Reliability of a New Portable Dynamometer for Assessing Hip and Lower Limb Strength

: The purpose of this study was to evaluate intra- and inter-session reliability of the new, portable, and externally ﬁxated dynamometer called MuscleBoard ® for assessing the strength of hip and lower limb muscles. Hip abduction, adduction, ﬂexion, extension, internal and external rotation, knee extension, ankle plantarﬂexion, and Nordic hamstring exercise strength were measured in three sessions (three sets of three repetitions for each test) on 24 healthy and recreationally active participants. Average and maximal value of normalized peak torque (Nm/kg) from three repetitions in each set and agonist:antagonist ratios (%) were statistically analyzed; the coefﬁcient of variation and intra-class correlation coefﬁcient (ICC 2,k ) were calculated to assess absolute and relative reliability, respectively. Overall, the results display high to excellent intra- and inter-session reliability with low to acceptable within-individual variation for average and maximal peak torques in all bilateral strength tests, while the reliability of unilateral strength tests was moderate to good. Our ﬁndings indicate that using the MuscleBoard ® dynamometer can be a reliable device for assessing and monitoring bilateral and certain unilateral hip and lower limb muscle strength, while some unilateral strength tests require some reﬁnement and more extensive familiarization. portable, and externally ﬁxated dynamometer MuscleBoard ® for assessing bilateral hip and lower limb muscle strength and agonist:antagonist strength ratios. Lower (i.e., moderate to good) intra- and inter-session reliability was generally shown for unilateral hip and lower limb muscle strength. These ﬁndings can be a good basis for further use of the described dynamometer in research- and clinical-based environments, where objective, easy to use, and reliable measurement of hip and lower limb strength is both important and necessary. Future studies are needed to evaluate the MuscleBoard ® device in sports and clinical settings.


Introduction
Measurement of athletes' hip and lower limb muscle strength is a common practice in amateur and professional sports. Indeed, numerous research studies have demonstrated the importance of muscular strength in athletic performance (for review, see Suchomel et al. [1]) and an association between insufficient hip and lower limb muscle strength and an increased muscle injury risk in field-based sports (e.g., soccer, rugby, ice hockey, American football, etc.) [2][3][4][5][6]. Aside from performance optimization and injury prevention in sports, assessment of hip and lower limb muscle strength is also relevant for prevention and treatment of injuries in older individuals [7].
Three main methods for measuring hip and lower limb muscle strength are being used in sports and clinical settings. Isokinetic dynamometry is currently considered as the gold standard for assessing human muscle strength; however, its use is often limited to a laboratory setting as these devices are expensive and not portable. Manual muscle testing is practical and a device-independent method; however, its reliability is questionable and depends to a large extent on the experience of the examiner [8]. Finally, hand-held dynamometry is a valid, reliable, portable, and more accessible strength assessment method compared with isokinetic dynamometry [9]. However, studies have shown that the validity and reliability of hand-held dynamometric measurements depend on external stabilization because measurement characteristics often rely on an examiner's strength [10,11]. In addition, there is considerable variability in test procedures, with substantial effects of body position (supine, prone, side-lying, standing, sitting), stabilization methods (manually 2 of 13 vs. non-elastic straps), and dynamometer placement (short vs. long lever arm) being reported [12].
Therefore, new, externally fixed (independent of examiner's strength), portable, easy to use, and accessible dynamometers with standardized test protocols are needed. One externally fixed dynamometer called GroinBar (later renamed ForceFrame) has recently been introduced and evaluated [13][14][15]; however, this device does not allow the measurement of lower body muscles. To this end, a new portable dynamometer called MuscleBoard ® was recently developed to assess the bilateral and unilateral strength of the hip muscles in all planes, as well as knee flexors, knee extensors, and plantar flexors. The purpose of the study was to verify the intra-and inter-session reliability of this novel strength testing device. We hypothesized that this new dynamometer system would produce good to excellent intra-and inter-session reliability results with low within-individual variation.

