate the optimal sample size prior to the study, the effect sizes were calculated for the time taken to reach a score of 7 between mood conditions.Based on these effect sizes and the sample size of 30, the estimated power (G*Power, 3.1.9.4) of the study to detect the differences (α < 0.05) of positive to negative conditions; power > 99%; positive to neutral conditions: power > 99%; neutral to negative conditions: power = 71%.

| Participant characteristics
Thirty participants (Males: Females 16:14; age 22.9 ± 2.5 years; height 170.9 ± 9.5 cm; body mass 68.4 ± 14.6 kg) completed the familiarization and the three experimental sessions for this study.Three participants withdrew from the study due to reasons unrelated to the study (time commitment and a sustained injury from personal extracurricular activities).

| Outcome variable: Mood during treadmill walking
During the walking tests, there was a significant (p < 0.001) difference in mean SAM scores between mood conditions (Figure 2).Post hoc analysis demonstrated significant differences between all conditions (p < 0.001), negative (2.4 ± 0.3), neutral (4.9 ± 0.6) and positive (6.6 ± 0.3).These results closely aligned with the IAPS allocated group mean valence scores for each mood domain, negative (2.7 ± 1.6), neutral (5.5 ± 1.4) and positive (7.3 ± 1.5) (Lang, 2005).From ~300 s, the number of participants continuing to walk began to reduce, due to having reached a 7/10 NRS score.To prevent the data from being visually skewed when displayed in a graph due to dropouts, the data after ~300 s were removed and the average mood score at the end of treadmill walking (7/10 NRS) was illustrated.During the walking tests, there was a significant (p < 0.001) difference in minute-by-minute changes in mood.

| Primary outcome: Treadmill walking duration for different mood interventions
Figure 3 represents the average treadmill walking duration taken to reach a pain score from 1 to 7 for all mood conditions.Three participants did not reach the intended 7/10 NRS at the end of the walking test (10 min).One participant did not achieve a 7/10 NRS in all three mood conditions, only achieving a maximum of 5/10 NRS.The other two participants achieved a 7/10 score for the negative condition; however, they did not achieve a 7/10 on the positive condition (2 participants) or the neutral condition (1 participant).These missing data were accounted for using the Last Observation Carried Forward method (Shao & Zhong, 2003).There was no significant difference in time taken to achieve a pain score of 1 and 2 between the three mood conditions.However, there was a significant difference (p < 0.05) between mood conditions in the times to achieve pain scores of 3, 4, 5, 6 and 7/10 NRS.Post hoc analyses indicated that the significant effect was only observed between positive and negative conditions indicating that F I G U R E 2 Average minute-byminute self-assessment manikin scores for each condition.From 318 s, incomplete data due to participants stopping exercise having reached a pain score of 7/10.End of exercise is the average valence score for all participants at the end of their exercise test.The number of participants walking at 400 s (negative: 20; positive: 26; neutral: 22); 500 s (negative: 7; positive: 20; neutral: 15); 600 s (negative: 2; positive: 4; neutral: 2).

Mood valence scores
Walking time (s)

