EXERCISE PREVENTS DEVELOPMENT OF AUTONOMIC DYSREGULATION AND HYPERALGESIA IN A MOUSE MODEL OF CHRONIC MUSCLE PAIN

Abstract

Chronic musculoskeletal pain (CMP) conditions, like fibromyalgia, are associated with widespread pain and alterations in autonomic function. Regular physical activity prevents development of CMP and can reduce autonomic dysfunction. We tested if there were alterations in autonomic function in sedentary mice with CMP, and if exercise reduced the autonomic dysfunction and pain induced by CMP. CMP was induced by two intramuscular injections of pH 5 in combination with a single fatiguing exercise task. A running wheel was placed into cages so that the mouse had free access for either 5 days or 8 weeks (exercise groups) and these animals were compared to sedentary mice without running wheels. Autonomic function and nociceptive withdrawal thresholds of the paw and muscle were assessed before and after induction of CMP in exercised and sedentary mice. In sedentary mice, we show decreased baroreflex sensitivity, increased blood pressure variability, decreased heart rate variability and decreased withdrawal thresholds of the paw and muscle 24h after induction of CMP. There were no sex differences after induction of the CMP in any outcome measure. We further show that both 5 days and 8 weeks of physical activity prevent the development of autonomic dysfunction and decreases in withdrawal threshold induced by CMP. Thus, this study uniquely shows development of autonomic dysfunction in animals with chronic muscle hyperalgesia that can be prevented with as little as 5 days of physical activity, and suggest that physical activity may prevent the development of pain and autonomic dysfunction in people with CMP.

1. Introduction

Chronic musculoskeletal pain (CMP) conditions, like fibromyalgia, affect between 11–24 % of the population [15;38] and result in significant disability [2]. CMP has been associated with autonomic dysfunction in humans, albeit with substantial heterogeneity of dysfunction among subjects [42;61]. Decreases in heart rate variability and baroreflex sensitivity, and abnormal autonomic responses to physiological challenges have been most consistently observed in people with chronic pain [16;19;26;32;54;69;72;73]. Autonomic dysfunction predicts premature death post-myocardial infarction and in heart failure [43], and is associated with cardiovascular and renal complications in diabetes [37]. It is unclear if induction of muscle pain induces autonomic dysfunction, or if these changes parallel the hyperalgesia. Therefore, a better understanding of autonomic regulation in chronic pain is needed.

Repeated muscle insults produce long-lasting and widespread hyperalgesia in mice and rats [22;84;98]. In particular, repeated intramuscular acid injections, or acid injection in combination with muscle fatigue, result in hyperalgesia of the muscle, skin and viscera without overt tissue injury [22;30;84;88;98;99] . The hyperalgesia in these models, once developed, is independent of peripheral input and requires continued activation of central pathways to maintain the hyperalgesia [20;84;93]. Furthermore, these models are sensitive to treatments commonly used in CMP including opioids, dual and serotonin reuptake inhibitors, pregabalin/gabapentin and exercise [22;64;85;98]. Thus, these animal models of chronic muscle pain mimic the signs, symptoms and responsiveness to treatment observed clinically in fibromyalgia, further validating the models.

Regular exercise may reduce cardiovascular risk in part by decreasing sympathetic tone and increasing parasympathetic tone and baroreflex sensitivity [46;52;67]. In addition, regular exercise can prevent and reduce pain and disability in humans with CMP [8;10], and hyperalgesia in animal models of pain [6;78;85;91]. In fact, regular physical activity reduces the risk of both heart disease [4] and development of chronic pain [44;45]. Whilst some studies have shown that exercise can correct autonomic dysfunction in fibromyalgia, others do not show beneficial effects [27;42]. It is possible that different types of exercise, different intensities of exercise, or different durations of exercise underlie these conflicting data. Our previous study shows that 5 days of physical activity prevented development of exercise-induced secondary hyperalgesia but not primary hyperalgesia, and 8 weeks of exercise prevented both primary and secondary hyperalgesia [85]. While it is generally recommended that several weeks of exercise are required to induce training effects [33], this has not been tested in animals with chronic muscle pain. The primary aims of this study were to 1) extensively characterize the autonomic dysfunction induced by CMP, 2) determine if regular physical activity prevents the autonomic dysfunction induced by CMP, and 3) determine if prevention of autonomic dysfunction occurs with a short-duration of physical activity (5 days) or requires a more long-term exercise task (8 weeks). In parallel, we examined the ability of short- and long-duration physical activity to prevent primary and secondary hyperalgesia induced by CMP.