Participants
Twenty-four participants (17 males, 7 females) were included in the study (average ± SD: age = 21.9 ± 3.6 years, height = 1.79 ± 0.1 m, mass = 75.6 ± 13.8 kg, body mass index = 23.4 ± 2.2 kg/m 2 , body fat = 16.6 ± 4.7%, muscle mass = 79.2 ± 4.5%). All participants self-reported as being recreationally active (>three hours of physical activity per week). Twenty of them were current/former athletes who participated in the following sports: soccer (10), basketball (4), martial arts (3), volleyball (1), cheerleading (1), powerlifting (1). Exclusion criteria were neurological, muscular, skeletal, or connective tissue injuries during the last 12 months in the area of the back, hips, and/or legs. All participants were informed about the purpose and content of the study and gave written informed consent prior to participation. The study was approved by the National Medical Ethics Committee (0120-690/2017/8) and conducted in accordance with the latest revision of the Helsinki Declaration.

Study Protocol
Anthropometrical variables were assessed with a stadiometer and a multi-frequency bio-impedance scale (Tanita MC-980MA, Tanita, Tokyo, Japan). Participants then performed a standardized warm-up, consisting of: (a) 6 min stepping on/off a 25 cm high box, changing the lead leg every minute; (b) dynamic stretching exercises (8 repetitions of hip circles; forward, backward, and side hip bending; leg swings in frontal and sagittal plane); and (c) bodyweight strength exercises with emphasis on the trunk and lower limb muscles (10 repetitions of squats, heel raises, hip bridges, Jackknife sit-ups, and hip extensions). After the warm-up, the participants underwent isometric strength measurements with the dynamometer MuscleBoard ® , as previously described by Marković et al. [16] on an earlier version of the device. Two "U"-shaped and padded aluminum braces (Figure 1), attached to uniaxial load cells (FL34-100 kg; Forsentek Co., Shenzhen, China), were used to measure forces (N), separately for each limb.
Nine tests (shown and described in Figure 2 and Table 1) were performed in a randomized order within a single session. Each test was repeated three times (maximal voluntary isometric contraction for 5 s) with 45-60 s break between repetitions. The break between the tests was set at 180 s. Three sessions were performed as follows: first and second sessions were performed on the same day, separated by a 45-60 min break (to analyze intra-session reliability), while the third session was performed one week after the first visit at the same hour (to analyze inter-session reliability). Hip abduction (HABD), hip adduction (HADD), hip internal rotation (HIR), hip external rotation (HER) were performed unilaterally and bilaterally, while knee extension (KEXT) and plantar flexion (PFLX) were performed only unilaterally. Nordic hamstring exercise (NHE) was performed bilaterally. In all unilateral variations, participant's non-measured leg was next to the dynamometer to avoid any contact with the sensors. Unilateral contractions of the KEXT and PFLX were measured Nine tests (shown and described in Figure 2 and Table 1) were performed in a randomized order within a single session. Each test was repeated three times (maximal voluntary isometric contraction for 5 s) with 45-60 s break between repetitions. The break between the tests was set at 180 s. Three sessions were performed as follows: first and second sessions were performed on the same day, separated by a 45-60 min break (to analyze intra-session reliability), while the third session was performed one week after the first visit at the same hour (to analyze inter-session reliability). Hip abduction (HABD), hip adduction (HADD), hip internal rotation (HIR), hip external rotation (HER) were performed unilaterally and bilaterally, while knee extension (KEXT) and plantar flexion (PFLX) were performed only unilaterally. Nordic hamstring exercise (NHE) was performed bilaterally. In all unilateral variations, participant's non-measured leg was next to the dynamometer to avoid any contact with the sensors. Unilateral contractions of the KEXT and PFLX were measured using a rigid bar that connected both sensor braces, so that a sum of forces from both load cells presented one (unilateral) measurement (shown in Figure 3).
