The Effect of Spinal Muscle Fatigue and Psychosocial Factors on Pressure-Pain Threshold in Healthy Adults

Objective Pain sensitivity decreases following isometric exercise. It is not clear whether this exercise-induced hypoalgesia (EIH) occurs to the same extent in men and women. It is also unclear if the effect is systemic or local to the exercised musculature. The aim of our study was to investigate whether fatiguing isometric exercise of the spinal and hip extensors would result in increased pressure pain threshold (PPT) at sites local to and remote from the exercised muscles in healthy men and women and whether there is a relationship between central sensitization, psychosocial factors, and PPT. Subjects 35 healthy adults (age 27.1 ± 4.5 years, 22 women). Methods This was a within-subjects cohort study. Participants completed questionnaires quantifying central sensitization, pain catastrophizing, sleepiness/insomnia, anxiety, and depression. PPT was assessed at the lumbar and thoracic paraspinals, hamstrings, gastrocnemius, wrist, and third digit before and immediately after participants performed the Biering–Sorensen test to failure. Results PPT increased postexercise in the thoracic paraspinals, hamstrings, and gastrocnemius in men and women and in the lumbar paraspinals in men only but did not change at the wrist and digit sites. A lower average PPT at baseline was associated with a higher central sensitization scores. A greater increase in average PPT postfatigue was significantly associated with higher average PPT at baseline. Conclusions Exercise-induced hypoalgesia occurs at sites overlying the muscles involved in fatiguing exercise, but not at remote sites, and is more evident in males than females. The magnitude of EIH depends upon baseline PPT. Even in healthy individuals, greater central sensitization is associated with lower baseline PPT.


Introduction
An individual's pain sensitivity is afected in the short term and long term by multiple factors. Pain sensitivity can be assessed using the measurement of pressure-pain threshold (PPT). Tis is a simple quantitative sensory test of pain that is often used in research and clinical practice [1][2][3][4]. Te PPT is defned as the minimum pressure applied to anatomical site that results in an individual perceiving the mechanical stimulus as pain [5]. Te measurement of PPT is accurate, valid, and reproducible [6,7]. Women typically display a lower PPT than men [8]. Individuals with persistent musculoskeletal pain disorders such as recurrent or chronic low back pain also demonstrate reduced pain thresholds when measured using PPT [9][10][11]. Tis reduced PPT may occur at the location of the clinical symptoms, suggestive of peripheral sensitization mechanisms, and also at locations that are not anatomically or neurophysiologically related to the symptoms, suggestive of supraspinal mechanisms [2,[10][11][12]. To understand the contribution of short-and long-term factors to localize and widespread pain sensitivity in individuals with persistent pain disorders, it is critical to frst examine potential contributing factors in asymptomatic men and women.
Pain sensitivity is modulated in response to exercise [13]. In the short term, isometric muscle exercise appears to decrease pain sensitivity [3,9,14,15]. Tis is called exerciseinduced hypoalgesia (EIH). Te greatest change in pain threshold occurs in response to long-duration, low-intensity isometric contraction that is maintained until failure [16]. However, it is unclear if the reduction in pain sensitivity following muscle fatigue is local to the afected muscle or if it is a generalized response that also occurs at sites remote to the fatiguing exercise [3]. Te Biering-Sorensen test is a long-duration, low-intensity isometric task that is often used to quantify the endurance of the spinal and hip extensor musculature in individuals with and without low back pain [17,18]. Previous preliminary work investigating the efect of the Sorensen test on PPT in the spinal extensors and remote sites suggested that the extent of EIH at remote locations is greater in women [3]. However, in this previous study, the Sorensen test was held for a standardized duration, and so the muscles may not have been fatigued to failure. Sex diferences in the local and systemic hypoalgesic response to fatiguing exercise may depend on testing location as conficting results across studies suggest that these diferences are muscle specifc [3,19].
In the short and long terms, pain sensitivity can increase because of the heightened responsiveness of nociceptive neurons to normal or subthreshold aferent input. Tis phenomenon is known as central sensitization [20]. While central sensitization is a neurophysiological mechanism that cannot be directly measured in vivo, signs and symptoms such as widespread increased pain sensitivity suggest its presence [21]. Central sensitization is known to contribute to many persistent pain disorders [22]. Recently, it has been recognized that central sensitization may also be elevated in individuals who do not have persistent pain. Elevated central sensitization may in fact be a precursor to the development of clinical pain [22,23]. It is not known if central sensitization infuences pain sensitivity in nonclinical populations or if it alters the extent of hypoalgesia following isometric exercise. Clinically, central sensitization can be assessed using a combination of quantitative sensory testing and selfreport measures. One such self-report measure is the Central Sensitization Inventory (CSI) [24], which quantifes the presence of biological and psychological symptoms and characteristics that are associated with central sensitization syndromes [24]. It is not clear, however, how quantitative measures such as PPT and self-report measures such as the CSI are related [25,26].
Multiple psychosocial factors are associated with pain sensitivity. Tese include depression, anxiety, pain catastrophizing, and sleep quality. Depression and increased pain sensitivity frequently occur together [27], but the mechanism underlying this relationship is unclear [28]. Evidence also suggests that in individuals with and without persistent pain, elevated anxiety is associated with decreased PPT [1,12]. Pain catastrophizing has been associated with a number of indicators of pain sensitivity in the context of experimental pain testing paradigms, both among healthy, pain-free participants and individuals with persistent pain disorders [29][30][31][32]. Tere is also increasing evidence of a relationship between sleep quality and an individual's pain experience [33]. In the short term, healthy individuals have increased pain sensitivity following experimentally-induced sleep deprivation [34,35]. Longterm sleep disruption is also associated with increased pain sensitivity in individuals with and without persistent pain disorders [36]. However, it is unclear if any of these psychosocial factors infuence EIH.
Te aim of our study was to investigate whether performing a long-duration, low-intensity fatiguing exercise to failure would result in increased PPT at local and remote sites in healthy men and women and whether there is a relationship between central sensitization, psychosocial factors, and PPT in healthy individuals. It was hypothesized that (1) the fatiguing isometric exercise would produce an increase in PPT at both local and remote sites; (2) the increase in PPT in response to fatiguing exercise would be more pronounced in women than in men; and (3) impaired sleep quality, and increased anxiety, pain catastrophizing, and central sensitization would be associated with lower PPT in healthy adults at baseline and would be associated with a smaller extent of change in PPT in response to fatiguing exercise.