2. Methods

2.1. Mice

All experiments were approved by the University of Iowa Animal Care and Use Committee and were conducted in accordance with National Institutes of Health guidelines. Mice used were on a C57BL/6J background and were bred at the University of Iowa Animal Care Facility or purchased from Jackson Laboratories. Both male (n=40) and female (n=32) mice were used in these studies. Animals were 9–10 weeks old when starting the study.

2.2. Physical activity in running wheels

Running wheels (Columbus Instruments or University of Iowa Animal Care Facility) were placed in home cages and all mice were housed individually. Thus mice had free access to running wheels. Running wheel activity was recorded and converted to average distance per day. Sedentary groups consisted of housing mice individually in cages for the same duration as the running wheel group.

2.3. Induction of fatigue-induced chronic musculoskeletal pain (CMP) model

After exercise training or sedentary living (5 days or 8 weeks), the CMP model was induced by combining 2 intramuscular injections of pH 5.0 sterile saline solution (20µl) with a 2h fatigue task as previously published [99]. To inject acidic saline, mice were briefly anaesthetized with isoflurane (3%), and the left gastrocnemius muscle was injected. Two injections were given 5 days apart. Just prior to the second injection, mice were run in exercise wheels at a self-selected speed. If mice stopped running, the exercise wheel was gently tapped to encourage continued running.

2.4. Measurement of Nociceptive Behaviors

For behavioral assays, mice were separated into 4 groups: sedentary 5 days in cages (n=12; 8 Males, 4 Females), 5 days running wheel (n=12; 8 Males, 4 Females), sedentary 8 weeks in cages (n=6; 3 Males, 3 Females), 8 weeks running wheel (n=8; 4 Males, 4 Females). Nociceptive behavioral tests were performed before the first injection of acidic saline, 5 days after the first injection of acidic saline, and 24 hours after the second injection of acidic saline. Behavioral testing was separated across multiple days, mice were randomly assigned to groups, and sedentary and running wheel groups were evenly distributed. In these experiments the behavioral tester was not blinded to group.

Mice were tested for mechanical withdrawal thresholds of the paw and muscle, both ipsilateral and contralateral to the injection site, as we previously published [85]. Mice were acclimated to behavioral procedures 2x per day for two days prior to beginning the experiment. To acclimate for mechanical withdrawal thresholds of the paw, mice were placed in small clear cubicles on the wire mesh table for 20 minutes. To acclimate for mechanical withdrawal thresholds of the muscle, mice were placed in a glove and their hind limb was gently stroked.

Mechanical withdrawal thresholds of the paw were tested using Von-Frey filaments (0.008–2 g) using the Dixon up down method. Testing was initiated with a 0.4-gm filament in the middle of the series and withdrawal was determined if a mouse elevated or licked the paw after the filament was pressed against the paw. A decrease in withdrawal threshold (force) was interpreted as cutaneous hyperalgesia.

Muscle withdrawal thresholds were measured with a pair of force-sensitive forceps, built in house, and applied to the belly of the gastrocnemius muscle [87]. Mice were held in a gardener’s glove while the hind quarters were pulled into extension. The muscle was squeezed with the forceps until the mouse withdrew the hindlimb. Three trials were conducted and an average was taken. A decrease in withdrawal threshold was interpreted as muscle hyperalgesia.

2.5. Radiotelemetry

For telemetry protocols, separate groups of mice were randomly divided into 4 groups: non-runners 5 days in cages (n=8–10; 3–5 Males, 5 Females), 5 days running wheel (n=8–12; 2–6 Males, 6 Females), non-runners 8 weeks in cages (n=5; 2 Males, 3 Females), 8 weeks running wheel (n=7; 4 Males, 3 Females). All mice underwent the CMP protocol after the wheel-running or sedentary intervention. The intervention of 5 days or 8 weeks was performed and running wheels removed from cages for the duration of the experiment. Telemetry recordings were obtained at 3 time points in every mouse: 1) baseline, control after recovery from implant surgery before the start of the wheel-running or sedentary intervention, 2) 5 days or 8 weeks after initiating the wheel-running or sedentary intervention and before induction of the CMP model, and 3) 24 hours after induction of the CMP model. After the first baseline testing period, the runners were put in cages with running wheels attached and allowed to run (8 weeks or 5 days), while non-runners (sedentary) were put in individual cages without running wheels for the same duration of time. The experimenter was blinded to group.

A telemetry probe (TA11PA-C10, DSI) was inserted into the thoracic aorta via the left common carotid artery in mice anaesthetized with ketamine (91 mcg/g, i.p.) and xylazine (9.1 mcg/g, i.p.) at 9–10 weeks of age as described previously [74]. Animals were allowed to recover from the surgery for 7 days before beginning the protocols.