For all isometric strength measurements, participants were instructed to place the measured limb(s) onto the appropriate side of the sensor braces or under the non-elastic strap (see Table 1 for more details) and asked to perform a maximum repetition by pushing their distal part of the lower leg against the brace/strap as hard as possible and holding it with maximum force for 5 s. For the NHE, participants were instructed to maintain straight knee-shoulder line during the descent, to hold arms and hands in front of the body, and to brake the forward fall for as long as possible. Standardized loud verbal encouragement was given during all measurements: "touch, get ready, push, push, push, relax".        Figure 2D); posterior side of thigh in contact with the box; for stabilization, participant held the straps attached to the box (black strap, shown in Figure 2D) the measurement was performed using non-elastic strap that connected participant's lower leg on the anterior side (5 cm proximal to medial malleoli) with a rigid rod mounted on both sensor braces on the superior part (shown in Figure 3A) ANKLE PFLX seated position, hip in 90 • flexion and neutral in other planes (0 • abduction/adduction; 0 • external/internal rotation); knee in 90 • flexion and neutral in transverse plane (0 • internal/external rotation); ankle in neutral position in all planes (0 • plantar/dorsal flexion, 0 • inversion/eversion); non-elastic strap that enabled isometric plantar flexion was placed across the thigh proximal to the knee; for stabilization, the participant held the straps attached to the box (black strap, shown in Figure 2E) the measurement was performed using a rigid rod that was mounted on both sensor braces on the inferior part; metatarso-phalangeal joints were placed on the middle of the rigid rod (shown in Figure  For all isometric strength measurements, participants were instructed to place the measured limb(s) onto the appropriate side of the sensor braces or under the non-elastic strap (see Table 1 for more details) and asked to perform a maximum repetition by pushing their distal part of the lower leg against the brace/strap as hard as possible and holding it with maximum force for 5 s. For the NHE, participants were instructed to maintain straight knee-shoulder line during the descent, to hold arms and hands in front of the body, and to brake the forward fall for as long as possible. Standardized loud verbal encouragement was given during all measurements: "touch, get ready, push, push, push, relax".  Knee extension isometric strength was measured using non-elastic strap that connected participant's lower leg with a rigid rod mounted on both sensor braces on the superior part. Ankle plantarflexion isometric strength was measured using a rigid rod that was mounted on both sensor braces on the inferior part. For both tests, sum of forces from both load cells presented one unilateral measurement. Knee extension isometric strength was measured using non-elastic strap that connected participant's lower leg with a rigid rod mounted on both sensor braces on the superior part. Ankle plantarflexion isometric strength was measured using a rigid rod that was mounted on both sensor braces on the inferior part. For both tests, sum of forces from both load cells presented one unilateral measurement.

Data Analysis
Custom-developed software (ARS Dynamometry, S2P, Ljubljana, Slovenia; created in Labview 8.1., National Instruments, Austin, TX, USA) was used for recording and analyzing the signals, which were sampled at 1000 Hz and pre-processed with a 10 ms moving average filter. The signals were transferred to a computer using a two-channel amplifier (InsAmp, Isotel, Logatec, Slovenia) and an analogue-to-digital card (NIUSB-6009, National Instruments, Austin, TX, USA). The peak force values were multiplied by the lever arms (m) of the participants and normalized by their body mass to obtain normalized peak torques (Nm/kg) for each repetition. Lever arms were measured prior each test to the nearest 0.5 cm using the measurement scale on both edges of the dynamometer (shown in Figure 1) from sensor braces to: (i) greater trochanter for HFLX, HEXT, HABD, HADD; (ii) femoral epicondyle for HIR, HER, KEXT, and NHE; (iii) lateral malleoli for PFLX. Mean and maximal values of three repetitions, and agonist:antagonist ratio (%) for each test were calculated and statistically analyzed. No gravity corrections were performed for any strength test.