Participants.
Tirty-fve healthy adults (mean age of 27.1 ± 4.5 years, 22 women and 13 men) participated in the study. Sample size exceeded the minimum necessary to detect a main efect of fatigue and the interaction efect of fatigue and sex with a power of 0.90 and an alpha of 0.05 (n = 22, G * Power version 3.1.9.7 [3,37]). Participants were recruited via word of mouth and study fyers at an academic institution and included faculty members, staf, and students. Participants were eligible for inclusion if they were aged between 18 and 60 years. Participants were excluded if they were currently experiencing any back or leg pain, had any history of back pain requiring treatment, or a change in activity for more than one week, if they were currently using any form of analgesic medication and if they had any history of neurological or cardiovascular diseases. Participants gave written consent prior to participating. Chapman University's Institutional Review Board approved this study prior to its commencement.

Psychosocial Factors.
All testing was conducted in the same teaching laboratory space on a consistent day of the week/time of day. Participants completed the Central Sensitization Inventory (CSI), the Pain Catastrophizing Scale (PCS), the Karolinska Sleepiness Scale (KSS), the Insomnia Severity Index (ISI), and the Hospital Anxiety and Depression Scale (HADS) prior to PPT assessment. Te CSI is a widely used questionnaire that is valid and reliable for assessing central sensitization [24,38]. Part A of the CSI contains 25 questions identifying symptoms associated with central sensitization and has a maximum score of 100. Part B assesses if the individual experiences any disorders associated with central sensitivity, such as fbromyalgia and temporomandibular disorder. Te PCS is a reliable and valid measure of pain catastrophizing and includes the domains of pain magnifcation, rumination, and helplessness, with a maximum score of 52 [39]. Te KSS is used to estimate the state-sleepiness of participants, with a score range from 1 to 9 [40]. It has high validity and correlates with EEG and behavioral indicators of sleepiness [41]. Te ISI is a valid and reliable tool to assess perceived insomnia, with scores ranging from 0 to 28 [42]. Te HADS is a validated selfreport measure of psychological status, encompassing depression and anxiety subscales (HADS-D and HADS-A, maximum scores of 21 for each subscale [43]).

