| | Pain Perception After Running a 100-Mile UltramarathonAbstract Hoffman MD, Lee J, Zhao H, Tsodikov A. Pain perception after running a 100-mile ultramarathon. ObjectiveTo determine if pain perception is affected by an extreme bout of exercise that causes ongoing exercise-related pain. DesignRepeated-measures design. SettingPre-race registration area and finish area of an endurance race. ParticipantsTwenty-one competitors in the 2005 Western States 100 Mile Endurance Run and 11 control subjects who were assisting at the race but not running. InterventionsNot applicable. Main Outcome MeasuresOverall pain and pain ratings on a pressure pain test before and after the event. ResultsMean overall pain ± standard deviation on a 100-mm scale increased (P<.05) from 3±6mm before the run to 39±28mm after the run among the runners. The faster runners showed a mean reduction (P<.05) in pain ratings after the race of 15±20mm (on a 100-mm scale), whereas there was no change for the slower runners and controls. Findings were confirmed by model-based analysis. ConclusionsThe faster runners in a 100-mile (161-km) running race experience a modest temporary reduction in pressure pain perception that does not appear to be augmented by ongoing pain related to the exercise. The lack of a reduction in pain perception among the slower runners may be because an extreme bout of exercise of this nature can “exhaust” the systems responsible for exercise-induced analgesia in all but the most well-trained of runners, or that these systems were not activated because the slower runners were unable to maintain a high enough exercise intensity during the later stages of the race. PREVIOUS RESEARCH HAS shown that aerobic exercise can induce an acute analgesic effect.1, 2 However, the duration and intensity of exercise required to elicit such an effect has not been fully clarified. Exercise-induced analgesia has been shown to occur from fairly high intensities for durations of as little as 8 to 10 minutes3, 4, 5 and up to 40 to 50 minutes,6, 7 but longer bouts of exercise at low intensities have not been studied. Some work has suggested that intensities of over 70% of maximal aerobic capacity are required, and that pain threshold increases with increasing intensities above this level.3, 4, 5 Our recent work8 showed that aerobic exercise of 30 minutes at 75% of maximal oxygen uptake (V̇o2max) resulted in a significant exercise analgesia effect, but 30 minutes of exercise at 50% of V̇o2max or 10 minutes at 75% of V̇o2max were inadequate to show the effect. This suggests that there may be thresholds for both intensity (>50%−70% of V̇o2max) and duration (>8−10min) required for exercise-induced analgesia. As far as we know, none of the studies examining postexercise analgesia have quantified the extent of naturally occurring pain as a result of the exercise. As such, no attention has been directed at how naturally occurring pain from the exercise might relate to the exercise-induced analgesic effect. An extreme bout of exercise, such as an ultramarathon competition, provides a venue to study pain perception in the presence of ongoing exercise-related pain. During and after such arduous exercise, some naturally occurring pain has been shown to be present.9 Whether the pain caused by this type of exercise would augment or reduce the usual exercise-induced analgesia that is observed after exercise is unknown. The purpose of the present study was to determine whether pain perception is altered among competitors after completing a 100-mile (161-km) running race. Given the previous findings that exercise intensities of greater than 50% to 70% of V̇o2max were required to elicit exercise-induced analgesia and the recognition that this may be close to the intensity that is sustained for a 161-km run,10, 11 we theorized that only the most trained runners, who were able to maintain a higher intensity throughout the race, would show exercise-induced analgesia. We also hypothesized that exercise-induced analgesia among the faster runners would be augmented by the presence of several hours of naturally occurring pain from the exercise. Methods  Participants We performed the study at the Western States 100 Mile Endurance Run (June 25−26, 2005), a point-to-point trail run through the Sierra Nevada Mountains in Northern California. With over 5500m of ascent and 6700m of descent, this 161-km race is considered to be among the most arduous organized running events in the United States. The 2005 race was particularly challenging due to sections of snow cover on the trail for nearly the first 30km. A 30-hour time limit is allowed for completion of the race. Two groups of subjects participated in this study. Runners were recruited from the group of competitors in the race. Several weeks prior to the race, information about the study was sent to all entries. Some runners agreed to participate in the study in advance and others were recruited during race registration 1 to 2 days before the race. Control subjects were recruited among the crew members and race assistants who would be available 1 to 2 days before the race and again at the race finish, and had also indicated they would remain awake during the event. Exclusion criteria for participation included the presence of any chronic pain condition, diseases affecting sensory