Large-Fiber Mechanoreceptors Contribute to Muscle Soreness After Eccentric Exercise Nivan S. Weerakkody, Nicholas P. Whitehead, Benedict J. Canny, John E. Gregory, and Uwe Proske Abstract: Muscles subjected to eccentric exercise, in which the contracting muscle is forcibly lengthened, become sore the next day (delayed onset muscle soreness). In subjects who had their triceps surae of 1 leg exercised eccentrically by walking backwards on an inclined moving treadmill, mapping the muscle 48 hours later with a calibrated probe showed sensitive areas were localized but not restricted to the muscle-tendon junction. Injection of 5% sodium chloride into a sensitive site in the exercised leg did not produce more pain than injections into the unexercised leg, suggesting that nociceptor sensitization was not responsible. Applying controlled indentations to a sensitive area showed that the pain could be exacerbated by 20-Hz or 80-Hz vibration. In an unexercised muscle, vibration had the opposite effect; it reduced pain. Pain thresholds were measured before, during, and after a pressure block of the sciatic nerve. The block affected only large-diameter nerve fibers, as evidenced by disappearance of the H reflex and a weakened voluntary contraction, leaving painful heat and cold sensations unaltered. Pain thresholds increased significantly during the block. It is concluded that muscle mechanoreceptors, including muscle spindles, contribute to the soreness after eccentric exercise. © 2001 by the American Pain Society Key words: Exercise, pain, contraction, vibration, allodynia, muscle spindles.
W
e have all had the experience a day after unaccustomed exercise of lying in bed the next morning feeling fine, but the moment we try to get up we realize that we are stiff and sore. This is the soreness that results from certain forms of exercise and is called delayed onset muscle soreness (DOMS).1 Not all forms of exercise bring about DOMS. Specifically, it is eccentric exercise, in which the contracting muscle is forcibly lengthened, that is associated with DOMS.2 In everyday life we use concentric contractions, in which the contracting muscle shortens, to initiate movements. We perform eccentric contractions whenever we use our muscles as brakes—to slow down a movement. Sports that have substantial components of eccentric exercise include downhill walking, skiing, and horse riding. Sports with a concentric exercise bias, such as cycling and swimming, are not usually accompanied by DOMS.3 The muscle soreness that we experience after eccentric exercise is characterized by sensations of pain during muscle stretch or contraction. Pressing on the affected muscles or palpating them causes discomfort. But importantly, provided we remain still and relaxed, Received June 16, 2000; Revised October 23, 2000; Accepted November 30, 2000. From the Department of Physiology, Monash University, Clayton, Victoria, Australia. Address reprint requests to Uwe Proske, DSc, Department of Physiology, Monash University, Clayton, Victoria 3800, Australia. E-mail:
[email protected] © 2001 by the American Pain Society 1526-5900/01/0204-0001$35.00/0 doi:10.1054/jpai.2001.22496
we do not have any persisting soreness. Thus, the soreness is evoked by stimuli that do not normally produce pain, and there is no persisting pain. After injury to the skin, there is a region immediately surrounding the injury site, referred to as the region of primary hyperalgesia, in which low threshold mechanical stimuli evoke pain. In this region there is also a heightened sensitivity to painful mechanical and heat stimuli (hyperalgesia). It is believed that these changes are the result of sensitization of nociceptors.4 A larger area surrounding the injury shows a sensitivity to nonnoxious mechanical stimuli and hyperalgesia but only a small change in heat pain threshold.5 This is referred to as the region of secondary hyperalgesia, which, it is believed, is generated by mechanisms within the central nervous system,6 and the afferent input includes large-fiber mechanoreceptors (allodynia). Because DOMS is a deep tissue pain, it is not known whether there are regions of primary and secondary hyperalgesia. It is, in fact, not known whether there is any hyperalgesia at all associated with DOMS. Given that there is no chronic pain, the mechanism for DOMS is likely to be different than for other kinds of injury. Although it remains a point of controversy, there is evidence in support of the idea that the primary event that ultimately leads to DOMS is mechanical.7 Stretch of the contracting muscle leads to an uneven distribution of the length change and overextension of some sarcomeres in muscle fibers into a region of no filament overlap. Repeated eccentric contractions produce sarcomere disruptions and membrane damage. The damage triggers a
The Journal of Pain, Vol 2, No. 4 (August), 2001; pp 209-219
209
210
Mechanism of Delayed Onset Muscle Soreness
local inflammatory response8 and invasion of the injured area by macrophages and monocytes.9 That, in turn, leads to breakdown of the tissue and local production of chemicals such as prostaglandins.8 These will sensitize nociceptors.10 Therefore, one plausible mechanism for DOMS is that it is the result of sensitization of nociceptors to the point where innocuous stimuli, such as stretch or contraction, are able to excite them. A rather different explanation is that, like the secondary hyperalgesia seen in the skin, the generation of DOMS is based, at least in part, on mechanisms operating within the central nervous system with input from large-fiber mechanoreceptors.6 Here we have sought evidence for receptor sensitization and tested the possibility that the primary endings of muscle spindles are involved in DOMS in experiments with muscle vibration and a large nerve fiber block.
Methods The experiments were carried out on young, healthy adults, 28 men and 7 women aged 18 to 28 years. Anyone undergoing regular training exercise or who had some physical disability was excluded from the study. The experiments were approved by the local human ethics committee and conformed to the ethical guidelines of the 1975 Declaration of Helsinki. Informed consent was obtained from each subject.
