Main outcomes

Passive resistive torque of the quadriceps muscle-tendon unit is evaluated using a computerized isokinetic dynamometer (Biodex system 3 Pro, Biodex Medical, Shirley, NY, USA). The participant is placed in a standardized position on the seat of the dynamometer (see Additional file 4). The pelvis as well as the thigh of the tested leg are fixed with restriction straps to minimize secondary movement. The opposite hip is fixed at 90° flexion to limit pelvic and lumbar motion. The knee axis is aligned with the rotational axis of the dynamometer.

To obtain PRT, the lower leg is moved from full knee extension (0°) to maximal achievable knee flexion angle with an angular velocity of 5°/s in passive mode of the dynamometer. Torque (T) and angle (θ) are recorded at 100 Hz. This procedure has been described as a reliable method to evaluate passive tissue properties for various positions and muscles (intraclass correlation coefficient (ICC) ranging from 0.88 to 1.00) [28–33]. Torque data is gravity corrected and filtered using a Butterworth, zero-lag, fourth-order low-pass filter with a 10-Hz cutoff frequency [22].

A fourth-order polynomial (FOP) model is fitted on the T-θ data (and stiffness is calculated using the slope of the FOP model [34, 35]. Passive resistive torque as well as stiffness values from four angles during the last 13° of passive tissue tensioning (1°, 5°, 9° and 13°) are calculated and serve as a quantification of resistance and stiffness during passive motion [22].

To monitor muscle activity, surface electromyography (sEMG) is used with two surface electrodes (Ambu Blue Sensor, Ambu GmbH, Bad Nauheim, Germany) placed on the head of the M. rectus femoris muscle with an 8-mm inter-electrode distance and one reference electrode on the patella, according to SENIAM recommendations [36]. Participants are provided with live biofeedback of muscle activity to prevent involuntary muscle contraction.

While assessing PRT, the probe of a high-resolution ultrasound (US) device (Siemens Acuson X300, Siemens Healthcare GmbH, Erlangen, Germany) is positioned on the thigh (for details, see below). Sliding of fascial layers is quantified with a frame-by-frame, cross-correlation algorithm of the generated US images obtained during the passive movement. The cross-correlation method developed in MATLAB (The MathWorks, Inc, Natick, MA, USA) by Dilley and colleagues [37] is used to calculate the correlation coefficient between the pixel gray levels for selected rectangle-shaped regions of interest (ROIs) in two adjacent images. The pixel shift providing the maximum correlation coefficient corresponds to the relative movement between two frames [37]. The method has been extensively used to quantify nerve movement and represents a reliable method to quantify tissue movement in vivo (ICC ranging from 0.70 to 0.99) [37–44].

The linear array US transducer used (4–11.4-MHz, 38.4-mm footprint) is placed on the proximal third of the muscle belly of the M. rectus femoris and sequences of 20 s are captured at 10 frames/s during passive knee flexion (starting at 0° until 100° of knee flexion at 5°/s). US transducer location is marked on the skin with a permanent marker. Participants are instructed to renew the marker on a daily basis to ensure equal transducer placement at all three testing sessions. Six ROIs are placed on the superficial and deep layers of the fascia lata, respectively, to quantify sliding of these layers during passive stretching of the underlying muscle (see Additional file 5). Maximal lateral movement of ROIs/fascial layers is calculated and analyzed as a quantification of fascial sliding.