2.6. Assessment of Cardiovascular and Autonomic Indices

For assessment of cardiovascular and autonomic indices, mice were transferred to the telemetry room, placed into a novel cage, and acclimated for an hour before data were collected. Blood pressure and heart rate were recorded continuously at 2000 Hz for 1–2 hours between 10:00 and 13:00 to enable collection of beat-to-beat data for assessment of baroreflex sensitivity, blood pressure variability, heart rate variability, and vasomotor sympathetic tone [74].

Baroreflex sensitivity for control of heart rate was calculated from spontaneous fluctuations in systolic blood pressure and heart rate measured over a 15 minute period when mice were relatively active (based on the locomotor activity signal) using the well-established sequence technique [47;74]. Sequences of four-or-more consecutive blood pressure pulses where systolic pressure and pulse interval (inversely related to heart rate) were positively correlated (r²>0.85) were detected using a customized software (Hemolab). Baroreflex sensitivity was calculated as the average slope of the systolic pressure-pulse interval relationships (Δms/ΔmmHg). In addition, blood pressure variability was calculated as the standard deviation (SD) of the systolic blood pressure values measured over the 15 minute period [74].

To estimate resting autonomic tone, heart rate variability was calculated from beat-to-beat measurements of pulse intervals collected over a 5 minute period when mice were relatively inactive using Hemolab and Batch Processor software. The standard deviation (SD) of the recorded pulse intervals and the root mean square of successive differences in pulse intervals (RMSSD) provided measures of overall heart rate variability and rapid, parasympathetic-mediated heart rate variability, respectively [11;74]. Resting sympathetic vasomotor tone was estimated by measuring the decrease in mean arterial pressure induced by injection of the ganglionic blocker chlorisondamine (12 mcg/g, IP, Tocris) when the mice were relatively inactive as described previously [74].

2.7. Data Analysis

The results are expressed as means ± standard error of the mean (SEM). Statistical evaluation was performed using paired and unpaired t-tests to compare measurements obtained before and after an intervention in the same mice and measurements between two groups of mice, respectively. Repeated measures, 2 factor ANOVA with Fishers PLSD post-hoc test was used to determine effects of CMP and wheel-running activity on nociceptive behaviors, and cardiovascular and autonomic indices (StatView SAS Institute, Cary, NC). A secondary analysis examined for sex differences by combining the control sedentary mice from the 5-day and the 8-week running groups (n=8 female, n=5–7 male). This allowed us to perform a repeated measures ANOVA (baseline, post CMP) with an analysis for sex. Significant differences were defined at P<0.05.

3. RESULTS

3.1. Effects of Inducing CMP Model on Cardiovascular, Autonomic, and Nociceptive Measures

As prior clinical studies have shown alterations in heart rate variability and other indices of autonomic nervous system activity in people with chronic pain, we examined if cardiovascular and autonomic measures are altered in our CMP model. We combined the sedentary comparison groups from the 5 day and the 8 week exercise groups to give us a total of 15 mice (n=8 female; n=5–7 male) for analysis of autonomic function. In comparison to baseline, induction of the CMP model significantly decreased baroreflex sensitivity by 30 % from 2.4±0.12 to 1.7±0.06 ms/mmHg, and increased blood pressure variability by 100 % from 3.2±0.23 to 6.5±0.44 mmHg (Fig 1). In addition, sedentary mice with CMP showed decreased heart rate variability measured as both the SD and the RMSSD of pulse intervals (Fig 1). The decrease in RMSSD, a specific measure of parasympathetic-mediated heart rate variability, was particularly pronounced averaging 49 % (10.2±1.3 to 5.2±0.92 ms, Fig 1). Interestingly, there were no changes in mean arterial pressure (117±2.2 vs. 119±1,8 mmHg), mean heart rate (664±8 vs. 665±14 bpm), or resting sympathetic vasomotor tone measured as the depressor response to ganglionic blockade (−59±2.9 vs. −63±1.6 mmHg) after induction of the CMP model in sedentary mice (Fig 1). Cardiovascular and autonomic measurements were measured twice under control conditions, first under baseline conditions after recovery from the telemetry implant surgery and again after an additional 5 days of sedentary cage-living. There were no significant differences between the two sets of measurements. All changes elicited by induction of the CMP model noted above were also significant when measurements were compared with the second set of control measurements (Table 1). Table 1 provides the mean and SEM for each measure in the sedentary non-runner groups before and after induction of chronic muscle pain.

Figure 1. Effects of inducing the chronic musculoskeletal pain (CMP) model on cardiovascular and autonomic indices in sedentary mice.

