Tensiomyographical responsiveness to peripheral fatigue in quadriceps femoris

Study design

A single group pretest-posttest design was used, which involved repeated TMG and MVIC measures of the dominant lower limb QF before and after a fatigue protocol within the same session. Participants were physiotherapy students recruited by email using the University of Valencia Intranet. This study was conducted from April to July 2018. All measurements were carried out between 10 a.m. and 2 p.m in the clinical research laboratory of the Department of Physiotherapy (University of Valencia) at an ambient temperature 21–22 °C. An experienced examiner in the measurement techniques evaluated the participants. He was a physiotherapist who had used TMG and hand dynamometers both in research and in clinical practice for several years. Before participation, participants were informed of the study procedures and their possible associated risks. All of them provided written informed consent. This study was completed following the principles outlined in the Declaration of Helsinki and it was approved by the Ethics Committee of the University of Valencia (Spain) (H1523633864087).

Participants

Thirty-nine recreational athletes were evaluated. All participants performed exercise 3 times per week and practiced activities such as running, swimming, cycling, or central strength training. The specific inclusion criteria were: (a) aged between 18 and 30 years, (b) not surgically

operated on the lower limb, (c) without pain in the lower limb in the 2 months before data collection, and (d) performing physical exercise a minimum of 3 days per week. The exclusion criteria were: (a) practicing a specific sport as an amateur or professional, (b) contraindication to the use of electrodes due to injury or allergy to the adhesive, and (c) nontolerance to electrical stimulation.

Procedures

Before starting the session, height was measured using a 1-millimeter sensitivity flexible tape measure, while body mass and body mass index (BMI) were assessed using a standardized body composition analyzer (Tanita BC 418 MA, Tanita Corp, Tokyo, Japan). Next, TMG parameters were measured and then, participants performed a warm-up, which consisted of 10 min cycling at comfortable speed (80 revolutions per minute) with low resistance and the performance of three submaximal isometric contractions of isometric knee extension (Martins et al., 2017). Following this, the MVIC test was performed. After the fatigue protocol, the order of the tests was reversed, and the strength test was performed first to reduce the time between MVIC and TMG tests in acute fatigue. A schematic representation of the experimental procedures is reported in the Fig. 1.

Tensiomyography measurements

First, participants were placed supine and resting on the stretcher. The knee was placed at 120° of flexion (considering full extension at 180°), fixing such position with a triangular foam cushion (García-García et al., 2016; Martín-San Agustín et al., 2020). The area where the TMG sensor and electrodes were placed was shaved and cleaned with gauze and alcohol. The position of the sensor for each QF belly was determined using the anatomical criteria described in the literature (Dahmane et al., 2005; Tous-Fajardo et al., 2010; Rey, Lago-Peñas & Lago-Ballesteros, 2012). This position was marked with a permanent marker so that it would remain throughout the evaluation. The sensor was finally placed on this point perpendicularly to the thigh and the electrodes were placed at five cm distance from it, forming an imaginary straight line along the belly (Fig. 2).

The contractile properties of each belly were evaluated during an maximal elicited contractions with the TMG electro stimulator (TMG-100 System). Starting from 20 mA with 1ms pulses, each stimulation was increased by 10mA until achieving the maximum radial deformation (Dm) of the muscular belly. A time of 10s was left between stimuli to minimize fatigue or potentiation effects (Krizaj, Simunic & Zagar, 2008). Before data acquisition, a pilot test was done to verify the functioning of the TMG. For each belly, spatial and temporal parameters were measured: Dm, contraction time between 10 and 90% of the Dm (Tc), contraction velocity between 10 and 90% of the Dm (Vc), and contraction velocity of the first 10% of the Dm (V10). TMG has proven to be a method with a high relative [ICC for Dm (0.91–0.99), Tc (0.70–0.98), and VC > 0.95] and absolute (low coefficient of variations for Dm, Tc, and VC) reliability (Martín-Rodríguez et al., 2017; Lohr et al., 2018).

Maximal voluntary isometric contraction test

MVIC of the QF was measured by a MicroFET2 handheld dynamometer (Hoggan Health Technologies Inc., Salt Lake City, UT). Participants were seated in an isokinetic dynamometer (Prima Plus, Easytech, Italy) with their torso and hips tied so they were stable, and with a 90° hip flexion. MVIC was evaluated in 90° knee flexion, considering 0° the complete extension (Fig. 3). MicroFET2 was fixed with a rigid belt perpendicular to the ankle five cm above the malleoli, with a pad between the tibia and the dynamometer to minimize the discomfort caused by the contact (Hansen et al., 2015).

After the warm-up, participants completed three MVIC for 5s, with a 60-second rest after each repetition. Through verbal stimuli, participants were instructed to exert and maintain the maximum effort during the session. MicroFET2 has proven to be a valid method to measure the MVIC of the QF with an excellent inter-examiner reliability (ICC: 0.93, 95% CI [0.83–0.97]) and a minimal detectable change (MDC) of 14.1 N*m (95% CI [9.23–22.01]) (Hansen et al., 2015).

Fatigue protocol test

After performing the baseline measurements, participants were requested to implement a protocol based on a 60s fatiguing isometric contraction at 70% MVC (Melchiorri & Rainoldi, 2011). The experimental setup was the same as the one adopted during the MVIC test. The handheld dynamometer, previously set at 70% MVIC, was used to display the feedback (Melchiorri & Rainoldi, 2011). It was considered that the fatigue was achieved when the reduction of the MVIC was higher than the upper limit of the MDC reported in a previous study (22.01 N*m) (Hansen et al., 2015).