nerves, pregnancy, the use of narcotic analgesics, and known cardiac, pulmonary, or metabolic disorders. Procedures were approved by the institutional review board. All subjects were provided general information about the intent of the study in order to obtain informed consent, but the hypothesized results were not discussed in detail prior to completion of their participation in the testing. Experimental Design All subjects underwent 2 pain tests during the 2 days prior to the race. These 2 tests were separated by at least 30 minutes. The first test was considered to be for practice in order to allow the subjects to become familiar with the procedures. The second test was used in the data analysis. After completing the race, the runners were directed to an area near the finish line where they underwent a third pain test. Given that our previous work has suggested that exercise-induced analgesia may persist for less than 30 minutes under some conditions,8 our goal was to perform the post-race test within 10 to 30 minutes after completion of the race. Control subjects who were assisting a runner in the study underwent the third pain test shortly before or after the runner finished, and those control subjects who were assisting with the race underwent the third test when they were available during the span of nearly 14 hours over which the runners were finishing. Prior to the pain tests, we asked the subjects to list any pain medications taken in the previous 24 hours and to indicate their overall pain level by placing a mark on a 100-mm visual analog scale (VAS) anchored with “no pain” and “worst possible pain.” Pressure Pain Testing We performed the pain tests with a pressure pain stimulator similar to that previously used by others12, 13, 14 and the same device as in our previous work.8, 15 The device uses a 6×0.25mm Lucite edge through which a constant force of 9.8N is applied against the dorsal surface of the middle phalanx of the nondominant index finger, halfway between the distal and proximal interphalangeal joints. Subjects listened to a recorded message reviewing the procedures prior to each test. At 10-second intervals the audio recording instructed subjects to indicate the level of perceived pain by placing a mark on separate 100-mm VASs anchored with “no pain” on the left side and “worst possible pain” on the right side. Each pain stimulus lasted 2 minutes. During testing, the subjects were seated with arms supported on a table. Good reproducibility of these techniques with repeat testing at a 15-minute interval, as well as across days, has been previously shown.8, 15 Statistical Methodology The main focus of the study was to examine for a change in pain perception as a result of running the race. This was a longitudinal study with 2 time points at which pain ratings were measured (before and after the race). Each pain test was also longitudinal in nature. However, the within-test pain ratings over time were not modeled, and were represented as a single summary (average) value. Each mark on the VAS was converted to a numeric value by measuring the distance in millimeters from the left side of the scale to the mark. Pain ratings were relatively stable during the last minute of the pain test, so the last 6 data points (ie, the final minute) were averaged to yield a single value for each trial. Repeated before-after measurements induce correlation because they are shared by the same subject. This was taken into account using linear random-effects models for the analysis.16 The models allowed for fixed effects of explanatory variables as well as a random normally distributed subject-specific intercept term. Main effects as well as interactions were examined in search of the best model. The following variables were included in the model selection procedure: time, a binary variable modeling the effect of the race on within-subject outcome score; group, modeling the effect in runners compared with controls; a cluster of anthropomorphic and other subject characteristics including age (continuous), sex (binary), height (continuous), body mass (continuous), weekly running distance, training volume during the previous month (a measure of subject’s physical activity level), finish time (continuous and categorical dichotomized at the median) defined in runners and modeling the performance in the race, and delay (continuous) modeling the effect of delay in pain testing after the race. The search for the best model was not automatic. It involved testing key meaningful hypotheses based on a hierarchical family of models and the likelihood ratio test presented in table 1. A number of those hypotheses involved interaction effects targeting analgesic response operating differently in specific subgroups of subjects defined by the main variables mentioned above. These included interaction of group with time, which allowed us to study the effect of race separately in runners and controls, and time with finish time modeling the effect of race separately in slower and faster runners. | ⁎ P value for the hypothesis. |