The Exercise The exercise was carried out by a total of 35 subjects. Of these, 13 subsequently participated in the nerve block experiment, 13 in the vibration experiment, 6 in the saline injection experiment, and 3 in the experiment with painful stimuli. Triceps surae of 1 limb was exercised eccentrically by asking subjects to step backwards on an inclined (13º), moving treadmill. The treadmill speed was adjusted so that subjects carried out about 30 steps per minute. Subjects were asked to step backwards by using a toe-to-heel action with their exercising leg. It meant that when the foot bore the weight of the body, the triceps underwent an eccentric contraction. The other leg was brought alongside and placed flat on the treadmill so that its triceps did not undergo an eccentric contraction. In fact, it contracted very little, and because it did not become sore, it could act as a control.11 The treadmill speed had been adjusted so that by this time the subject would be carried back to the top of the treadmill ready for the next step backwards. To ensure sufficient load on the exercising muscles, subjects carried a weight belt with 5 to 10 kg of weight. Details of the method have been described previously.11
Mapping Tender Areas For 7 of the 13 subjects used in this experiment, measurements were restricted to a single sore spot in the exercised muscle. Subjects were studied on the second day after the exercise, when DOMS reaches its peak.11 They were asked to lie prone on a mattress. In the
Figure 1. Distribution of areas of sensitivity to local pressure in 1 subject’s triceps surae of the right leg that had undergone eccentric exercise 48 hours earlier. The muscle surface has been subdivided into a 1.5 × 1.5 cm grid. Medial side to the left, lateral side to the right. Pain threshold was measured by applying a calibrated compression gauge with 1.5-cm diameter plunger in each square of the grid. Thresholds have been represented as circles, with the size of the circle inversely proportional to the measured threshold given in newtons at the top of the figure. Similar measurements were carried out on the left, unexercised leg and there all pain thresholds were greater than 35 N.
prone position the ankle was plantar flexed so that the mapping was performed on a rather short muscle. In the majority of experiments testing was carried out with a compression gauge equipped with a 2.5-cm diameter tip. In 2 experiments in which we wanted to map the tender areas in more detail, a 1.5-cm diameter probe tip was used (Fig 1). The curved surface of the muscle was represented in 2 dimensions as a rectangle subdivided into 1.5-cm squares. An outline of the sore calf was drawn and divided into a 1.5-cm or 2.5-cm square grid. Tenderness threshold was measured within each square made by the intersecting grid lines (Fig 1). The compression gauge was slowly pressed onto the muscle, and the gauge reading was recorded when the subject reported the onset of pain. Tenderness threshold values were obtained for all parts of the muscle, and a suitably placed site was identified where threshold was particularly low.
Saline Injection This experiment was carried out on a total of 6 subjects. A volume of 0.2-mL saline was injected into medial or lateral gastrocnemius of each leg. Subjects
ORIGINAL REPORT/Weerakkody et al
211
were asked to rate perceived soreness by using a Visual Analog Scale (VAS) in which they turned a dial with a range of 0 to 10 over 300º of rotation. The dial was the moving arm of a potentiometer in which output could be recorded by computer.12 All subjects received injections of 5% saline into muscles of both legs before the exercise. They then received injections of 0.9% and 5% saline into both legs 2 days after the exercise. The rationale was that a control was needed for possible mechanical effects from the volume of fluid injected into the tender muscle. Injections were given in a random order, and subjects were unaware of the composition of a particular solution. For 3 subjects, injections of 0.9% saline also were given before the exercise. When subjects were tested after the exercise, the exercised muscle was first mapped for tender areas and the saline injected into the center of a tender region. The equivalent spot on the other muscle was used for the control injections. Injections were made with a 25-gauge needle. The injection process was carried out in 2 stages. The needle was pushed through the skin to a depth of 2 to 3 cm and then held there for 10 to 20 seconds before saline was injected. It allowed subjects to distinguish between pain from the needle prick and pain from the injected saline.
matching of peak forces with and without vibration. Vibration amplitude lay in the range of 0.8 to 1.3 mm. Subjects were asked to rate pain on a scale of 5, with 0 being no pain at all and 5 being intolerably intense pain. They were asked to rate the pain in steps of 0.5. Trials in which the muscle was just pressed by the actuator were interspersed with trials that included vibration, each separated by a 1-minute interval. For each subject, 10 trials were recorded with vibration at 20 Hz and 10 without, followed by 20 trials with and without vibration at 80 Hz. This was a sufficient number for each subject, given the reproducibility of subjects’ scores. Detailed vibration measurements were made on 7 subjects. For an additional 6 subjects, measurements were made at the sore spot, defined here as having a threshold of less than 15 N to applied pressure, a less sore area (threshold, 20 to 30 N), and in a comparable area of the unexercised leg. The pressure applied by the actuator was adjusted so that pain ratings at the sore spot and at the other sites were roughly comparable. Subjects were asked to report the intensity of the pain with and without vibration, which in this case was at a frequency of 80 Hz only. Here a VAS was used in which subjects recorded the level of pain by turning a dial with a scale of 0 to 10.
Vibration Experiment
Skin Anesthesia
Subjects were asked to lie on their side so that the tender area was within reach of the probe tip of a moving coil electromagnetic actuator supplied with position feedback. The diameter of the probe tip was 1 cm. Inserted between the tip and the shaft of the actuator was a force transducer. In parallel with the shaft of the actuator was a displacement transducer. It meant that both the amplitude and the force of displacement applied to the leg could be measured. A gating circuit was used to control the duration of indentation. For each set of measurements, first the movement was applied, typically a 20-mm indentation at a rate of 30 mms–1. The subject was asked whether the prod was being applied at the center of the tender area and, if not, minor adjustments were made to the position of the leg and the placement of the actuator. When a satisfactory position had been obtained, the leg was held fixed in that position by supports placed on each side. The subject also was asked, from this point on, to move as little as possible. Several trial indentations of the muscle were carried out to ensure that reproducible pain ratings were obtained in each trial. Here it was important to leave sufficient time between trials, typically 1 minute, to prevent receptor desensitization. The amplitude of the prod, the force applied, and the amplitude of the superimposed vibration were altered slightly from subject to subject, depending on how tender they were. Forces applied with the actuator lay in the range of 15 to 25 N. The amplitude of the displacement was 20 to 24 mm when not accompanied by vibration or with 20-Hz vibration. When 80 Hz was used, amplitude had to be reduced to 14 to 18 mm to allow
In making measurements of muscle soreness, it was important to be sure that some of the sensation had not been referred from the overlying skin. Therefore, in 3 subjects sore areas were mapped, and then a region of overlying skin, 2 cm in diameter, was treated with local anesthetic cream (EMLA, lignocaine/prilocaine 5%; Astra Pharmaceuticals, Sydney, Australia) for 2 hours. After treatment, the region of affected skin had become insensitive to touch and pin-prick stimulation. Stronger mechanical stimulation evoked some sensation, but this seemed to be coming from skin immediately adjacent to the treated area, which became dimpled by the stimulus. Measurements of threshold for pain were made with the compression gauge before and after anesthesia of the skin.
