This study was designed as a case-control study (level of evidence III). The local ethics committee approved the study (FF 33/2009). The registration trial number is DRKS00021589 on DRKS. 'Retrospectively registered'. Date of registration: April 4, 2020.
Patients and control group participants
We recruited patients and control group participants from the orthopaedic clinic of our sports medicine centre. The control group participants were selected to resemble the patient group with respect to anthropometric data and sport/running behaviours (all p > 0.05; Table 1). Eligible participants were screened according to our inclusion and exclusion criteria. Selected participants received detailed oral and written information about the study project, design, and operational aspects of the study and provided written informed consent.
Participants between 15 and 50 years of age were included in either group. All participants had to perform regular recreational or competitive sport activity on a regular basis (≥ 2×/week and/or ≥ 2 h/week). Patients were included in the patient group if CECS was diagnosed from a typical history and unremarkable physical examination. Uni- or bilateral anterior and deep posterior CECS with a spontaneous, non-traumatic onset were accepted. Specifically, a minimum of 12 weeks history of running-induced pain was required, which reproducibly irradiated over more than 10 cm of the deep posterior or anterior leg compartment. Finally, the increasing pain had to force the patient to interrupt the running activity. After cessation of the inducing activity, the symptoms had to be relieved completely within less than five minutes. For the control group, no history of exercise-induced leg pain and an unremarkable physical exam were used to exclude CECS.
Exclusion criteria for both groups were: exercise induced leg pain (EILP) different from CECS, such as bone stress injuries, pain of osteo-fascial origin, particularly medial tibial stress syndrome, pain of muscular origin, pain due to nerve entrapment, and pain due to a temporary vascular compromise [2]. We excluded non-athletes, persons with abnormal alignment of the lower extremity, uncertain compliance, acute or degenerative spine diseases, systemic diseases (e.g., diabetes), and persons who participated in other clinical studies up to one year ago. We also excluded patients with acute or traumatic onset of symptoms or with previous leg surgery.
When CECS occurred bilaterally, we tested both sides. To avoid a ‘double-dipping effect’ [21], we chose the more painful leg of the patients for further analysis (Fig. 1) [22]. For the control persons, the side to evaluate was chosen randomly.
Microdialysis and experimental procedure
With the tested person lying supine, the microdialysis probe was inserted under ultrasound guidance into the middle of the respective muscle belly. According to the manufacturer’s instructions and under sterile conditions, a sterile, single use microdialysis probe (CMA 63, M Dialysis AB, Stockholm, Sweden) was inserted into the muscle parallel to the fibres with a splitable introducer. The shaft was 40 × 0.9 mm and the membrane length was 30 mm. The membrane cut-off was approximately 20,000 Daltons. The microdialysis catheter was fixed on the leg with tape. The microdialysis catheter’s inlet tube was connected to a microdialysis pump. The outlet tube ended with a microvial holder where the sample was collected into small microvials. The probes were perfused with sterile Ringer solution at a flow rate of 2 µl/min by means of a precision infusion pump (CMA 100, CMA Microdialysis, Stockholm, Sweden). Sampling was performed for 6–12 × 10 min. Then, the whole microdialysis system was removed and the point of insertion was dressed with a sterile tape. The patients then ran in the adjacent forest at his/her preferred speed. The patients returned when the typical CECS pain forced them to discontinue. Control persons ran a distance of 5 km with an exhaustive speed. Then, another microdialysis measurement was commenced within 10 min and sampling was performed for another 6–8 × 10 min.
Dialysate analysis
Samples in the microvials (10–20 µL) were initially frozen at − 80 degrees C and were later analysed using micro-analysers as previously described [13;15]. Glucose and lactate concentrations represent energy metabolism while glycerol and glutamate represent mediators of cell damage and pain, respectively. Samples obtained 6 × 10 min before and after the exercise load, respectively, were analysed and included for further statistical processing. In the first step, means and standard deviations were calculated for each pre- and post-load 10 min sampling period. We defined a ‘rest’ phase including the 6 × 10 min samples before exercise (first microdialysis probe), a ‘peak’ phase including the initial two samples after exercise and implantation of the second microdialysis probe, and a ‘recovery’ phase as the 4 × 10 min measurements following the ‘peak’. For the longitudinal pre- and post-exercise analyses within the CECS and control group, respectively, the 6 × 10 min ‘rest’ dialysates were compared with the ‘peak’ dialysates after exercise and implantation of the second microdialysis probe. The respective results were averaged within the groups.
Statistical analysis
Anthropometric data and running duration between groups were compared by t-tests for independent samples. Repeated measures two-way ANOVA (Software: GraphPad® Prism 4.0, San Diego, CA) was performed to identify between-group differences (CECS vs. control), time effects, and group × time interaction effects. There were 22/1056 (2.1%) values missing within all individual data sets. For further statistical analysis, these values were filled in by interpolation. One-way ANOVA was used to test for differences between rest, peak, and recovery phases. For the determination of statistical power, we used the equation N = 2 SD2 × power index / delta2. We expected standard deviations of 25% in each group, and we aimed to find significantly different values (p < 0.05) at differences of 30% between healthy subjects and CECS patients. We respectively calculated the required number of participants as 8.61 patients per group. With 6 healthy volunteers and 12 patients, our study has a power of approximately 80%. The power index of 6.2 was taken from Harvey Motulsky´s textbook “Intuitive Biostatistics” [23].