Observed versus predicted residual plots (not shown) were used to confirm model adequacy. Data Analysis There were 25 runners and 13 control subjects who agreed to participate in the study. Twenty-one of the runners completed the study, because 2 did not finish the race, one was in the medical tent for an extended period after completing the race, and one forgot to return for the pain test after the race. Eleven of the controls completed the study because 2 forgot to present for the final pain test. The time of day for the post-race pain test was distributed from approximately 5:00 to 11:00 am for the controls, and approximately 12:30 to 11:30 am for the runners. Two controls reported having obtained some sleep during the night. We initially used descriptive statistics to summarize data by marginal means before and after the race, standard errors (SEs), and paired t tests targeting mean within subject change in pain test scores from before to after the race. Runners were split into 2 equally sized groups (slower, faster) by using a cutpoint at the median of finish time of 25.55 hours. A histogram of finish time is shown in figure 1. The results shown in table 2 strongly suggest that the analgesic effect of the race predominantly manifests itself in the faster runners. The descriptive observation in table 2 was then confirmed in rigorous model-based analyses. A key sequence of models considered in the search and associated hypotheses is given in table 1. The search for the best model started with a model having 3 fixed effects modeled by categorical variables: time (post-race vs pre-race); group (runners vs controls), and time by group (interaction term modifying post- vs pre-race difference in runners as compared with the controls). The above variables directly pertain to the effect of exercise on analgesia, the main hypothesis of this study. A round of forward selection attempted to include age (continuous), sex (categorical), height (continuous), body mass (continuous), weekly running distance (continuous), and finish time (continuous) in the initial time plus group plus time by group model. Likelihood ratio test was not significant and Akaike information criterion was worse than the starting model for all attempts. Having completed the above exploratory stage of best model selection, we refined the model to allow fine measurement of performance level by introducing finish time as a modifier for the effect of the race. Because continuous finish time showed no significance, we treated finish time as a categorical variable in the subsequent analysis. We were testing whether a subgroup of subjects with better performances showed a significant effect of the race. The target group was defined based on the finish time variable dichotomized using a cutpoint at the median. When analyzing a model containing the effect of time in runners (a group by time interaction), runners showed a mean ± SE within-subject reduction of pain rating of 6.6±4.0mm. This effect showed a trend toward significance at P equal to .109 and we followed with a targeted analysis of pain rating effects in runners. Post hoc analysis using variable cutpoints revealed the lowest (unadjusted) P value of .008 at a cutpoint of 26 hours with the estimated effect ± SE of 13.8±4.9. Using this cutpoint, the mean finish times ± standard deviation (SD) were 22.8±2.1 and 28.5±1.4 hours for the faster and slower runners, respectively. Pain ratings among the faster runners were reduced from 52±20 to 37±18mm (fig 2). The ongoing exercise-related pain showed no significant correlation with analgesia in the runners (P=.50) when the overall pain level was added to the model of race-induced changes in pain ratings. The faster runners included 8 men and 4 women, and the slower runners included 8 men and 1 woman. Examination of some mutually related variables shed light on the difference in characteristics between the faster and slower runners (table 3). Faster runners tended to be younger (P=.053) and ran greater (P=.001) distances in training during the month before the race. At the same time, there were no differences in body mass index (BMI) (P=.080), pre-race pain ratings (52±20 vs 51±30, P=.92) (see fig 2), pre-race overall pain levels (1±3 vs 3±4, P=.30) (fig 3), post-race overall pain levels (39±28 vs 40±27, P=.94) (see fig 3), or time interval between finishing the race and performance of the last pain test (35±44 vs 17±7min, P=.25). This latter observation was confirmed in model-based analysis including the variable delay (see table 1). | ⁎ Statistical differences between groups were identified with unpaired t tests. |
The control subjects included 5 men and 6 women. This group was of similar age and BMI to the runners, but was running less (P=.036 vs slower runners; P<.001 vs faster runners) than the runners (see table 3). Compared with the runners, the controls had no difference in pre-race pain ratings (42±17, P=.25) (see fig 2) and pre-race overall pain levels (5±7, P=.16) (see fig 3). Model-based analysis further showed no difference in pain ratings between runners and controls at baseline (main effect of group, P=.23) (see table 1). Additionally, there was no change in pain rating (P=.91) (see fig 2) or in overall pain levels (P=.30) (see fig 3) among the controls between the pre-race and post-race evaluations. The correlation of pre-race and post-race pain ratings for the control group was highly significant (r2=.871, P<.001). The use of anti-inflammatory medications and acetaminophen was common among the runners. Of the 12 faster runners, one (8%) reported using anti-inflammatory