Assessment of Pressure-Pain Tresholds.
Prior to the PPT assessment, participants warmed up by walking on a treadmill at a speed of 2.8 mph for 5 minutes. Upon completion of the warm-up, participants quantifed their perceived exertion during the walk on the 0-10 Rating of Perceived Exertion (RPE) scale [44,45]. Participants also rated any pain/discomfort from 0 to 10 after the warm-up on a numeric pain rating scale.
Te sites for PPT assessment were palpated and marked with an indelible pen so that the location of each site was identical prefatigue and postfatigue. Pressure-pain threshold was assessed at the following sites on the participants' dominant sides ( Figure 1): the erector spinae at the level of T9, 3 cm lateral to the spinous process (thoracic site); erector spinae at the level of L4, 2 cm lateral to the spinous process (lumbar site); the hamstrings halfway between the ischial tuberosity and the popliteal crease (hamstring site); the gastrocnemius one third of the distance between the calcaneus and the popliteal crease (calf site); the palmar aspect of the middle phalanx of the third digit (digit site); and the dorsal wrist at the midpoint between the ulnar and radial styloid processes (wrist site).
Pressure-pain threshold was measured utilizing a handheld pressure algometer with a 1 cm diameter circular application area (FDX Digital Force Gauge, Wagner Instruments, and CT). Increasing pressure was applied at a rate of approximately 0.5 kg/s. Te same male experimenter performed all PPT assessments for the study and was blinded to the results. Participants were instructed on the testing procedure and were told to verbally indicate immediately when the sensation of pressure changed to pain. A practice trial at the dorsal wrist was administered to the participant while seated. Following this, PPT was assessed while the participant was prone on a plinth. Two PPT measurements were taken at each site, with 20 seconds between each measurement [3]. Te order in which the sites were assessed was randomized across participants, but for each participant the order was the same before and after exercise.

Isometric Fatiguing Exercise.
Following the baseline PPT measurements, participants completed the Biering-Sorensen test (Figure 1(a)). Te Sorensen test fatigues the spinal extensor, latissimus dorsi, gluteus maximus, and hamstring muscles [46,47]. Participants were positioned on a Roman chair, with the anterior superior iliac spines aligned with the edge of a pelvic pad. Participants were instructed to hold their body parallel to the ground with the arms crossed across their chest for as long as possible. To assist in monitoring the performance of the test, a plumb bob at the end of a lanyard was placed around the neck of each subject during the test. Te researchers adjusted the lanyard so that the plumb bob was one inch above the seat of a standardheight chair placed in front of the participant. Te test was concluded when the participant placed their hands back down onto the chair or if the plumb bob dropped to the chair, indicating failure to maintain the test position [48], and the duration of the test was noted. Immediately following completion of the Sorensen test, participants rated the intensity of exertion during the test using the same RPE scale and rated the intensity of any pain/discomfort during the test on the numeric pain rating scale. Te PPT measurement was then repeated using identical methods to the prefatigue testing. Postfatigue PPT assessment for all sites was completed within 5 minutes of the end of the fatiguing exercise. Tis ensured that the postfatigue testing was applied during the window of time when the muscles were still measurably fatigued [49].