Measurements were obtained under baseline control conditions (black bars) and 10 days later, after induction of the CMP model (gray bars) in the same mice (n = 13–15). Sedentary animals from the 5-day group and from the 8-week group were combined. Induction of CMP decreased baroreflex sensitivity and heart rate (HR) variability, and increased systolic blood pressure (BP) variability, without altering mean BP, mean HR or sympathetic vasomotor tone. *P<0.05, CMP vs. control period, paired t-test. Data are means ± SEM. SD of PI, standard deviation of pulse intervals, a general measure of HR variability. RMSSD, root mean square of successive differences in pulse intervals, a measure of parasympathetic-mediated HR variability.

Table 1.

Cardiovascular and autonomic indices in sedentary (Non-Runners) and exercised (runners) mice before (Control) and after induction of chronic musculoskeletal pain (CMP). Corresponding values from 8 weeks and 5 day protocols are shown in the table.

	Non-Runners		Runners
	Control	CMP	Control	CMP
Body Weight (g)
8 weeks	20±4	20±3	22±4	21±2
5 days	21±3	20±3	20±3	20±1
Mean BP (mmHg)
8 weeks	112±3	117±5	105±2	104±2*
5 days	121±2	119±1	108±1*	105±2*
BP Variability (mmHg)
8 weeks	3.2±0.6	6.9±1.1†	2.2±0.3	2.5±0.6*
5 days	3.5±0.7	6.3±0.9†	3.2±0.4	3.0±0.7*
Baroreflex sensitivity (ms/mmHg)
8 weeks	2.5±0.2	1.6±0.1†	2.9±0.2	2.6±0.2*
5 days	2.3±0.2	1.7±0.1†	2.6±0.1	2.4±0.1*
Sympathetic Tone (AmmHg)
8 weeks	−53±4	−59±4	−35±4*	−26±3†*
5 days	−63±1	−66±2	−37±3*	−34±3*
Mean HR (bpm)
8 weeks	655±15	645±8	527±8*	510±6*
5 days	680±13	664±8	535±14*	505±7*
HR Variability-SD of PI (ms)
8 weeks	18.4±4.0	5.7±2.1†	24.1±3.9	21.9±2.7*
5 days	13.8±1.2	9.0±1.0†	18.8±1.6	16.4±1.1*
HR Variability-RMSSD (ms)
8 weeks	11.1±2.5	2.7±1.2†	15.8±3.3	15.1±2.7*
5 days	9.8±1.2	4.7±0.7†	12.7±1.5	11.6±1.6

BP: blood pressure; HR: heart rate; SD: overall HR variability; RMSSD: parasympathetic-mediated HR variability. Data are means ± SEM.

^*Runners vs. Non-Runners, P<0.05

^†CMP vs. Control, P<0.05

As we have previously shown [83;99], the combination of exercise-induced fatigue with pH 5.0 intramuscular injections resulted in decreased both paw and muscle withdrawal thresholds, bilaterally, 24h after the second intramuscular acid injection in the group of mice that did not have access to running wheels (n=18/group). Decreases were significant (P=0.001–0.00001) when comparing the withdrawal thresholds prior to the first or second injection with the withdrawal thresholds after the second injection (Fig 2).

Figure 2. Effects of fatigue combined with acid injections on paw and muscle withdrawal threshold.

Bar graphs representing the paw withdrawal threshold and the muscle withdrawal threshold before injection 1, before injection 2 and 24h after the second acid injection for all animals (n=18) for the ipsilateral (black bars) and the contralateral (gray bars). A significant decrease in withdrawal threshold of the paw and muscle occurred bilaterally in both groups, *, p<0.001. Data are the means ± SEM.

As a secondary exploratory analysis, we pooled the data from the control sedentary groups from the 5-day and the 8-week running wheel groups and examined for sex differences. For autonomic dysfunction measures (n=8 females; n=5–7 males), there were significant differences for sex for MAP (p=0.01), for sympathetic tone (p=0.01), and for baroreceptor sensitivity (p=0.01) but not for HR, heart rate variability (SD, RMSSD), or blood pressure variability (Table 2). There was no interaction between time of measurement and sex suggesting that differences in MAP, sympathetic tone and baroreceptor sensitivity occur at both time periods and are unaffected by induction of CMP. For nociceptive measures there were no sex differences in muscle or paw withdrawal thresholds before or after induction of CMP (Table 2).

Table 2.

Sex Differences for Autonomic and Nociceptive Measures at Baseline and After CMP.