medication within 6 hours of the pre-race test and 9 (75%) reported taking pain medication (anti-inflammatory medication and/or acetaminophen) that may have been within 6 hours of the post-race test. Among the 9 slower runners, 3 (33%) indicated they had taken anti-inflammatory medication within 6 hours of the pre-race test, and 8 (89%) indicated they had taken pain medication (anti-inflammatory medication and/or acetaminophen) that may have been within 6 hours of the post-race test. Timing of the pain medication prior to the post-race test was not always clear due to the difficulties some runners had in remembering such information after completion of the race. Control subjects reported no use of pain medications within 6 hours of the pain tests. Results Our model selection procedure described in the Data Analysis section converged at the final model that included separate effects of race (time) in slow and fast runners. The group of faster runners was defined prior to the analysis by placing a cutpoint at 25.55 hours, the median of finish time. A significant effect of time in faster runners was identified (P=.017) (table 4). No other effects were significant. | | |  | Parameter | Effect | SE | LR | df | P | |
|---|
| Intercept | 48.13 | 4.03 | 57.80 | 1 | <.001 | | | Race effect in faster runners | −12.76 | 5.20 | 5.68 | 1 | .017 | | | Race effect in slower runners | 0.12 | 5.44 | <0.00 | 1 | .984 | | | | |
Discussion The present study shows that faster runners perceived a painful stimulus to the finger to be less painful shortly after finishing a 161-km running race than before the race. The most interesting aspect of this study is that the exercise-induced analgesia effect was only present for the faster runners, and was not observed among the slower runners. As might be expected, the faster runners tended to be younger and ran greater distances in training than the slower runners. Otherwise, the faster and slower groups of runners appeared similar in various characteristics including BMI, baseline pain ratings to the pain test, and use of pain medications during the race. Both groups also had similar increases in overall pain levels after the race. These mean overall pain levels of 39 to 40mm on a 100-mm VAS were comparable to the post-race “muscle soreness” values of about 4 to 5 on a 10-point Likert scale determined from another group of runners after the same event.9 Presumably, overall pain levels would have been even higher had they been measured during rather than after the race. Activation of the endogenous opioid system is a mechanism that has been proposed to be involved in exercise-induced analgesia.1, 2 Other mechanisms have also been suggested, such as activation of large afferent fibers incorporating the gate-control theory of pain, enhanced psychologic well-being, distraction, and activation of endogenous cannabinoids.7, 17, 18 Yet, even in the situation where there is ongoing pain, the evidence supports the involvement of the endogenous opioid system. Two animal studies have examined exercise-induced analgesia under such conditions. One study used an acute tissue injury model in mice19 and the other study used a chronic muscle pain model in rats.20 Both showed the presence of exercise-induced analgesia that was mediated through the endogenous opioid system. It has been suggested that dysfunction of the endogenous opioid system may play a role in chronic pain. An impairment of the endogenous opioid antinociceptive system may contribute to the elevated pain sensitivity often observed among more disabled chronic pain patients.21 Yet, among less disabled patients, it has been suggested that chronic pain may serve as a primer producing up-regulated opioid antinociceptive responses to acute pain.21 The modest reduction in pain perception observed in the present study among the faster runners, which was comparable to what we have previously observed from moderate exercise,8, 15 suggests that there was no augmentation of the opioid antinociceptive response from the ongoing pain associated with the run. Because the post-race overall pain levels were the same for the faster and slower runners, it would seem that the stimulus to the antinociceptive system due to naturally occurring pain from the exercise was similar between groups. Yet, we cannot be certain that the endogenous opioid antinociceptive system responded similarly in the 2 groups. Nonetheless, a more plausible explanation for the lack of evidence for exercise-induced analgesia in the slower runners might be that these runners were exercising at a lower intensity during the later stages of the race. We have no data on the exercise intensity at which our subjects were working throughout the competition, but limited literature on this topic suggests that the percentage of V̇o2max that can be sustained for the duration of a 161-km ultramarathon is on the order of 50% of V̇o2max.10, 11 It would be anticipated that the slower runners tended to exercise at a lower percentage of V̇o2max compared with the faster runners due to greater limitations on their performance from general fatigue. This could be particularly likely during the later stages of the competition. Given that exercise intensities of over 50% to 70% of V̇o2max may be required to elicit a measurable exercise-induced analgesic effect,3, 