Statistics.
Variables were checked for assumptions of normality and homoscedasticity. Variables that did not meet the assumptions of normality were log-transformed. Psychosocial characteristics were compared between males and females using independent t-tests and the chi-square test of independence. Te efect of fatigue at each assessment site, and averaged across all assessment sites, was tested using a mixed model ANOVA with the main efect of fatigue (within subject efect), main efect of sex (between subject efect), and interaction efect of fatigue * sex. In the case of a signifcant interaction, pairwise comparisons were made using paired or two sample t-tests with Bonferroni correction for multiple comparisons.
Linear relationships between psychological variables and (a) the average PPT at baseline and (b) average change in PPT were examined using Pearson correlation coefcients.
Statistical analyses were conducted in IBM ® SPSS ® statistical software (Version 26, IBM, Armonk, NY, USA). Level of signifcance was set at 0.05 for all tests.

Results
Participant demographic characteristics are shown in Table 1. Te duration of the Sorensen test did not difer between sexes (female duration 126.6 ± 52.8 s; male duration 126.3 ± 64.8 s). Tere was also no diference between sexes for the rating of perceived exertion at the end of the test or discomfort reported at the end of the test (see Table 1, p > 0.05 for all comparisons). Female participants scored higher on the HADS (anxiety subscale) and tended to score higher on the CSI than males (see Table 1, p � 0.035 and 0.052, respectively). Eight individuals (22% of the cohort) reported at least one disorder in Part B of the CSI. Six individuals reported one disorder. Tese were most commonly headache (n � 2), neck injury (n � 2), and anxiety (n � 2). Two individuals reported two disorders (chronic fatigue/temporomandibular joint disorder and headache/irritable bowel syndrome, respectively).

Baseline PPT and Change in PPT in Response to Fatigue.
Data from one male participant were excluded due to his PPT exceeding the maximum possible pressure for the device at several sites. Group data from three sites were logtransformed to meet assumptions of normality (calf, wrist, and digit). Group data for all sites, prefatigue and postfatigue, are shown in Figure 2. At the wrist, there was no efect of fatigue on PPT (main efect of fatigue F � 0.191; p � 0.665). Males had a higher PPT at the wrist than females (main efect of sex F � 8.346; p � 0.007). At the digit site, there was no main efect of fatigue (main efect of fatigue F � 0.359; p � 0.554) or sex (F � 2.567; p � 0.120). Tere was a signifcant fatigue by sex interaction (F � 4.581; p � 0.041) but none of the post hoc comparisons were signifcant.

Relationships between Baseline PPT, Change in PPT, and
Psychological Characteristics. Data from the PCS and the KSS were log-transformed to meet assumptions of normality. Relationships between baselines PPT, change in PPT, and psychological characteristics across the entire group and for each sex individually are shown in Table 2 and signifcant fndings are reported below.
For baseline PPT, greater central sensitization was signifcantly associated with lower average PPT at baseline (r � −0.352; p � 0.041; Figure 3(a)). In males only, a greater history of insomnia was associated with lower PPT at baseline (r � −0.593; p � 0.042).
For change in PPT following the Sorensen test, greater increase in average PPT postfatigue was signifcantly associated with higher average PPT at baseline (r � 0.576; p � 0.001;  Figure 3(b)). With the sexes considered separately, this relationship was only signifcant in females (r � 0.630; p � 0.002). Greater pain catastrophizing was signifcantly associated with a smaller change in PPT in response to fatigue in females (r � −0.461; p � 0.031).