Measure		Baseline		CMP
		Male	Female	Male	Female
HR (bpm) (n=8 female, n=7 male)		660±12	653±20	649±8	666±9
MAP (mmHg) (n=8 female, n=7 male)*, p=0.01		122±2.8	112±2.6	121±1.5	116±2.6
SympTone (ΔmmHg) (n=8 female, n=7 male)*, p=0.01		−65±2.4	−53±2.6	−66±2.4	−61±2.4
BRS (ms/mmHg) (n=8 female, n=7 male)*, p=0.02		2.1±.04	2.6±.17	1.6±.10	1.7±.07
SDNN (ms) (n=8 female, n=5 male)		13.1±1.8	17.9±2.7	6.0±1.7	8.8±1.1
RMSSD (ms) (n=8 female, n=5 male)		8.3±1.1	11.5±1.7	2.7±3.9	4.7±.69
Blood Pressure Variability (n=8 female, n=5 male)		2.7±.46	3.6±.20	6.9±.59	6.2±.62
Muscle withdrawal threshold (mN) (n=7 female, n=11 male)	Ipsilateral	1724±26	1566±77	1011±52	1027±88
	Contralateral	1758±24	1540±68	1188±69	1288±115
Paw withdrawal threshold (g) (n=7 female, n=11 male)	Ipsilateral	.35±.06	.14±.03	.037±.007	.031±.011
	Contralateral	.45±.10	.33±.08	.14±.08	.039±.013

^*significant effect for sex (repeated measures ANOVA), p<0.05

3.2. Effects of Regular Exercise on Cardiovascular, Autonomic and Nociceptive Measures

Regular physical activity reduces the risk for development of chronic muscle pain in human subjects, prevents the development of chronic muscle pain in mice [44;45;85] and has protective effects on the cardiovascular and autonomic nervous systems [46;52;97]. Therefore we tested if regular physical activity, by allowing free access to running wheels, prevented cardiovascular and autonomic dysfunction in our CMP model. To produce cardiovascular and autonomic adaptations with exercise training, guidelines typically recommend several weeks of regular exercise to induce training effects [33;62]. In contrast, we have shown that a relatively short duration of wheel-running activity of 5 days can prevent hyperalgesia in rodent models of CMP [85], and changes in vasculature can occur within days [92]. Therefore, we examined effects of both 8 weeks and 5 days of voluntary wheel-running activity on, arterial blood pressure, heart rate, autonomic indices, and nociceptive behaviors in the present study. Sex differences were not analyzed due to the small number of mice per sex.

3.2.1. Overall physical activity

Mice in the 5 day running group ran an average of 3.4±0.44 km/day over the 5 day period. Mice in the 8 week running group ran an average of 6.5±0.19 km/day over the 8 week period. The first several days of running in the 8 week group were lower (4.2 km/day, 3.6 km/day, and 4.7 km/day on days 1–3) and this increased to a peak of 7.8 km/day by the 3^rd week. Body weights of the mice with running wheels were similar to sedentary mice (Table 1).

3.2.2. 8 weeks of exercise prevents autonomic dysfunction and hyperalgesia in CMP model

Mice subjected to 8 weeks of wheel-running activity exhibited significantly lower mean heart rate and sympathetic vasomotor tone than age-matched sedentary mice prior to induction of the fatigue-induced pain model (Fig 3, Table 1). Mean arterial pressure tended to be lower and baroreflex sensitivity and heart rate variability tended to be higher in runners vs. non-runners, but the differences were not significant (Fig 3, Table 1).

Figure 3. Effects of 8 weeks of wheel running on cardiovascular and autonomic indices during the control period and after induction of the CMP model.

Measurements were obtained after either 8 weeks of ‘sedentary’ living (Non-Runners, black bars, n = 5) or 8 weeks of wheel-running activity (Runners, gray bars, n =7). Mice subjected to 8 weeks of wheel running exhibited lower mean heart rate (HR) and lower sympathetic vasomotor tone during the control period than sedentary mice. Baroreflex sensitivity, blood pressure (BP) variability, and HR variability did not differ between the groups. Mean BP, mean HR, and sympathetic vasomotor tone measured after induction of CMP were lower in mice previously subjected to wheel running than in sedentary mice. Furthermore, the prior period of wheel running prevented CMP-induced decreases in baroreflex sensitivity and HR variability, and the CMP-induced increase in BP variability. *P<0.05, Runners vs. Non-Runners; †P<0.05, CMP vs.control period.