4, 5, 8 it is conceivable that the exercise intensity was inadequate for the slower runners to induce an exercise-analgesic effect. The present study was not directed at clarifying the underlying mechanism of exercise-induced analgesia. Still, it is interesting to examine the present results in view of some previous work done at the Western States 100 Mile Endurance Run. Bortz et al22 reported the findings of elevations in β-endorphin concentrations in the blood at 97km and at the finish among competitors in the 1980 race. Although the subjects who were tested at the 2 points on the course were not the same and the authors did not comment on the relative performance level of the subjects who participated in their study, the β-endorphin concentrations appeared to be lower at the finish than at 97km. This could represent a reduction in β-endorphin production during the later stages of the race simply from “exhaustion” of this physiologic function, or due to a reduction in exercise intensity to the point that the endogenous opioid system was not fully activated in some runners. Study Limitations There is always a concern about reproducibility of a measurement that involves some subjectivity as is the case with the pressure pain test that was used in this study. Our previous work has shown acceptable reproducibility for repeat pain tests on a given day after a single practice test on a previous day, as well as for repeat pain tests across days.8, 15 As such, we believe that our design that involved 2 pain tests before the race, with the first serving as a practice, was adequate to assure acceptable reproducibility. Indeed, the lack of change in pain ratings among the control subjects, in which no alteration would be expected, provides further support for the reproducibility of the pain test after a single practice test. It should be noted that we were vigilant in recruiting a control group that included subjects of similar characteristics to the runners. Although it turned out that the control group included a higher percentage of female subjects and was not training at the same level as the runners, there is no reason to suspect that there would have been a variation in pain ratings across tests had there been more men or more active subjects in the control group. Most importantly, the majority of the control subjects were busy and awake throughout the duration of the race. This shows that the reduction in pain ratings among the faster runners was not simply a matter of the stress from having been awake for nearly 24 hours or longer. Field studies are often fraught with variables that are difficult to control. Like other field studies, this one has some potentially confounding issues that warrant comment. Given that this study was performed around a challenging competition for which these athletes had invested considerable effort in order to participate, it was necessary for the study to avoid significant intrusions. As such, we were unable to control the use of pain medications before and during the race. In fact, 4 of the runners reported using anti-inflammatory medications within 6 hours before undergoing the pre-race pain test, and all but 3 of the runners reported taking a pain medication (acetaminophen and/or anti-inflammatory drug) sometime during the race. In contrast, none of the control subjects reported having used a pain medication within 6 hours of any of the pain tests. Because the use of pain medication before (1/12 in the faster group, 3/9 in the slower group) and after (9/12 in the faster group, 8/9 in the slower group) the race appeared comparable between the faster and slower runners, we do not think that this factor can account for the finding of a different effect for the 2 groups of runners. It was also important that none of the subjects exercised shortly before the pre-race pain tests, and that the control subjects did not exercise shortly before the post-race pain tests. Each subject was queried before each pain test, and we found that there was no significant exercise shortly prior to pain testing that should have confounded the results. One variable that we were unable to control to our desired level was the time interval from finishing the race to the post-race pain test. Three of the faster runners were not willing to undergo the post-race pain test until 0.85 to 2.78 hours after finishing, whereas all of the slower runners were tested within the desired interval. Nonetheless, given that the exercise-induced analgesic effect is likely to be less evident the longer it has been after the completion of exercise,8, 15 this issue is not thought to account for the difference between the faster and slower runners. When excluding the 3 runners showing more than a 30-minute interval from finishing to post-race pain testing, the effect of the race was no longer significant (P=.125). However, the estimate of the effect in faster runners (−11.51mm) still points at a strong trend toward increased analgesia. Therefore, we attribute the loss of significance to a reduction in sample size. Furthermore, in the model-based analysis the delay variable showed no significance (P=.51). Another potentially concerning issue relates to differences in air temperature at the time of the post-race testing. Because the faster runners finished between 12:00 pm and 7:00 am, the