The influence of long distance running on sonographic joint and tendon pathology: results from a prospective study with marathon runners

The data presented here were gathered during the yearly Munich marathon race with more than 14.000 participants. In contrast to most trials published so far, we decided to use high resolution musculoskeletal ultrasound instead of MRI, given the fact that ultrasound has shown to have a comparable sensitivity and specificity [14‐16, 21]. For our purpose, the advantage of this approach is obvious as ultrasound is easily available, quick to perform, and can be moved to the location where test subjects can be reached. These advantages resulted in the recruitment of more than 100 individuals, which to our knowledge has not been reached by other similar trials performed so far.

According to our data, extreme physical stress in highly trained athletes does not influence developement or change of joint effusion in the knee or ankle. This does not differ between male and female participants and was also independent from other covariates such as body-mass-index, age, average weekly training and whether the full or the half marathon distance was covered (Table 2 and Fig. 2). As the follow-up examination was performed within two hours after the race without further follow-up examinations, it can be argued that we could demonstrate only short term effects of extreme exercise on knee and ankle joints. However, we also analysed whether joint effusions at baseline depend on the mean weekly training of the participants but we could not find any significant differences (data not shown). We therefore conclude that there seems to be neither a long term effect of regular sportive activity nor a short term effect of extreme physical stress on the amount of effusion in the respective joints examined in this study.

This however was different for the ultrasound assessment of the patellar tendon, where we could detect significantly more runners with increase of patellar tendon hypervascularity at follow-up. The difference was preserved in the male subgroup and also in most of the other subgroups analysed, identifying acute physical stress as the main contributor to the observed patellar tendon changes (Tables 2 and 3). In contrast to the patellar tendon, the tendons of the anterior tibial, posterior tibial and peroneal aspects of the ankle did not show findings of hypervascularity or tendon sheath effusions before or after the race. Moreover, the ultrasonographic findings were not associated with joint or tendon pain as reported by the participants (data not shown).

Our results partly confirm previously published data of smaller studies. In knee joints of eight runners for example, Hohmann et al. could not demonstrate bone marrow edema, periosteal stress reactions or joint effusions on MRI images after exercise and hence concluded that physical stress in long-distance running is well tolerated [5]. Another study using MRI of the knees before, immediatly after and up to eight weeks after a marathon performance also could not detect negative long-term-effects of physical stress [8]. In contrast to our results, some studies including talocrural joints were able to detect post-exercise findings in this location. Lohmann et al., for instance, demonstrated an increased amount of joint fluid in the talocrural joint. Another study demonstrated the presence of post-exercise bone marrow edema [1, 22] which of course cannot be shown by ultrasound.

Concerning the patellar tendon, ultrasound findings were described in studies involving athletes such as rugby or soccer players and ballet dancers and included microcalculi, thickened tendons, areas of cystic degenerative change, macrocalculi and neovascularisation [23, 24]. Of note, these studies did not compare pre- and post-exercise findings but rather assessed chronic tendon damage while we were interested in acute exercise-related tendon alterations. Considering the given time frame, we found power Doppler ultrasound to be a valid method for this purpose. One small study that did pre versus post exercise tendon assessements including the patellar tendon found abnormalities such as infrapatellar bursitis or proximal patellar ligament thickening after only mild exercise, but power Doppler was not assessed [9]. Kubo et al. were able to demonstrate changes in the blood circulation of the patellar tendon during exercise which appeared to be related to the plasticity of the tendon’s mechanical properties, a finding that was different from the Achilles’ tendon in that study [11, 25]. One large study performing ultrasound assessments of 95 semiprofessional badminton players found that hypervascularity was present in a high percentage of examined tendons. No significant association to pain was observed but the findings significantly decreased during the active season of the players [10]. Other studies focused on parameters such as the cross-sectional diameter of the patellar tendon rather than power Doppler and detected an increased thickening of the tendon diameter after exercise [26, 27].

There are some noteworthy limitations of our study. First, selection bias can not be excluded considering we assessed 105 out of 14.000 potential participants, but recruitement of more participants would not have been feasible with our ressources. Still, compared to other studies on the topic, we present the largest cohort investigated so far. By actively approaching the participants on a random basis, we are confident, that a selection bias was excluded as far as possible. In addition, presence of joint effusion in both knee and ankle joints at baseline and follow-up correlated with presence of joint effusion of the respective contralateral joint, which may indicate expectancy bias during ultrasound assessment. Although we analysed multiple covariates to account for confounders as much as possible, we can not exclude that additional factors may contribute to non-independence of our observations. Also, the observation that there seemed a trend to less effusion after running may be attributed to different positioning of the probe and inter-reader variability. Further, while we used an ultrasound-based approach to the issue, some other studies were performed with MRI, so caution must be given when comparing results of studies using different methods. However, our main intention as rheumatologists was to clarify whether extensive physical stress influences our determinants for joint and tendon abnormalities, for which ultrasound is an evidential method [17, 18, 28]. Furthermore, the examination of structures like the Achilles‘ tendon, the hip joint or the metatarsophalangeal joints would have been interesting as well, but for these examinations, a relocation of the participants respective a switching of the ultrasound probe would have been necessary, both of which would not have allowed us to investigate the prospected number of subjects in the given time frame. Finally, the question how long the patellar hyperperfusion persists after the acute stress is interesting, but sonographic follow-up examinations were not performed because it was not feasible with our study design. Therefore, we cannot premise that the patellar hypervascularity that we visualize by ultrasound is pathological or just a physiologic reaction to extreme physical stress. The data from Boesen et al. would argue in favor of the latter assumption [10].