Responses to Painful Stimuli To obtain an indication of whether DOMS was associated with a heightened sensitivity to painful stimuli, a mechanical stimulus was devised that was strong enough to produce mild discomfort in an unexercised muscle. The muscle belly was gripped with a force of 50 N between the ends of a pair of large calipers held together by a spring. The caliper ends were broad and blunt, covered with taped cloth to prevent any skin soreness. Each arm of the calipers had a skin contact area of 0.7 cm2. When the calipers were applied to skin overlying bony structures, no soreness resulted; when applied to gastrocnemius, they evoked a dull sensation of deep pain. The testing procedure was to press the calipers open and to place them so that the ends just touched the skin overlying the muscle. Then the spring was released, leading to rapid compression of the mus-
212 cle. As soon as subjects reported a pain rating, typically within 1 to 2 seconds, the calipers were reopened. Measurements of the intensity of the pain evoked by a muscle pinch were made in a sensitive area of the exercised muscle and in the corresponding area of the unexercised muscle of the control leg.
Nerve Block Experiment This was carried out on a total of 13 subjects and included a number of subjects on whom the vibration measurements had been made beforehand. Subjects were seated in a chair, and the foot on the exercised side was strapped to a footplate. The plate was bolted to a steel shaft coincident with the ankle joint, and the shaft had strain gauges cemented to it. This allowed measurements of torque about the ankle joint. A stimulating cathode was strapped to the popliteal fossa, and an anode plate coated with electrode jelly was taped to the top of the knee above the patella. A 0.2ms duration shock of 20 to 40 mA, delivered across the knee, stimulated the tibial nerve. Stimulus strength was adjusted to elicit a maximal H reflex, the monosynaptic reflex evoked by Ia afferents of muscle spindles. The size of the H reflex, measured as the electromyogram (EMG), was approximately half of the maximal motor response. Sometimes it was necessary to adjust the position of the stimulating cathode before a satisfactory response had been obtained. The cathode was firmly strapped in place with velcro straps once an optimal position had been reached. For details see Wood et al.13 Stimulating conditions were adjusted so that the H reflex was accompanied by a small direct (M) response. This provided a convenient control for electrode movement. The triceps surae EMG was recorded with AgAgCl adhesive electrodes (3M Red Dot, Borken, Germany). Reflexes were displayed on an oscilloscope and recorded on a Macintosh (Apple Computer, Cupertino, CA) computer and a Maclab 8s system running Chart software (ADInstruments, Castle Hill, NSW, Australia). A differential compression block of the sciatic nerve was achieved by placing a wooden bar 6 cm high and 2 cm wide under the thigh of the exercised leg, just distal to the ischial tuberosity.14 Subjects were asked to lean toward the blocking side so that their full weight bore down on the bar. Once the subject was settled and the wooden bar was in place, reflex amplitude remained reasonably constant for the period before onset of the block. It took about 20 minutes to achieve a complete block of the H reflex. Once the block was complete, the area of desensitized skin was mapped by asking subjects where they were able to detect a light-touch stimulus. Care was taken to measure latencies to warm and cold stimuli within the boundary of the anesthetized area, which, experience dictated, was always on the side of the leg overlying the tibialis anterior muscle. Skin of the inside of the leg and overlying triceps surae is supplied by the saphenous nerve, which, as part of the femoral nerve, had not been blocked.
Mechanism of Delayed Onset Muscle Soreness Pain threshold for a sore region of the muscle was measured with the compression gauge. Integrity of conduction in small afferents was checked by using warm and cold stimuli applied to the skin. Here it was assumed that the pressure block of the sciatic nerve acted preferentially on large nerve fibers, regardless of their skin or muscle origins.14 Latency to a painfully warm stimulus (50°C) was measured with a 2.5-cm diameter probe. Contact with the skin was indicated by a digital counter on the stimulator so that accurate latencies could be measured. Cold was measured with a stainless steel bar with circular 2.5-cm diameter tip that had been immersed in ice between tests. Control measurements of pain threshold, latency to painful heat, and cold sensation were made 3 to 4 times before the block was applied. Throughout the block, measurements were initially carried out at 5-minute intervals, and when the block began to take effect, measurements were continued every 2 minutes. During the period of the block, the subject was asked at regular intervals to try to contract their triceps muscle. Any active torque and the triceps EMG were recorded.
Data Analysis In each experiment, individual measurements from all subjects were pooled. Values are given as means ± standard errors of the means calculated from the pooled data, unless otherwise stated. Analysis of variance testing was used to determine significance levels. Where an analysis of variance was significant (P < .05), a least significant difference post hoc test was applied. In addition, a pooled t test was used to determine the significance level between pain thresholds before and after skin anesthesia. The analysis program used was Data Desk (Data Description Inc, Ithaca, NY).
Results In a preliminary experiment, we tested for nociceptor sensitization by measuring responses to locally injected sodium chloride. We then sought evidence for a contribution to DOMS by muscle spindles in 2 different ways. First, the intensity rating of the pain was measured in response to a mechanical prod applied with and without concomitant vibration of the muscle. Here the underlying idea was that nociceptors were unlikely to respond to the vibration whereas muscle spindles were known to be vibration sensitive. Second, the effect of local pressure on the perceived intensity of pain was measured before, during, and after a conduction block of large nerve fibers to triceps surae.