air temperatures were lower (range, 12°−16°C) than when the slower runners finished (range, 16°−21°C). It is recognized that cooling slows sensory nerve conduction velocities,23, 24, 25 although it is not known if this amount of variation in air temperature would affect pain ratings in the type of test we performed. Nonetheless, because nearly half of the controls performed the last pain test within the temperature range of the faster runners and the other half performed the test within the temperature range of the slower runners, and there was no apparent relationship of pain ratings with air temperature among the controls, we have discarded this issue as having affected the results. We were concerned that some of the runners might have swelling in the fingers that could alter the pain ratings during the post-race pain test. Some finger swelling is not unusual among competitors in endurance events due to various factors including the development of low serum sodium levels.26 If finger swelling was present, it could, in effect, cause some “cushioning” between the edge of the pressure pain stimulator and the finger. This could either lower the pain ratings or cause a shifting of the pain rating–time curve to the right. However, we are unable to formulate a plausible explanation for why this effect would be greater among the faster runners than the slower runners. Thus, we do not believe that this factor accounts for the reduced pain ratings that were observed among the faster runners. Small sample size is recognized as an important limitation in this study. Important effects may have been masked because the study was only powered to detect large differences. It is quite likely that we would have discovered a continuous race effect modification by finishing time in a larger study, as well as other predictors of the association between exercise intensity and analgesia. Conclusions Fastest runners in a 161-km running race experience a modest temporary reduction in pressure pain perception. The magnitude of this analgesic effect was comparable to that previously observed after modest exercise, without evidence for an augmented response due to ongoing pain from the exercise. It is unknown whether the lack of a reduction in pain perception among the slower runners was because an extreme bout of exercise of this nature can “exhaust” the systems responsible for exercise-induced analgesia in all but the most well trained of runners, or if these systems were not activated because the slower runners were unable to maintain a high enough exercise intensity during the later stages of the race. Acknowledgment We thank the leadership of the Western States Endurance Run for support of the research. References 1. 1Janal MN. Pain sensitivity, exercise and stoicism. J R Soc Med. 1996;89:376–381. MEDLINE 2. 2Koltyn KF. Analgesia following exercise: a review. Sports Med. 2000;29:85–98. MEDLINE |
CrossRef
3. 3Kemppainen P, Paalasmaa P, Pertovaara A, Alila A, Johansson G. Dexamethasone attenuates exercise-induced dental analgesia in man. Brain Res. 1990;519:329–332. MEDLINE |
CrossRef
4. 4Kemppainen P, Pertovaara A, Huopaniemi T, Johansson G, Karonen SL. Modification of dental pain and cutaneous thermal sensitivity by physical exercise in man. Brain Res. 1985;360:33–44. MEDLINE |
CrossRef
5. 5Pertovaara A, Huopaniemi T, Virtanen A, Johansson G. The influence of exercise on dental pain thresholds and the release of stress hormones. Physiol Behav. 1984;33:923–926. MEDLINE |
CrossRef
6. 6Black J, Chesher GB, Starmer GA, Egger G. The painlessness of the long distance runner. Med J Aust. 1979;1:522–523. 7. 7Janal MN, Colt EW, Clark WC, Glusman M. Pain sensitivity, mood and plasma endocrine levels in man following long-distance running: effects of naloxone. Pain. 1984;19:13–25. Abstract |
Full-Text PDF (935 KB)
|
CrossRef
8. 8Hoffman MD, Shepanski MS, Ruble SB, Valic Z, Buckwalter JB, Clifford PS. Intensity and duration threshold for aerobic exercise-induced analgesia to pressure pain. Arch Phys Med Rehabil. 2004;85:1183–1187. Abstract | Full Text |
Full-Text PDF (113 KB)
|
CrossRef
9. 9Nieman DC, Henson DA, Dumke CL, et al. Ibuprofen use, endotoxemia, inflammation, and plasma cytokines during ultramarathon competition. Brain Behav Immun. 2006;20:578–584. MEDLINE |
CrossRef
10. 10Davies CT, Thompson MW. Estimated aerobic performance and energy cost of severe exercise of 24 h duration. Ergonomics. 1979;22:1249–1255. MEDLINE |
CrossRef
11. 11Dumke CL, Shooter L, Lind RH, Nieman DC. Indirect calorimetry during ultradistance running: a case report. J Sports Sci Med. 2006;5:692–698. 12. 12Forgione A, Barber TX. A strain gauge pain stimulator. Psychophysiology. 1971;8:102–106. MEDLINE |
CrossRef
13. 13Koltyn KF, Arbogast RW. Perception of pain after resistance exercise. Br J Sports Med. 1998;32:20–24. MEDLINE |
CrossRef
14. 14Koltyn KF, Garvin AW, Gardiner RL, Nelson TF. Perception of pain following aerobic exercise. Med Sci Sports Exerc. 1996;28:1418–1421. MEDLINE |
CrossRef
15. 15Hoffman MD, Shepanski MA, MacKenzie SP, Clifford PS. Experimentally-induced pain perception is acutely reduced by aerobic exercise in persons with chronic low back pain. J Rehabil Res Dev. 2005;42:183–190. MEDLINE |
CrossRef
16. 16Laird NM, Ware JH. Random-effects models for longitudinal data. Biometrics. 1982;38:963–974.