Statistical Analysis
The SPSS statistical package (version 25.0, IBM, New York, NY, USA) was used for statistical analysis. Descriptive statistics were calculated and reported as mean ± standard deviation, in addition to the 95% confidence intervals. Shapiro-Wilk test was used to verify normality of data. To assess absolute reliability, the coefficient of variation (CV) was calculated as the quotient between the typical measurement error and the mean value of both sessions, expressed in %. The relative reliability was assessed with intra-class correlation coefficient, using a 2-way random model with absolute agreement (ICC 2,k ) and was interpreted as poor (<0.5), moderate (0.5-0.75), good (0.75-0.9), and excellent (>0.9) [17]. The systematic bias was analyzed by analysis of variance with repeated measurements. The level of statistical significance was set at p < 0.05.
Calculated agonist:antagonist ratios had in general lower intra-session reliability than peak torques (Table 3). In bilateral strength tests, the calculated agonist:antagonist ratios had no systematic bias, good to very good test-retest reliability (ICC 2,k = 0.66-0.83), and acceptable within-individual variation (CV = 8.1-11.9%). In contrast, the calculated agonist:antagonist ratios for unilateral strength tests had lower intra-session reliability. Significant systematic bias was observed for unilateral HIR:HER and HFLX:HEXT ratios (d = 0.49 and 0.40, respectively), with ICC coefficients ranging from low (0.39 for the right HIR:HER ratio) to high (0.8 for the right FLX:EXT ratio) and CVs ranging from low (7.5% for the right FLX:EXT ratio) to high (22.5% for the right HIR:HER ratio). ADD = adduction, ABD = abduction; IR = internal rotation; ER = external rotation; EXT = extension; FLX = flexion; UNI = ratio, calculated from unilateral contractions; BI = ratio, calculated from bilateral contractions; R = right leg; L = left leg; S1 = session 1; S2 = session 2; CI = confidence interval; CV = coefficient of variance; ICC = interclass coefficient; d = effect size-Cohen's d.
The calculated agonist:antagonist ratios in bilateral strength tests had no systematic bias, with very high inter-session reliability for HADD:HABD ratios (ICC 2,k = 0.89-0.92, CV = 3.2-4.4%) and good inter-session reliability for HIR:HER ratios (ICC 2,k = 0.58-0.67; CV = 12.5-16.2%) ( Table 5). For unilateral strength tests, the calculated agonist:antagonist ratios had lower inter-session reliability. Significant systematic bias was observed for unilateral left HFLX:HEXT ratios (d = 0.99), with ICC coefficients ranging from low (0.30 for the left HADD:HABD ratio) to high (0.78 for the right HIR:HER ratio) and CVs ranging from moderate (11.9% for the right HIR:HER ratio) to high (25.6% for the left HADD:HABD ratio).

Discussion
The purpose of this study was to evaluate inter-and intra-session reliability of the new, portable, and externally fixated dynamometer MuscleBoard ® for assessing strength of hip and lower limb muscles. Overall, the results demonstrate high to excellent intraand inter-session reliability with low to acceptable within-individual variation for average and maximal peak torques in all bilateral strength tests. Furthermore, the calculated agonist:antagonist ratios for bilateral strength tests also had good to high intra-and intersession reliability. Conversely, the peak torques and calculated agonist:antagonist ratios for unilateral tests had lower intra-and inter-session reliability.