Mapping the Tender Areas Subjects lay prone on a mattress while the triceps muscle on the sore side was systematically examined and the sore areas were mapped with a compression gauge.3 In nonpainful regions of the exercised muscle and in all parts of the muscle of the other leg the gauge could be pushed in to a maximum extent (50 N force)
ORIGINAL REPORT/Weerakkody et al
213
without the stimulus becoming painful. From the distribution of pain thresholds, it was clear that some parts of the exercised muscle were very much more tender than others. Areas 3 cm apart could differ in thresholds by as much as 30 N. It suggested that the foci of damage underlying the soreness were discrete and separated by regions of uninjured muscle. All subjects tested showed discontinuous distributions of thresholds across the muscle. However, the precise distribution of thresholds was unique for each subject. In the example shown in Fig 1, there is a suggestion that lateral gastrocnemius had become more sore than medial gastrocnemius. This was not a consistent finding. There was no suggestion that one part of the muscle like, for example, the muscle-tendon region was more prone to soreness than another.3 It was our impression that the 2 gastrocnemius muscles were more sensitive to touch than the soleus, although most of the soleus is overlaid by the gastrocnemii and only a limited portion could be accessed directly through the skin.
Saline Injections Before and after the exercise, injection of 0.9% saline into both the exercised and control muscles produced very little soreness, giving a mean rating of 0.63 (±0.22). Injecting 5% saline before the exercise produced a mild, dull soreness that peaked in about 1 minute and had completely gone by 5 to 7 minutes (Fig 2). Mean soreness ratings for the 2 legs gave a value of 3.9 (±0.48). Injecting 5% saline into the sore region of the exercised muscle 48 hours after exercise gave a mean rating of 3.2 (±0.73). This compared with a mean value of 3.1 (±0.37) for the unexercised muscle. This difference was not significant. Although there was a small difference in the pain intensity evoked by 5% saline in muscles of both legs before and after the exercise, this difference was not significant. A point noted in passing was that some subjects claimed that the pricking sensation as the needle penetrated the skin was more intense at a sore spot in the exercised muscle. But all subjects were able to clearly distinguish between the needle prick and sensations arising from the injected saline 10 to 20 seconds later. For 2 subjects, pain ratings to 5% saline injections were measured after the skin had been rendered anesthetic by treatment with EMLA cream. Pain ratings with the skin anesthetized did not differ significantly from values when the skin was fully sentient.
Muscle Vibration In preliminary experiments, the exercised muscle surface was explored with a hand-held vibrator. It was apparent that at a sensitive spot the vibrator had to be pressed into the muscle much less than a nonvibrating probe for the subject to report pain. In other words, vibration by itself could initiate pain. Provided muscle length was kept short, vibration, like the measurements with a compression gauge, revealed discrete areas of tenderness surrounded by less sore areas. However, if
Figure 2. Pain ratings reported by subjects in response to injecting isotonic (0.9%) and hypertonic (5%) saline. Top panel, sample record of the output of the potentiometer as the subject turned the dial, subdivided into a scale of 0-10 representing no pain up to unbearable pain. Bottom two panels show mean (±SEM) pain ratings for the exercised leg and the unexercised leg before and after exercise. Both 5% and 0.9% saline were injected. Because in preliminary experiments 0.9% was not used before the exercise, mean values here are for only 3 subjects. All other means are for 6 subjects.
muscle length was increased by passive dorsiflexion of the foot, the area from which pain could be evoked by vibration increased dramatically. Presumably the longer, stiffer muscle fibers transmitted the vibratory stimulus more effectively. The point emphasized that whatever neural structures were being vibrated, they lay deep within the muscle and the stimulus was reaching them via vibrating muscle fibers. The possibility was considered that responses of vibration receptors, such as pacinian or paciniform corpuscles lying deep in the tissue and associated with bony structures,15 were responsible for DOMS. Vibrating bony structures adjacent to the muscle at the knee and ankle did not produce any pain, nor did vibration of the Achilles’ tendon.
214
Figure 3. Measurements of intensity of pain evoked from a sensitive area of the exercised muscle in response to compression of the muscle at that site with and without vibration. Top panel: upper trace, tension; lower trace, skin displacement. Although not shown to scale, each test was separated from the next by a 1-minute rest period. The records, from left to right, show length and tension during compression combined with 20-Hz vibration, compression without vibration, with vibration at 80 Hz, and a repeat without vibration. Notice that with 80-Hz vibration, the amplitude of displacement had to be reduced to allow matching of peak compression forces with and without vibration. Bottom panel: histogram distribution of mean pain ratings (±SEM) in response to compression of the muscle without vibration (Press 1, Press 2) and with 20-Hz or 80-Hz vibration (Press +20 Hz; Press + 80 Hz). A rating of 0 was no pain, a rating of 5 was intolerably intense pain. Subjects were encouraged to rate pain in steps of 0.5. Asterisks indicate significant differences between ratings measured with and without vibration.
Once a conveniently located site of tenderness had been identified, it was marked and a supported electromagnetic vibrator was brought up to the muscle so that its probe tip just contacted the skin. Subjects were asked to rate the degree of soreness in response to a prod of the muscle. Every second prod was accompanied by vibration of the probe tip at 20 Hz or 80 Hz (Fig 3). Subjects were quite consistent in their scoring. For example, for pressure without vibration, 1 subject reported scores of 2.5 to 3.5 (mean, 2.95 ± 0.09). For pressure with 80-Hz vibration, the range was 3.5 to 4.5 (mean, 3.85 ± 0.11). It meant that 10 trials for each series were sufficient without risking subjects losing their concentration.