While induction of the CMP model in sedentary mice (non-runners) did not change mean arterial pressure, heart rate, or sympathetic vasomotor tone, it markedly decreased baroreflex sensitivity (P=0.001) and heart rate variability (SD and RMSSD of pulse intervals, P=0.016); and increased blood pressure variability (P=0.017) (Fig 3, Table 1). All of these CMP-induced changes observed in sedentary mice were abolished in the mice previously subjected to 8 weeks of wheel-running activity (Fig 3, Table 1). After induction of the CMP model, the wheel-running mice exhibited significantly lower mean arterial blood pressure (104±2 vs. 117±5 mmHg), blood pressure variability (2.5±0.6 vs. 6.9±1.1 mmHg), mean heart rate (510±6 vs. 645±8 bpm) and sympathetic vasomotor tone (−26±3 vs. −59±4 ?mmHg); and significantly higher baroreflex sensitivity (2.6±0.2 vs. 1.6±0.1 ms/mmHg), and heart rate variability (21.9±2.7 vs. 5.7±2.1 ms for SD of PI; and 15.1±2.7 vs. 2.7±1.2 ms for RMSSD) compared with sedentary mice.

Eight weeks of wheel-running activity prevented the decrease in withdrawal threshold of the paw after induction of the CMP model. The exercise-trained mice exhibited significantly greater withdrawal thresholds when compared with sedentary controls (P=0.04 ipsilateral; P=0.009 contralateral) (Fig 4). The decreased withdrawal threshold of the muscle was also significantly higher ipsilaterally in the group that performed 8 weeks of wheel-running activity when compared with sedentary mice (P=0.01, Fig 4).

Figure 4. Effects of 8 weeks of exercise training on paw and muscle withdrawal thresholds in the fatigue-induced pain model.

Line graphs showing the mechanical withdrawal threshold of the paw before the first injection, before the second injection on day 5 (D5) and 24h after the second acid injection (D6) in the group with access to running wheels for 8 weeks (solid circles) when compared to the sedentary group (open circles). The withdrawal thresholds were greater in the running wheel group compared to the non-running group 24h after the second acid injection (*, p<0.05). Data are the means ± SEM.

3.2.3. 5 days of exercise prevents autonomic dysfunction and hyperalgesia in CMP model

Mice with access to running wheels for 5 days exhibited significantly lower mean arterial pressure, mean heart rate, and sympathetic vasomotor tone; and higher heart rate variability than age-matched sedentary mice prior to induction of the CMP model (Fig 5, Table 1). Induction of the CMP model did not change mean arterial pressure, heart rate, or sympathetic vasomotor tone in sedentary mice (Fig 5, Table 1). In contrast, induction of the CMP model decreased baroreflex sensitivity (P=0.01) and heart rate variability (SD and RMSSD, P=0.002), and increased blood pressure variability (P=0.001)(Fig 5, Table 1). The CMP-induced changes observed in sedentary mice were abolished in the mice previously subjected to 5 days of wheel-running activity (Fig 5, Table 1). After induction of the CMP model, the runners exhibited significantly lower mean arterial blood pressure, blood pressure variability, mean heart rate, and sympathetic vasomotor tone; and significantly higher baroreflex sensitivity and heart rate variability compared with sedentary mice (Fig 5, Table 1). Interestingly, the effects of 5 days of wheel-running on cardiovascular and autonomic measures (Fig 5) were similar to the effects of 8 weeks of wheel-running (Fig 3, Table 1).

Figure 5. Effects of 5 days of wheel running on cardiovascular and autonomic indices during the control period and after induction of the CMP model.

Measurements were obtained after either 5 days of ‘sedentary’ living (Non-Runners, black bars, n = 8–10) or 5 days of wheel-running activity (Runners, gray bars, n = 8–12). Mice subjected to 5 days of wheel running exhibited lower mean blood pressure (BP) and heart rate (HR) and lower sympathetic vasomotor tone during the control period than sedentary mice. Baroreflex sensitivity and BP variability did not differ between the groups, whereas HR variability (SD of PI) was increased in the runners. Mean BP, mean HR, and sympathetic vasomotor tone measured after induction of CMP were lower in mice previously subjected to wheel running than in sedentary mice. Furthermore, the prior period of wheel running prevented CMP-induced decreases in baroreflex sensitivity and HR variability, and the CMP-induced increase in BP variability. *P<0.05, Runners vs. Non-Runners; †P<0.05, CMP vs. control.

As expected, 5 days of wheel-running activity prevented the decrease in withdrawal threshold of the paw in the CMP model with mice showing a significantly greater withdrawal threshold of the paw (ipsilateral P=0.03; contralateral P=0.04) when compared to sedentary mice (Fig 6). In contrast, 5 days of wheel-running activity had no effect on the decreased withdrawal threshold of the muscle with mice showing similar decreases in muscle withdrawal thresholds when compared to sedentary mice after induction of the CMP model (Fig 6).

Figure 6. Effects of 5 days of exercise training on paw and muscle withdrawal thresholds in the fatigue-induced pain model.

Line graphs showing the mechanical withdrawal threshold of the paw before the first injection, before the second injection on day 5 (D5) and 24h after the second acid injection (D6) in the group with access to running wheels for 5 days (solid circles) when compared to the sedentary group (open circles). The withdrawal thresholds were greater in the running wheel group compared to the non-running group 24h after the second acid injection (*, p<0.05). Data are the means ± SEM.