| DISCUSSION AND CONCLUSION
This study found that while walking with a fixed nociceptive input, participants who viewed negative imagery reported a decreased mood (sad/angry) compared with the neutral and positive imagery, whereas when viewing positive imagery participants reported an increased mood (happy) compared with neutral and negative imagery.
Under the negative mood condition, participants were less pain tolerant and walked for a shorter duration on the treadmill before experiencing a 3, 4, 5, 6 and 7/10 Pain NRS when compared to a positive mood shift.Neither the negative nor the positive mood condition was significantly different to the neutral condition with regard to the time taken to reach each pain score and duration of walking.The current study found that, when using a fixed nociceptive stimulus, the walking duration to achieve a target pain intensity during walking can be altered with acute changes in mood in healthy individuals.These results are consistent with others who found that participants who were positively primed reported reduced pain perception and an increased tolerability to a normally pain-provoking task compared with negatively primed participants (Arnold et al., 2007;de Wied & Verbaten, 2001).Our results would appear consistent with the motivational priming hypothesis (Lang et al., 1997) and supports the claim that motivational priming is a reasonable theoretical construct for pain perception in humans (Meagher et al., 2001;Tang et al., 2008).
The mechanism by which mood influences the pain experience appears to involve the central pain modulatory network, including the anterior cingulate cortex, bilateral insula and somatosensory cortices (Kamping et al., 2013).Furthermore, the biopsychosocial model supports the interaction between neurobiological and cognitive, emotional, behavioural and social factors in the development and maintenance of the pain experience.While the cause/ effect nexus is not yet clearly established, the health benefits of a positive affect (i.e.feeling state) that is characterized by pleasant moods or emotions (Fredrickson et al., 2008) are well established and include improved function in the neuroendocrine and immune systems (Segerstrom & Sephton, 2010;Steptoe et al., 2009) and the development of self-efficacy to help cope with pain (Finan & Garland, 2015).
Interestingly, we found no difference in the walking duration to achieve a pain score between the positive and neutral and between the negative and neutral mood states.While we anticipated greater separation between the neutral and negative/positive conditions, the lack of significant difference may be a function of the sample size.
F I G U R E 3 All data presented as mean ± SD for all participants.*p < 0.01, time taken to achieve pain score significantly shorter for negative images versus positive images.
Others have similar results in healthy individuals et al., 2007;de Wied & Verbaten, 2001).In their study, de Wied and Verbaten (2001) proposed that differences in pain tolerance related to a negative mood state may be more related to the type of negative images used (i.e.pain-related images) to induce the mood state.The collection of negative images from the IAPS database includes the option to use images of humans exhibiting pain-related expressions and injuries, for example mutilated bodies, infected open wounds and burn victims (i.e.pain inclusion images).It is plausible that these graphic images could emotionally evoke pain-related memories and traumas, which contribute to the construct of the current pain experience (Chapman & Nakamura, 1999;Veilleux et al., 2019).de Wied and Verbaten (2001) found that participants viewing negative images associated with pain-related cues (pain inclusion group) had lower tolerance to a painful task compared with those viewing negative images without pain-related cues (pain exclusion group) (de Wied & Verbaten, 2001).The negative images employed in the current study had fewer pain-related cues similar to the pain exclusion group in De Wied and Verbaten's study, which may explain the lack of difference in perceived pain intensity and treadmill walking duration between negative and neutral interventions.
Similarly, a positive mood state may be related to the type of positive images used, with erotic images known to induce a higher level of arousal and valence in males than in females (de Wied & Verbaten, 2001).We reduced the likelihood of sex differences in arousal confounding the effects of valence by removing erotic images from the groups of images in this study, as we were not powered to assess males and females separately.
The significant difference in perceived pain between negative and positive mood states may also be related to aspects of the stimuli and the specificities in human interpretation.Torta et al. (2017) also supports this rationale by elaborating on the influence of attentional factors, such fear, to be a stimulant for increasing pain perception in humans.While we did not assess pain catastrophising or fear avoidance in the current study, this previous work supports the potential neurophysiological mechanism behind the amplification of the pain experience when viewing negative images compared to positive images.
Demographic characteristics may have relevance to the current study in which participants were significantly younger than participants in previous studies (Arnold et al., 2007;de Wied & Verbaten, 2001;Meagher et al., 2001).As such, the IAPS images may be interpreted differently and subsequently not have the same emotional impact on this younger cohort compared to an older cohort.Further investigation is necessary to examine the efficacy of the IAPS images in different demographic and clinical populations as previous studies have reported variable responses which may be explained by variances in participant demographics (Arnold et al., 2007;Kamping et al., 2013;Rhudy et al., 2013).
The strengths of this current study include assessment of pre-intervention mood state (BRUMS) to ensure there were no day-to-day differences in general mood that could impact on treadmill walking duration, prior to viewing the different IAPS image sets.In addition, the intensities of the noxious stimulus (pressure cuff) and the treadmill walking (speed, slope) were determined at the familiarization session and then standardized for each test session.This ensured that any reported difference in pain intensity or time to the pain limit of 7/10 NRS could be associated with the IAPS image effects.
There are some limitations in the current study which should be considered before generalizing these findings to other populations.Participants were young, healthy individuals, so the effects of acute mood change on painlimited treadmill walking in other populations with an underlying condition that could alter pain perception cannot be elucidated.Moreover, we did not record information on whether any of the young individuals walk regularly, particularly on a treadmill.The current study did not control for the differences in how males and females may respond to images, nor did we control for the level of arousal.The image valence scores will differ for males and females and hence individuals of different sex may respond differently.The current study was not powered to examine these sex differences; however, we note that were no significant difference in the valence scores for males and females during the three conditions recorded at the same time point.
Another limitation was an absence of a strict control condition whereby no images were applied during the painful treadmill walking.Instead, the neutral image condition acted as the control.It is possible that the neutral images did evoke a small change in mood.Conducting a painful walking trial without any IAPS images would have been advantageous in providing a validation of the neutral images and the effects of each mood condition on the pain experience and treadmill walking duration.We only tested the immediate effects of mood induction on induced pain, so longer-term sustained effects are as yet unknown.In addition, there were three participants who did not reach the target pain intensity of 7/10NRS within the 10-min time period.The study protocol used treadmill walking duration as the primary outcome measure, with 10 min set a priori as the maximum duration for each walk test.The intensity of the noxious stimulus was determined during the initial familiarization session to ensure that participants would reach a NRS of 7/10 within the 10-min time limit.Despite this, three participants did not reach at least one experimental session, and the protocol did not allow for extending the treadmill walking duration for these individuals who exhibited greater tolerance to the noxious stimulus.While treadmill walking duration may be a component of physical activity prescription, the maximum limit of 10 min used in our study limits our ability to examine changes in pain intensity over a longer duration.Future studies might consider removing or extending the time constraint to investigate the extent to which mood modulation impacts both the pain experience and exercise capacity.
To our knowledge, the current study is the first to examine the effects of acute mood manipulation on pain perception during a treadmill walking task in healthy individuals.The current study examined the interrelationship between mood, pain and treadmill walking duration in a trifactor model.Future research is directed towards testing the protocol in different clinical populations.The noxious stimulus chosen in the current study stimulated an ischemic pain experience, which attempted to mimic a common debilitating symptom in individuals suffering from peripheral vascular disease.Future studies should confirm these findings in a peripheral vascular disease population, as ischaemic pain, caused by intermittent claudication, not only causes limitation in everyday function but can also act as a barrier to exercise therapy (Hamburg & Balady, 2011;Watson et al., 2008).The effects of acute mood change on pain-limited treadmill walking in participants with chronic pain should also be explored.Discovering more holistic strategies to enhance a positive mood may strengthen compliance to therapy designed to improve physical activity through altering the pain experience and therefore maximizing therapeutic outcomes.If successful in clinical populations, employing these strategies into routine therapy could also reduce the economic and personal burden associated with chronic pain.
Overall, the findings of the study support the hypothesis that inducing a positive mood state while walking with a fixed nociceptive input results in slower increases in pain intensity and longer treadmill walking duration compared to a negative induced mood state in healthy, young participants.Harnessing this interaction may help strengthen compliance to therapy designed to improve physical activity and optimize rehabilitation outcomes, especially for populations in which exercise capacity is limited by pain.