Mechanism of Delayed Onset Muscle Soreness At a sore spot, the 7 subjects gave mean soreness ratings of 2.57 (±0.08), increasing to 2.92 (±0.08) with 20-Hz vibration and 2.65 (±0.09) increasing to 3.54 (±0.09) with 80-Hz vibration (Fig 3). The difference between the prod without vibration and with vibration for both 20 Hz and 80 Hz was significant (P < .05). Subjects consistently reported that with vibration the intensity of the perceived pain was greater than without vibration, especially at vibration onset, but that it then began to fade. For 6 subjects, measurements of soreness, with and without vibration, also were made in less sore regions adjacent to the sore spots and at comparable nonpainful sites on the unexercised leg. Sore spots were defined as having a compression threshold of less than 15 N, with less sore areas having a threshold of 20 to 30 N. Once these were located, the electromagnetic actuator was brought up to apply vibration. When moving away from a sore spot, the force of the actuator needed to be increased for subjects to be able to report levels of pain that were approximately comparable with those they had experienced at a sensitive spot. For the exercised leg it meant only a small increase in force, but for the unexercised leg force had to be nearly doubled (Fig 4). Measurements were made at a sore spot, an adjacent area where subjects declared that they were only about half as sore, and on the unexercised leg (Fig 4). Although for the exercised leg, vibration of a sore spot
Figure 4. Pooled data from 6 subjects experiencing DOMS in 1 leg showing mean (±SEM) soreness ratings and levels of applied force in response to compression with (open symbols) and without vibration (filled symbols). Measurements were made at 2 locations in the exercised muscle and at comparable places in the unexercised muscle. Circles, sore spot; squares, moderately sore spot, triangles, comparable area in unexercised muscle. Force values for a given level of soreness were lowest at a sensitive spot, and here vibration produced the largest increase in soreness. In the unexercised muscle vibration reduced the perceived level of soreness.
ORIGINAL REPORT/Weerakkody et al
215
or less sore areas increased perceived soreness, on the unexercised side it reduced it.
Skin Anesthesia It was necessary to be sure that subjects’ reported sensations were due to stimulating the muscle and not the overlying skin. In practice, subjects insisted that the pain was deep and was coming from the muscle. Nevertheless, to control for any skin sensations or skinmediated reflex effects, in 3 subjects the skin overlying the stimulated area was treated for 2 hours before the compression measurements with anesthetic cream. By this time the region of treated skin had become insensitive to tactile and pin-prick stimulation. Values for pain thresholds with skin anesthesia (12.2 ± 1.4 N) were not significantly different from values when skin sensation was intact (9.7 ± 1.3 N).
Responses to Painful Stimuli The following question was posed: Was DOMS associated with any hyperalgesia? To test this idea, mildly painful stimuli were applied to a tender area of the exercised muscle to determine whether subjects reported a heightened sensation of pain. Stimuli were generated by gripping the muscle between the blunt ends of a pair of calipers that exerted a force of 50 N generated by a tension spring. Care was taken to avoid cutaneous nociceptive input by wrapping the caliper ends in soft cloth. Pain rating for the unexercised control leg of 3 subjects was in the range of 2.0 to 2.8 on a scale of 0 to 10 with a mean of 2.66 ± 0.16. For an equivalent region of the exercised muscle 2 days after exercise, values were in the range of 7.0 to 8.0 with a mean of 7.29 ± 0.25. This difference was significant (P < .05). In addition, subjects reported that the rate of onset of pain seemed to be more rapid when the calipers were applied to the exercised muscle.
Nerve Block This experiment was carried out on the same 7 subjects after their vibration responses had been measured. Further observations were made on an additional 6 subjects. For 11 of the 13 subjects an effective block of large afferents was achieved, whereas for 2 subjects conduction was not blocked, presumably because the wooden bar had not been placed correctly. The sensitive spot used for the vibration measurements also was used to test the effect of nerve block, provided it was conveniently placed. In the tender area, threshold values to muscle compression before the block were in the range of 15 to 20 N. Thresholds in unaffected parts of the muscle were greater than 50 N. Data from 1 experiment are shown in Fig 5. The amplitude of the H reflex began to decrease about 13 minutes after the onset of pressure on the sciatic nerve, and the reflex was fully blocked within 3 minutes. It remained blocked for an additional 15 minutes, and when the wooden bar was removed from under the thigh, the reflex recovered very rapidly within 1 to 2 trials. Subjects
Figure 5. Examples of measurements of tenderness threshold for 1 subject before, during, and after a pressure block of the sciatic nerve. Top panel shows the size of the H (Hoffman) reflex, measured as mV of recorded surface EMG from triceps surae during the 40 minutes of recording before, during, and after the block. Middle panel shows the tenderness threshold in response to compression of a sensitive region of the muscle exercised 48 hours earlier. Bottom panel shows the latency of pain to a hot (50°C) probe applied to the region of skin adjacent to the muscle and innervated by a branch of the sciatic nerve. Dashed horizontal lines indicate mean control thresholds and latencies; vertical dotted lines, the duration of the block. Notice that the H reflex disappears within 2-3 trials and after removal of the block it recovers rapidly. Changes in tenderness ratings are more gradual. There was no change in painful heat latency.
reported that recovery from block was accompanied by some paresthesia in the lower leg. During the block, there was an increase in pain threshold to a mechanical prod, with values increasing from about 17 to 30 N, whereas latency for painful heat sensation and the sense of cold remained unchanged. There was, in fact, a small increase in latency for warm stimuli in 4 subjects, but for the pooled values from the 11 subjects, this was found not to be significant (Fig 6). By contrast, the pooled data showed a significant increase in pain threshold from a control value of 16.2 N (±0.6 N) to 27.4 N (±1.6 N) during the block (P < .05) and recovery of normal thresholds (18.1 ± 0.6 N) after the block (Fig 6).