Discussion
Our study demonstrates that isometric contractions held to failure result in an increase in pressure-pain threshold in healthy individuals at local muscle sites but not at sites that Pain Research and Management are remote to the fatiguing exercise. Contrary to our hypothesis, the increase in PPT in response to fatiguing exercise was more pronounced in men than in women. Importantly, individuals with higher PPT at baseline had the greatest increase in PPT in response to fatigue. Even in our healthy participants, elevated CSI scores were associated with lower baseline PPT. Te infuence of psychosocial factors on baseline PPT and fatigue-induced changes in PPT were sex-dependent. Our fndings confrm the hypoalgesic efect of isometric exercise. Earlier studies have indicated that reduced pain sensitivity in response to exercise is most pronounced following prolonged, low-intensity, isometric contractions.
Tis suggests that recruitment of high-threshold motor units may be an important factor in the hypoalgesic response [14]. Mechanistic studies investigating the causes of EIH have predominantly focused on adaptations in response to chronic, whole-body aerobic exercise in animal models [13]. Tis work has highlighted the infuence of endogenous opioid, serotonergic, and endocannabinoid systems. Tese mechanisms produce generalized pain inhibition. Other work probing localized and generalized responses to isometric exercise in humans has suggested that generalized EIH may occur as a result of increased blood pressure or altered attention, but some research study has also demonstrated that efects are largest in the contracting muscle,  suggestive of a local or segmental efect [9]. In the present study, signifcant increases in PPT following the fatiguing exercise occurred at the sites overlying the spinal and hip extensor muscles. Tese muscles are known to fatigue during the Sorensen test [49,50]. Te local modulation of PPT in the fatigued muscles may be due to aferent inhibition in response to stimulation of sensory fbers during the fatiguing contraction [9]. In this study, the PPT at the calf site also increased following the Sorensen test. Te contribution of the calf musculature to maintenance of the Sorensen test position has not previously been reported. However, it is likely that the gastrocnemius is active during the test to develop a knee fexor moment that keeps the heels stabilized against the test equipment. In our study, we did not observe generalized EIH, as there were no changes in PPTat the wrist and digit sites that were remote to the exercise. Previous work reported increased PPT at the thenar eminence following the Sorensen test [3]. We speculate that the amount of muscular tissue overlying remote sites may infuence the extent of measured EIH due to diferences in tissue thickness, tissue stifness, and density of mechanoreceptors. Additionally, our fndings are consistent with recent work suggesting that generalized EIH is less likely to occur following fatiguing isometric exercise of the back musculature in comparison with fatiguing exercise of the limb musculature [51]. Te reasons for this are unclear but may be related to varying exercise intensity when exercising axial or limb muscles or difering muscle fber types [51].
In contrast with earlier work using the Sorensen test paradigm [3], we found that the increase in PPT in response to fatiguing exercise occurred more consistently in male participants. Other studies using PPT have variously reported that EIH following isometric contraction is the same in men and women, that it is greater in women, or that it is greater in men [52][53][54]. Te conficting fndings from these studies may in part be due to varying fatigue protocols [53]. In the present study, participants held the Sorensen position for as long as possible, until failure, rather than for a standardized length of time. Tis ensured that participants of all capabilities reached the limit of their muscle endurance. Although the duration of the hold time varied widely across individuals (20 seconds to 250 seconds), the average hold time (126 seconds) was greater than the standardized time used in the previous study (120 seconds) [3]. Our study supports earlier work suggesting that sex diferences in EIH emerge primarily in response to more demanding exercise [53]. Te smaller amount of EIH evident in female participants in our study may also have been due to the female group tending to have higher levels of anxiety and higher scores on the CSI than male participants. Given the heterogeneity in the evidence for the infuence of sex on EIH and the potential interaction with psychological and physiological factors, future studies would beneft from large cohorts of male and female participants matched for psychosocial characteristics and should investigate EIH in response to a range of exercise intensities and durations.
Consistent with earlier work [55], baseline PPT was higher in males than females in this study. Factors that may account for the sex diference in pain sensitivity include the infuence of sex hormones on nociception and diferences in the function of the descending opioid system [55]. Sex diferences in baseline PPT may also be due in part to diferences in muscle size. Binderup et al. [56], proposed that as women have smaller muscle bulk on average, the size of the algometer probe head is relatively larger, leading to a lower PPT due to spatial summation. Te results of our study lend support to this hypothesis, as we found PPT diferences by sex at every site except for the 3rd digit, the only site where there is little muscle tissue.