4. Discussion

4.1. Autonomic dysfunction in CMP

The current study provided a comprehensive assessment of autonomic function and uniquely shows CMP decreases baroreflex sensitivity, increases blood pressure variability, and decreases heart rate variability without altering HR, MAP, or sympathetic vasomotor tone. The marked decrease in RMSSD indicates decreased heart rate variability reflects loss of parasympathetic modulation. Abnormalities in autonomic regulation occur in people with CMP, with substantial heterogeneity between studies that is attributed to co-morbidities, subgroups with different underlying mechanisms, methodological issues, medication effects, age, or sex [42;61]. Despite the heterogeneity, decreased heart rate variability and baroreflex sensitivity are found consistently in people with CMP [14;16;19;26;32;54;69;72;73]. The current data are consistent with our prior study showing changes in heart rate variability and baroreceptor sensitivity after induction of CMP by repeated acid injections in rats [65], and in people with CMP. We extend our prior study by performing a more comprehensive analysis and show changes in heart rate variability reflect altered parasympathetic, but not sympathetic, activity.

The current study shows no sex differences in autonomic dysfunction after induction of CMP. We further show that there are no sex differences in nociceptive measures after induction of CMP which is consistent with our prior studies using repeated intramuscular acid injections [86], but not with our prior study which induced muscle pain by combining a localized isometric contraction with pH 5.0 injections [30]. These data suggest that different muscle stimuli result in uniquely different manifestations across the sexes. Prior work summarizes the complexities of sex differences in pain and analgesia concluding that the biological mechanisms are complex and could include sex hormones, endogenous opioid function, and genetics [5].

The baroreceptor reflex minimizes blood pressure variability by altering heart rate and vascular resistance in response to blood pressure fluctuations [40]. As a result, low baroreflex sensitivity manifests as decreased heart rate variability and increased blood pressure variability [55] that can cause end-organ damage and increase cardiovascular risk [50;55]. As predicted, the decrease in baroreflex sensitivity in our mouse model of CMP was accompanied by a 2-fold increase in blood pressure variability. Furthermore, activation of arterial baroreceptors and vagal afferents reflexively inhibit nociception in animals and humans [9;71], and spontaneous baroreflex sensitivity is inversely related to severity of pain in humans [14;24;72;73]. To our knowledge, the possibility that blood pressure variability may contribute to CMP has not been investigated previously.

People with fibromyalgia exhibit increases in the ratio of low frequency (LF) to high frequency (HF) heart rate variability (LF/HF)[16;26;54]. While LF/HF is indicative of cardiac sympathovagal balance, it does not correlate with sympathetic nerve activity to either heart or skeletal muscle [28;39;51]. Prior evidence suggests that decreased parasympathetic-mediated HF heart rate variability is responsible for the increase in LF/HF ratio in people with fibromyalgia [16;26;32;54;72;73]. We further show no difference in sympathetic vasomotor tone in our model of CMP, and prior studies suggest, sympathetic nerve activity contributes to pain [3;53;75]. However, direct measurement of sympathetic nerve activity to skeletal muscle shows conflicting results in people with fibromyalgia - one study shows increased activity [26], and another no difference [25]. Taken together, these results suggest a strong positive correlation between decreased baroreflex sensitivity and heart rate variability and increased pain.

4.2. Effects of exercise on autonomic dysregulation

The current study shows regular physical activity prevents development of autonomic dysfunction induced by CMP, and this effect can be prevented with as little as 5 days of running wheel activity. These studies are the first to show that autonomic dysfunction induced by CMP can be reversed by regular physical activity, and that 5 days of exercise produces an equivalent effect on normalizing autonomic dysfunction as 8 weeks of exercise. Regular exercise produces numerous adaptations that influence autonomic regulation in humans and animals. Exercise training for 4 weeks to 6 months decreases resting heart rate in part by increasing parasympathetic tone [1], and is associated with increased heart rate variability, decreased sympathetic tone and blood pressure, and altered baroreflex sensitivity [1;13;23;48;52;57;66;76]. Responses vary between studies depending on subject age, presence of disease, and assessment methods. Most notably, baroreflex sensitivity for control of heart rate is decreased [13;23;63;89] or unchanged [48;76;79] after exercise training in healthy humans and animals, whereas it is consistently increased by training in pathological states associated with impaired baroreflex sensitivity, such as hypertension and heart failure [46;48;52;57;66;82]. Similarly, we show that neither 5 days nor 8 weeks of wheel running altered baroreflex sensitivity in control mice, but both training regimens restored baroreflex sensitivity and heart rate variability in mice with CMP.