Statistical analysis

Baseline data were summarized as means and standard deviations (SD) for continuous variables and as absolute and relative frequencies for categorical variables. Variables were checked for normality with the Kolmogorov–Smirnov test and homogeneity of variances with Levene’s test. A summary was also provided for participants with and without fatigued QF.

Paired t-tests were used to compare changes in the TMG parameters and MVIC within each sex group. These changes were also compared between sex groups by using non-paired t-tests.

Internal responsiveness was determined by the paired t-test and supplemented with an effect size statistic, as recommended by Husted et al. (2000) [30]. To reduce the probability of getting false positives, we increased the acceptance level from 0.05 to 0.01 for paired t-test because multiple comparisons were made on the same data set. Of the current effect size statistics we used the standardized response mean (SRM), which provides an estimate of the magnitude of change that is not influenced by sample size (Navarro-Pujalte et al., 2018). It was calculated as (MeanFollowup _ MeanBaseline)/Standard deviationFollowup-Baseline and the 95% confidence intervals were calculated using the bootstrapping estimation method. Values of 0.20, 0.50, and 0.80 or higher have been proposed in the literature (Husted et al., 2000) to represent small, moderate, and large responsiveness, respectively. Besides, we calculated the percentage of participants that exceeded MDC. This statistic examines the extent to which change score exceeds the amount of variability accounted by measurement error (Pardasaney et al., 2012), which is calculated as $SEMx1.96x\sqrt{2}$, where SEM is the standard error of measurement.

External responsiveness was determined by correlations, regression models, and receiver operating characteristic (ROC) curves (Husted et al., 2000). The external criterion for assessing the external responsiveness of the TMG tool was the magnitude of change in the MVIC.

We assumed that: (i) changes in the external criterion (MVIC) in participants with fatigue would be associated with changes in the TMG parameters; (ii) participants without fatigue would have the smallest change in the TMG parameters (and therefore change in these TMG parameters can be useful to classify participants’ QF as fatigued or not fatigued). To test the first hypothesis, correlations and simple and multiple linear regression models were used. In the regression models the explanatory variable was the change of each TMG parameter while the response variable was the change in MVIC between before and after protocols. Each model was controlled by sex, and comparisons were carried out between the presence or absence of this control. Goodness-of-fit of the model was assessed by r2. To test the second hypothesis, we calculated the area under the ROC curve (AUC), which represents the probability that the measure of correctly classifying participants has (Husted et al., 2000). An AUC >0.70 was used as a generic benchmark to consider acceptable its discriminant ability (Menaspà, Sassi & Impellizzeri, 2010).

For sample size calculation, we selected the multiple regression as the main statistic of responsiveness because it allowed us to examine change relationships controlling by a covariate relevant in our study (sex). Regarding this statistic, we used the usual rule of thumb that 15 participants per predictor are needed for a reliable equation in multiple regression models (Tabachnick & Fidell, 2007). We recruited a minimum of 30 participants assuming a maximum of 2 explanatory variables (TMG parameter and sex). Statistical significance was set at p < 0.05. All analyses were performed using the Statistical Package for the Social Sciences software program (SPSS version 24.0; IBM SPSS, Chicago, IL, USA).

Participants’ characteristics

Baseline characteristics of participants are listed in Table 1. A total of 35 (89.7%) participants achieved QF fatigue after the application of the fatigue protocol. They were 19 of 20 females (95%) and 16 of 19 males (84.2%). Participants with and without fatigue showed no significant differences (p > 0.05) in any of their baseline characteristics.

Changes associated with the fatigue protocol

Participants with peripheral fatigue (n = 35) had a significant decrease (31.5%) on their MVIC after the fatigue protocol (from 203.3 N*m to 138.9 N*m). Table 2 shows that both sex groups had a similar pattern of change: males reduced 30.8% and females 32.1%. Table 2 also shows patterns of change by sex groups for TMG parameters of the RF, VL, and VM. All these parameters, except for the Tc of the RF and VM, had significant differences within but not between sex groups.

Figure 4 shows changes in TMG parameters for all participants with peripheral fatigue. All parameters, except for Tc, showed a significant difference (p < 0.001) for the three bellies of the QF. Dm’s decrease ranged from 18.22% to 21.65%; Vc decreased from 15.62 to 22.20%, and V10 decreased from 14.80% to 23.77%.

Internal and external responsiveness

Internal and external TMG responsiveness to fatigue of QF bellies is shown in Table 3. Internal responsiveness statistics suggest that all TMG parameters, except for Tc of RF and VM, showed large internal responsiveness (SRM > 0.8) among participants with QF fatigue. Dm and V10 in RF were the parameters in which most of the participants exceeded the MCD (91.3% and 97.1%, respectively). Only Dm, Vc, and V10 of the RF showed to be linearly associated with changes in the MVIC. After controlling by sex, adjusted models typically provided b coefficients and r2 with small variations regarding their respective unadjusted model (range 0.01 to 0.05). Consequently, Dm and V10 of RF were still statistically associated with b coefficients of 0.40 and 0.43, respectively. Moreover, the models of these parameters explained the 22% of the total variance.

The AUC analysis suggests that changes of several TM G parameters (Dm in RF and VL, Tc in VL, and V10 in RF and VM) were >0.70 and could discriminate between QF with and without fatigue. Also, the overlapping among their 95% CI suggests that none of these TMG parameters is superior to the others to discriminate fatigue.