216
Mechanism of Delayed Onset Muscle Soreness stimulation. The regions of soreness are not distributed uniformly throughout the muscle; rather, there are areas of heightened sensitivity that can be found in all parts of the muscle (Fig 1).3 The central question was whether DOMS resulted from a heightened sensitivity of muscle nociceptors to mechanical stimuli or whether large-diameter afferents were involved, as occurs in the region of secondary hyperalgesia surrounding a cutaneous injury site.16
Saline Injections
Figure 6. Pooled measurements from 11 subjects of H-reflex amplitude (top panel), tenderness threshold (middle panel), and latency to painful heat (bottom panel) before, during, and after block of the H reflex. All values are given as means (± SEM) averaged over the period of recording, typically 20 minutes before onset of the block, 10 minutes during the block, and 10 minutes after recovery from the block.
Throughout the experiment each subject was asked to attempt to contract their triceps. Torque and EMG recordings both showed reductions in amplitude during the block, suggesting conduction failure in some, but not all, motoneurons during the period of H-reflex block. This supported the view that only large myelinated fibers had been blocked because some skeletal motor neurons supplying triceps were still conducting through the region of the block. A similar result was obtained with all but 2 of the 11 subjects. In these 2 subjects, there was a complete motor nerve block.
Discussion The results of these experiments are presented in support of the view that muscle mechanoreceptors served by large diameter, group I nerve fibers play a role in DOMS. DOMS is characterized by a mild soreness to mechanical stimuli such as stretch, contraction, or palpation. Pain onset is prompt with no persistence at the end of
It has been known for a long time that muscle receptors respond to hypertonic sodium chloride. 17 Experiments on the gastrocnemius muscle of the dog18 indicated that close arterial injection or topical application of sodium chloride excited 89% of units with Aδ and C fiber conduction velocities, presumed to be polymodal nociceptors. Only 19% of stretch receptor afferents responded. The findings suggested some degree of selectivity in the action of sodium chloride. Although conduction velocity of the stretch receptor afferents was not given, responses from muscle spindle and tendon organs would be expected to be limited because of the presence of a capsule surrounding the sensory endings that would act as a diffusion barrier. Injection of hypertonic saline is a convenient method of inducing controllable levels of muscle pain in humans.19 Recent work has suggested that saline produces its effects by increasing extracellular sodium concentrations leading to depolarization of excitable membranes.20 In our experiments, soreness evoked by hypertonic saline was not significantly different in muscles of subjects experiencing DOMS when compared with unexercised muscles (Fig 2). It means that our observations are not consistent with the nociceptor sensitization hypothesis. If nociceptors were in a sensitized state, they might have been expected to produce a larger response to the saline stimulus to produce more pain. It might be argued that, in our subjects, nociceptors were in a temporarily desensitized state as a result of generating DOMS-related activity. However, that does not fit the observations. There was no evidence of systematic changes in measured pain thresholds to local muscle pressure throughout the half hour or so during which subjects lay prone while test solutions were injected into their muscles. We conclude that the saline injection experiment does not support the nociceptor sensitization hypothesis. To shore up this idea, we plan to test the effects of vibration on saline-evoked soreness. Our prediction is that vibration will not exacerbate the pain.
Vibration Responses Support for the involvement of large-fiber mechanoreceptors in DOMS came from the finding that pressing on the muscle in a tender area produced mild soreness that became more intense during vibration (Fig 3). In an unexercised muscle, any pain from strong
ORIGINAL REPORT/Weerakkody et al pressure was reduced by vibration, an exactly opposite effect (Fig 4). The primary endings of muscle spindles are likely to be the only mechanoreceptors in a passive muscle with a sustained response to vibration at 80 Hz. The secondary endings of spindles do not respond much above 30 Hz,21,22 whereas tendon organs remain unresponsive at all frequencies unless the muscle is contracting.21 That leaves only the occasional paciniform corpuscle, served by a group III axon. In our view, there are too few of these in the muscle to be able to account for DOMS.23 Pacinian corpuscles associated with structures such as the interosseous membrane15,23 do not seem to be involved because vibration of bony structures in the lower leg evoked no sensations of pain. Applying 20-Hz vibration to the muscle is likely to excite secondary endings of spindles.24 We cannot exclude the possibility of a contribution to DOMS from secondary endings because a small, but significant, increase in pain rating was reported by subjects during 20-Hz vibration (Fig 3). More importantly, the further increase in perceived pain when vibration was increased from 20 to 80 Hz specifically points to the involvement of the primary endings of muscle spindles. Interestingly, several attempts at evoking pain by vibrating the Achilles tendon in subjects experiencing DOMS were unsuccessful, presumably because spindles were not excited sufficiently strongly.
Other Receptors It is not known whether sensitized muscle nociceptors are able to respond to vibration, although they are known to respond to non-noxious mechanical stimulation.25 Receptors served by unmyelinated axons are unlikely to show vibration responses.26 That leaves group III nociceptors as possible candidates. As well as group III nociceptors, skeletal muscle contains a group of mechanoreceptors served by group III axons that are not stretch sensitive but respond to local pressure.18,27,28 It is not known whether these receptors respond to vibration. It is conceivable that they too are involved in the generation of DOMS. However, DOMS can be evoked by muscle stretch so that they would have to alter their response properties as part of the process leading to development of DOMS. There are some observations from animal experiments that suggest that the responses of sensitized group III nociceptors are not consistent with the characteristics of DOMS.29 The compound carrageenan was injected into cat hindlimb muscles. Within 1 to 2 hours this evoked a powerful inflammatory response in the muscle associated with the release of 5-hydroxytryptamine, histamine, bradykinin, and prostaglandins. These substances stimulate or sensitize muscle nociceptors. The main effect of the inflammation on muscle group III nociceptors was that they developed resting activity. Interestingly, their mechanical thresholds did not
217 change. The finding led the investigators to conclude that this group of receptors might be associated with the spontaneous pain and dysesthesias associated with myositis, an inflammatory disease of the muscle, at least to the extent that carrageenan-evoked inflammation is a realistic model of the disease. The symptoms associated with myositis are quite unlike those of DOMS. With DOMS there is no pain in the resting subject. Pain is evoked only from muscle stretch, contraction, or palpation.