The ability of 5 days of wheel-running activity to alter cardiovascular responses was surprising since guidelines and numerous studies recommend several weeks of exercise to induce training effects [33]. Consistent with the current study, treadmill exercise (5 days/wk) improves baroreflex sensitivity in spontaneously hypertensive rats within 1 week and wheel running begins to decrease heart rate in mice after 5 days [1;57]. However, exercise-induced decreases in blood pressure occur much later between 4 and 8 weeks of training [1;57]. It also is possible that the rapid changes in cardiovascular responses observed after 5 days of wheel running reflect attenuation of stress-induced autonomic responses that occur when moving sedentary mice to a novel cage to obtain measurements. Another possibility is that ‘postexercise hypotension’ contributes to our results as a single bout of aerobic exercise produces long-lasting decreases in sympathetic activity and blood pressure and increases in baroreflex sensitivity after exercise is terminated [12;17;31;90].

4.3. Effects of exercise on hyperalgesia

The current study shows that prior running wheel activity can prevent the onset of muscle and paw hyperalgesia induced by muscle insult, and is similar to our prior studies using free access to running wheels [85], and are consistent with clinical studies showing that physical activity reduces the risk for development of chronic musculoskeletal pain [44;45]. We further show that 5 days of running wheel activity prevented the hyperalgesia of the paw but not the muscle in our CMP model induced by combining fatigue with intramuscular acid, and is similar to our prior study in CMP models [85]. On the other hand, 8 weeks of exercise prevents both muscle and paw hyperalgesia in several models of CMP [85]. It is likely that once developed the hyperalgesia in the CMP models becomes independent of nociceptive input and has a strong central component [22]. Indeed, prior studies show that our CMP models increase central neuronal activity, including enhanced phosphorylation of the NMDA receptor, in brainstem neurons in the rostral ventromedial medulla (RVM) [83;87], and that regular physical activity prevents this increased activity [85]. Although long-term effects of exercise were not examined in the current study, we previously show the effects of exercise are temporary with hyperalgesia returning 1–2 week after stopping exercise [85]. Analgesia produced by regular exercise is reversed by blockade of opioid receptors with naloxone systemically in animal models of muscle pain and nerve injury [6;77;91], there is release of endogenous opioids in central inhibitory pathways including the RVM [91], and blockade of central but not peripheral opioid receptors reverses the analgesia [91]. Thus, these studies suggest that exercise produces analgesia by activation of central inhibitory pathways that reduce increased excitability of neurons in the RVM.

4.4. Potential overlapping mechanisms between autonomic dysfunction and pain

Extensive anatomical and functional interactions exist between central cardiovascular and pain regulatory systems [7;18;29;56;68;94]. The mechanisms underlying the changes in autonomic function induced by CMP, and the normalization of those by regular exercise remain to be determined. Sustained periods of physical exercise induce neuroplasticity in nervous system circuits regulating efferent parasympathetic and sympathetic activity, sometimes rapidly [12;41;59;60;81]. Many of the neurotransmitter pathways affecting autonomic regulation also inhibit pain, including activation of central α-2-adrenergic receptors, release of endogenous opiates, and modulation of NMDA receptors [6;80;83].

Anatomical and functional characteristics of the RVM strongly suggest that it could play a role in the integration of nociceptive and autonomic responses. The RVM receives inputs from the periaqueductal grey (PAG) which is known to coordinate autonomic and nociceptive stimuli [34;49] and projects to the spinal cord where it targets autonomic and nociceptive neurons [35;36]. In CMP models, the PAG and RVM are both involved in facilitation of nociceptive responses as evidenced by increases in glutamate release, increases in phosphorylation of NMDA receptors (NR1 subunit), and reversal of hyperalgesia after blockade of NMDA receptors in the RVM [20–22;70;85;93]. For autonomic function, microinjection of glutamate or NMDA agonists into the RVLM or RVM increases blood pressure, and autonomic responses induced by stimulation of RVLM are blocked by anesthetic in the RVM [18;94;95]. Furthermore, exercise produces analgesia through central release of endogenous opioids in the PAG and RVM [58;91], and blockade of opioids in the RVM alters autonomic responses associated with opioid withdrawal [96]. Together these data suggest an interaction between motor and cardiovascular systems occurs within brainstem nuclei particularly the RVM and PAG.

4.5.

In summary, induction of CMP alters the cardiovascular system in sedentary mice, but not those that are physically active. These data suggest that physical activity may prevent development of pain and autonomic dysfunction observed in people with CMP. It further suggests an overlap between nociceptive and autonomic pathways underlies these co-morbid conditions.