CrossRef
17. 17Simons CW, Birkimer JC. An exploration of factors predicting the effects of aerobic conditioning on mood state. J Psychosom Res. 1988;32:63–75. MEDLINE |
CrossRef
18. 18Sparling PB, Giuffrida A, Piomelli D, Rosskopf L, Dietrich A. Exercise activates the endocannabinoid system. Neuroreport. 2003;14:2209–2211. MEDLINE |
CrossRef
19. 19Carmody J, Cooper K. Swim stress reduces chronic pain in mice through an opioid mechanism. Neurosci Lett. 1987;74:358–363. MEDLINE |
CrossRef
20. 20Hoeger Bement MK, Sluka KA. Low-intensity exercise reverses chronic muscle pain in the rat in a naloxone-dependent manner. Arch Phys Med Rehabil. 2005;86:1736–1740. Abstract | Full Text |
Full-Text PDF (152 KB)
|
CrossRef
21. 21Bruehl S, Chung OY, Ward P, Johnson B. Endogenous opioids and chronic pain intensity: interactions with level of disability. Clin J Pain. 2004;20:283–292. MEDLINE |
CrossRef
22. 22Bortz WM, Angwin P, Mefford IN, Boarder MR, Noyce N, Barchas JD. Catecholamines, dopamine, and endorphin levels during extreme exercise. N Engl J Med. 1981;305:466–467. MEDLINE |
CrossRef
23. 23Abramson DI, Chu LS, Tuck S, Lee SW, Richardson G, Levin M. Effect of tissue temperatures and blood flow on motor nerve conduction velocity. JAMA. 1966;198:1082–1088. MEDLINE 24. 24Nelson R, Agro J, Lugo J, et al. The relationship between temperature and neuronal characteristics. Electromyogr Clin Neurophysiol. 2004;44:209–216. 25. 25Zankel HT. Effect of physical agents on motor conduction velocity of the ulnar nerve. Arch Phys Med Rehabil. 1966;47:787–792. MEDLINE 26. 26Hew-Butler T, Almond C, Ayus JC, et al.Exercise-Associated Hyponatremia (EAH) Consensus Panel Consensus statement of the 1st International Exercise-Associated Hyponatremia Consensus Development Conference, Cape Town, South Africa 2005. Clin J Sports Med. 2005;15:208–213. a Department of Physical Medicine & Rehabilitation, Department of Veterans Affairs, Northern California Health Care System, and University of California-Davis Medical Center, Sacramento, CA b Department of Biostatistics, School of Public Health, University of Michigan, Ann Arbor, MI.  Reprint requests to Martin D. Hoffman, MD, Dept of Physical Medicine & Rehabilitation (117), Sacramento VA Medical Center, 10535 Hospital Way, Mather, CA 95655-1200
No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the authors are associated. PII: S0003-9993(07)00330-9 doi:10.1016/j.apmr.2007.05.004 © 2007 American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation. Published by Elsevier Inc. All rights reserved. | |
|
FULL TEXT