Nerve Block The H reflex is the electrically evoked monosynaptic reflex following the pathway of the tendon jerk reflex. The reflex is thought to be elicited largely by group I fibers. In these experiments, block of the reflex by nerve compression led to a substantial increase in pain threshold at a DOMS site in the muscle in response to slowly applied pressure. Often the increase in threshold began at a time when subjects were still able to contract the muscle voluntarily, suggesting that α motoneurons were still conducting impulses although, typically, the contraction was weaker, suggesting some blocked fibers. It makes it unlikely that at this stage substantial numbers of small-diameter nerve fibers were blocked. In support of that view was the finding that the latencies of the senses of cold (Aδ fibers) and painful heat (Aδ and C fibers) remained unchanged during the block. Significantly, however, subjects still seemed to report a distinct pain threshold during the block, although several said that they were not sure whether the discomfort was any greater than that experienced from pressing the probe into an unaffected part of the muscle. They declared that the sensations had become rather vague. This point needs to be explored in future experiments, but the balance of our data suggests that some component of DOMS persists after a group I fiber block. If correct, it would imply that DOMS is generated as a result of combined inputs from nociceptors and large mechanoreceptor afferents. It could be argued that at the point of disappearance of the reflex, the group I fibers had not been blocked, their conduction just slowed in the region of the block, leading to dispersion of impulses so that motoneurons were no longer excited sufficiently synchronously to bring them to firing threshold. However, evidence of a partial large fiber block was indicated by the weakened voluntary contraction. The simplest interpretation of the accompanying increase in pain threshold was that it was due to withdrawal of input from large afferents. There is some support from the work of others for our conclusion about the role of large muscle afferents in DOMS. Barlas et al30 showed that there was an increase in mechanical pain threshold in elbow flexor muscles with DOMS after application of an ischemic block.
Mechanism How are the afferents of muscle spindles able to access the pain pathway? It has been known for some
218
Mechanism of Delayed Onset Muscle Soreness
time that in superficial layers of the dorsal horn of the spinal cord there are relay cells in the principal pain pathway, the spinothalamic tract, that receive nociceptive inputs from muscle.31 A proportion of these cells (20%) had thresholds in the innocuous range so they could respond to inputs from both nociceptive and non-nociceptive receptors. Cells responding to a range of inputs have been called wide dynamic range (WDR) cells. Therefore, one mechanism by which spindle afferents could access the pain pathway is by acting through the WDR cells. Interestingly, studies of cells in the dorsal horn are confounded by the presence of many cells with proprioceptive inputs, cells that have been thought not to be involved in the processing of pain information but to be part of the Clarke’s column– dorsal spinocerebellar tract system.32 It would now be interesting to reinvestigate dorsal horn neurons, seeking evidence specifically for cells that received both proprioceptive and noxious inputs before and after a period of eccentric exercise. It has recently been proposed that the nociceptor sensitization process associated with tissue injury leads to a raised excitability in postulated presynaptic inhibitory interneurons between large-diameter mechanoreceptor afferents and nociceptor afferents.33 These interneurons could include WDR cells. In normal, uninjured muscle, input from large afferents inhibits nociceptive input by presynaptic inhibition, the basis of the “gate control” theory of pain.34 This mechanism would account for our finding that in unexercised muscles, painfully strong pressure becomes less painful during vibration (Fig 4). In the excitable state triggered by muscle injury, the inhibitory interneurons in response to mechanoreceptor input are able to generate a dorsal root reflex in nociceptor afferents. That, in turn, leads to generation of pain. Our own data provide no information on which of the 2 mechanisms might be involved in DOMS. We do not even know
what the trigger for DOMS is. Presumably at some point there is sufficient nociceptive input produced by the muscle injury to initiate the central sensitization process. An alternative or perhaps supplementary hypothesis for a mechanism for DOMS is that the muscle damage from eccentric exercise triggers an inflammatory response, and the release of substances such as prostaglandins sensitizes nociceptors to the point at which they respond to non-noxious mechanical stimulation but without generating chronic pain. However, administration before and after exercise of nonsteroidal anti-inflammatory drugs that block the enzyme cyclooxygenase, which produces prostaglandins, does not lead to significant reductions in DOMS after eccentric exercise.35,36 It suggests that if a sensitization process is involved in DOMS, its pharmacology is different from that of sensitized cutaneous nociceptors. Our experiments on saline-evoked muscle pain also support the view that a simple sensitization process is not responsible for DOMS.
References
on the nature of cutaneous hyperalgesia. J Clin Invest 29:115-140, 1950
1. Armstrong RB, Warren GL, Warren JA: Mechanisms of exercise-induced muscle fibre injury. Sports Med 12:184207, 1991
7. Morgan DL, Allen DG: Early events in stretch induced muscle damage. J Appl Physiol 87:2007-2015, 1999
2. Jones DA, Round JM: Skeletal Muscle in Health and Disease. Manchester, NY, Manchester University Press, 1990, chap 10
8. Smith LL: Acute inflammation: The underlying mechanism in delayed onset muscle soreness? Med Sci Sports Exerc 23:542-551, 1991
3. Newham DJ, Mills KR, Quigley BM, Edwards RHT: Muscle pain and fatigue after concentric and eccentric muscle contractions. Clin Sci 64:55-62, 1983
9. Stauber WT, Fritz VK, Vogelbach DW, Dahlmann B: Characterisation of muscles injured by forced lengthening 1. Cellular infiltrates. Med Sci Sports Exerc 20:345-353, 1988
4. Treede RD, Meyer RA, Raja SN, Campbell JN: Peripheral and central mechanisms of cutaneous hyperalgesia. Prog Neurobiol 38:397-421, 1992
10. Mense S: Nociceptors in skeletal muscle and their reaction to pathological tissue changes, in Belmonte C, Cervero F (eds): Neurobiology of Nociceptors. Oxford, UK, 1996, chap 7
5. Kilo S, Schmilz M, Koltzenzurg M, Handwerker HO: Different patterns of hyperalgesia induced by experimental inflammation in human skin. Brain 117:385-396, 1994 6. Hardy JD, Wolff HG, Goodell H: Experimental evidence
Significance What is the significance of DOMS? Presumably the individual will try to minimize the pain by limiting use of the affected muscles during the period of repair after damage from eccentric exercise. It is known that within a week of a period of unaccustomed eccentric exercise and the accompanying muscle soreness, there is a remodeling of the muscle so that a second period of exercise leads to much less soreness.11 Nevertheless, DOMS can have a seriously debilitating effect on athletes’ performance. It means training programs must be devised to minimize DOMS by subjecting the individual to regular, mild eccentric exercise. The insight into the mechanism of DOMS put forward here may help to generate new therapeutic strategies to minimize the soreness whenever it arises.