Importantly, we found that the extent of EIH was linearly related to baseline PPT. Participants who demonstrated lower pain sensitivity at baseline had greater EIH. Vaegter et al. [57] have previously reported that increased pain sensitivity, and higher intensity of clinical pain at baseline, reduces exercise-induced hypoalgesia in individuals with back pain. Our fndings demonstrate that the infuence of baseline pain sensitivity on EIH extends to healthy individuals without pain. Tis indicates that a uniform hypoalgesic response to isometric exercise should not be assumed, even in healthy adults. We also demonstrate in this study that even in a population of healthy adults, higher scores on the Central Sensitization Index are associated with lower PPT at baseline. Tis is despite the fact that average scores on the CSI in our study were well below the threshold of 40 that has been suggested as the cutof score for clinically relevant central sensitization in patient populations [58][59][60] and were also lower than scores previously reported in nonpatient cohorts [59]. Te frequency of our participants reporting diagnoses associated with central sensitization (22%) was also far lower than that reported in patient populations [59]. Existing evidence for the relationship between scores on the CSI and quantitative measures of pain sensitivity in patient populations is mixed [26], but our study suggests an association between biopsychosocial characteristics identifed by the CSI and pain sensitivity in healthy adults.
Other factors infuencing baseline PPT or change in PPT postfatigue were sex-dependent. Females with higher pain catastrophizing scores demonstrated less hypoalgesia in response to the fatiguing exercise. Our fndings build upon work by Naugle et al. [52], who reported that levels of pain catastrophizing were predictive of exercise-induced changes in the temporal summation of heat pain. Te relationship between pain catastrophizing and reduced EIH may be a result of reduced descending opioid pain inhibition. It is not clear why the infuence of pain catastrophizing on hypoalgesia was only evident in women, as, unlike in previous studies [52], there was no diference in pain catastrophizing scores between men and women in our cohort. Our study did not fnd signifcant relationships between sleep quality and baseline PPT or a change in PPT following fatiguing exercise in the combined group analysis. However, in males, we demonstrate that greater insomnia was associated with reduced PPT at baseline. Te diference between our fndings and previous work may be due to the fact that the KSS and ISI scales quantify state-sleepiness and the perceived level of current insomnia [41,42], rather than long-term sleeping impairment. Additionally, levels of sleep Pain Research and Management 7 disturbance reported in this study were below the threshold for clinically signifcant impairment [42] and were far lower than in studies involving experimental sleep deprivation. Tere were some limitations to our study. Te sample had an unequal distribution of male and female participants. Of note, changes in PPT in response to exercise were more pronounced in the smaller male cohort, indicating that our fndings were not infuenced by a lack of statistical power in the male group. Additionally, as in all studies investigating fatigue, it is difcult to prove objectively that muscle fatigue has occurred. We did not quantify the loss of force production following the Sorensen test, but loss of force production at the end of the test was indicated by the individuals' inability to continue maintaining the test position. In a previous study, using the same Sorensen paradigm in a cohort of young healthy adults, we confrmed with electromyography that fatigue occurs in the lumbar extensor muscles during the test [49]. We have also demonstrated that this fatigue persists for fve to ten minutes after completion of the test, which was more than enough time to complete our PPT testing at all sites [49]. Finally, in this study, the mechanical stimuli were applied manually using a pressure dynamometer. It is possible therefore that the rate of pressure application varied slightly across testing repetitions. However, we minimized this potential confound by having the same trained tester apply the PPT stimuli to all participants.

Clinical Implications.
Te Biering-Sorensen test is a simple way to induce spinal and hip extensor fatigue and to assess EIH in healthy individuals and in clinical pain populations. Our fndings of a signifcant relationship between PPT and scores on the CSI suggest that PPT testing may be a useful quantitative adjunct to self-report central sensitization questionnaires in the clinical setting. Isometric strengthening exercise is often recommended as an intervention for individuals with low back pain, and training that includes isometric extensor exercises such as "planks" or "bird dogs" is a common component of rehabilitation.
However, since our study shows that the extent of EIH depends upon pain sensitivity at baseline, individuals with LBP with high baseline pain sensitivity may not respond in the same way to isometric exercise as back-healthy controls. Future studies should clarify if fatiguing isometric exercise reduces pain sensitivity in all patients with low back pain.

Conclusion
Isometric spinal and hip exercise held to failure produces a local hypoalgesic response. Tis occurs in both males and females, but to a greater extent in males. In healthy participants, there is greater EIH in those with lower pain sensitivity at baseline. In individuals without a clinical pain condition, a higher self-reported rating on the CSI is associated with a lower baseline PPT.

Data Availability
Data are available upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.