11. Jones C, Allen TJ, Talbot J, Morgan DL, Proske U: Changes in mechanical properties of human and amphibian muscle after eccentric exercise. Eur J Appl Physiol Occup Physiol 76:21-31, 1997
ORIGINAL REPORT/Weerakkody et al 12. Graven-Nielsen T, Arendt-Nielsen L, Svensson P, Staehelin-Jensen T: Stimulus-response functions in areas with experimentally induced referred muscle pain: A psychophysical study. Brain Res 744:121-128, 1997 13. Wood SA, Gregory JE, Proske U: The influence of muscle spindle discharge on the human H reflex and the monosynaptic reflex in the cat. J Physiol 497:279-290, 1996 14. Garland SJ: Role of small diameter afferents in reflex inhibition during human muscle fatigue. J Physiol 435:547558, 1991 15. Hunt CC, McIntyre AK: Characteristics of responses from receptors from the flexor longus digitorum muscle and the adjoining interosseous region of the cat. J Physiol 153:74-87, 1960 16. Torebjörk HE, Lundberg LER, LaMotte RH: Central changes in processing of mechanoreceptive input in capsaicin-induced secondary hyperalgesia in humans. J Physiol 448:765-780, 1992 17. Iggo A: Non-myelinated afferent fibres from mammalian skeletal muscle. J Physiol 155:52-53P, 1961 18. Kumazawa T, Mizumura K: Thin-fibre receptors responding to mechanical, chemical and thermal stimulation in the skeletal muscle of the dog. J Physiol 273:179194, 1977 19. Kellgren JH: Observations on referred pain arising from muscle. Clin Sci 3:175-190, 1938 20. Graven-Nielsen T, McArdle A, Phoenix J, Arendt Nielsen L, Jensen TS, Jackson MJ, Edwards RHT: In vivo model of muscle pain: Quantification of intramuscular chemical, electrical and pressure changes associated with salineinduced muscle pain in humans. Pain 69:137-143, 1997 21. Brown MC, Engberg I, Matthews PBC: The relative sensitivity to vibration of muscle receptors of the cat. J Physiol 192:773-800, 1997 22. Hagbarth KE, Eklund G: Motor effects of vibratory muscle stimuli in man, in Granit R (ed): Nobel Symposium I Muscular Afferents and Motor Control. Stockholm, Sweden, Almqvist & Wiksell, 1966 23. Barker D: The structure and distribution of muscle receptors, in Barker D (ed): Symposium of Muscle Receptors. Hong Kong, University Press, 1962, pp 227-240
219 24. Bianconi R, van der Meulen JP: The response to vibration of the end organs of mammalian muscle spindles. J Neurophysiol 26: 177-190, 1963 25. Mense S, Meyer H: Bradykinin-induced modulation of the response behaviour of different types of feline group III and IV muscle receptors. J Physiol 398:49-63, 1988 26. Bessou P, Burgess PR, Perl ER, Taylor CB: Dynamic properties of mechanoreceptors with unmyelinated (C) fibres. J Neurophysiol 34:116-131, 1971 27. Bessou P, Laporte Y: Étude des recepteurs musculaires innervé par les fibres afférentes du Groupe III (fibres myelinisées fines) chez le chat. Arch Ital Biol 99:270-292, 1961 28. Paintal AS: Functional analysis of Group III afferent fibres of mammalian muscle. J Physiol 152:250-270, 1960 29. Berberich P, Hoheisel U, Mense S: Effects of a carrageenan-induced myositis on the discharge properties of group III and IV muscle receptors in the cat. J Neurophysiol 59:1395-1409, 1988 30. Barlas P, Walsh DM, Baxter GD, Allen JM: Delayed onset muscle soreness: Effect of an ischaemic block upon mechanical allodynia in humans. Pain 87:221-225, 2000 31. Craig AD, Kniffki K-D: Spinothalamic lumbosacral lamina I cells responsive to skin and muscle stimulation in the cat. J Physiol 365:197-221, 1985 32. Mense S: Nociception from skeletal muscle in relation to clinical muscle pain. Pain 54:241-289, 1993 33. Cervero F, Laird JMA: Mechanisms of touch-evoked pain (allodynia): A new model. Pain 68:13-23, 1996 34. Melzack R, Wall PD: Pain mechanisms: A new theory. Science 150:971-979, 1965 35. Bourgeois J, MacDougall D, MacDonald J, Tarnopolsky M: Naproxen does not alter indicies of muscle damage in resistance-exercise trained men. Med Sci Sports Exerc 31:49, 1999 36. Croisier JL, Camus G, Monfils T, Debydupon G, Fafchamps F, Venneman I, Crielaard JM, Juchmesfrir A, Lhermerout C, Lamy M, Deby C: Piroxicam fails to reduce myocellular enzyme leakage and delayed onset muscle soreness induced by isokinetic eccentric exercise. Mediators Inflamm 5:230-254, 1996