The results of strength assessments can be reported with different units of measurements (N, Nm, Nm/kg, N/kg, % of bodyweight, etc.), which makes the direct comparison of peak torques/forces from different studies difficult. Nevertheless, Tourville et al. (2013) [18] used externally fixated load cells to measure hip strength of ice hockey players and found similar results for HADD and HABD (157.5 ± 48.9 Nm vs. 158.8 ± 48.2 Nm and 155.7 ± 51.4 Nm vs. 154.8 ± 40.3 Nm from our study, respectively), higher HFLX (174.1 ± 66.9 Nm vs. 146.8 ± 40.8 Nm from our study) and lower HEXT (146.7 ± 56.7 Nm vs. 199.8 ± 52.6 Nm from our study). Furthermore, HADD:HABD ratio of~103% from bilateral measurement and HFLX:HEXT ratio of~75% from unilateral measurement is in agreement with previous studies using the GroinBar device on football players [13,15], while measured HIR:HER ratio of~96% from unilateral and 124% from bilateral measurement in our study is comparable to bilaterally measured the HIR:HER ratio of 108% from the study by Desmyttere et al. (2019) [15]. In addition, peak torques during the NHE showed high intra-session reliability, which is comparable to other similar devices for measuring forces during the NHE-namely, Opar et al. (2015) [19], who tested the NordBord device, and Lodge et al. (2020) [20], who recently tested the Hamstring Solo Elite device. This is a particularly important measurement characteristic of dynamometers since eccentric strength of the hamstrings (measured during the NHE but not on the isokinetic dynamometer) [21,22] is a significant risk factor for the occurrence of hamstring strain injuries [23]. Another important measurement characteristic of the device in the context of prevention/rehabilitation of hamstring injuries is the ability to measure forces for each limb separately since inter-limb strength asymmetry during the NHE is another risk factor for hamstring injuries [24].
We observed high to excellent intra-and inter-session reliability for bilateral tests of HADD, HABD, HIR, and HER and moderate to good intra-session reliability results for unilateral tests of HFLX and HEXT. This is comparable to the results of the study by Desmyttere et al. (2019) [15], in which they used the GroinBar device, a similar externally fixated dynamometer. Results are similar despite the following differences in body positions and lever arm lengths: HFLX was measured in 90 • hip flexion with short lever arm compared to our measurement of HFLX in neutral position with long lever arm; HEXT in prone position in 4-point support and 90 • knee flexion with short lever arm compared to prone lying position with neutral knee position and with long lever arm; HADD and HABD in supine position with 45 • knee and hip flexion with short lever arm compared to supine position and neutral position in hip and knee joint with long lever arm; and HIR and HER in supine position with hip and knee in 90 • flexion compared to prone position in 4-point support. Furthermore, Ryan et al. (2018) [14] found similar low within-individual variation (CV < 10%) and excellent reliability using the GroinBar device for bilateral HADD peak force compared with results from our study. In addition, hip strength measurements using hand-held dynamometry displayed comparable (moderate to excellent) reliability [25][26][27]. Based on the abovementioned findings, it can be concluded that the externally fixated dynamometer MuscleBoard ® analyzed in our study could be a reliable alternative for assessing hip and lower limb strength compared to hand-held dynamometry, as it is not dependent on examiner's strength. Additionally, handling this dynamometer requires less knowledge and experience compared to hand-held dynamometry.
When measuring HADD, HABD, HIR, and HER, bilateral contractions had systematically better (good to excellent) intra-and inter-session test-retest reliability compared with unilateral contractions. There could be two main reasons that made maximal voluntary isometric contractions (more) difficult to perform during unilateral contractions: (i) lack of stabilization/fixation of body position and (ii) participants' division of attention. Participants were instructed to abduct or rotate the non-measured leg outside of sensor braces and to maintain this position during unilateral measurements in addition to generating maximal isometric torque with the measured leg. This seemed even more difficult during unilateral abduction and adduction in the supine position, as the peak torque generated with the measured leg was partly dependent on strong fixation of the trunk, upper body, and non-measured leg. Thus, in our case, participants resisted with the non-measured leg at the outer edge of the dynamometer to improve stabilization during unilateral adduction. During unilateral abduction, participants tried to stabilize themselves by pushing the non-measured leg into the ground, which occasionally caused a slight movement of the dynamometer in the direction of abduction of the measured leg. To prevent this movement, the examiner had to provide additional external stabilization (by holding the dynamometer in place). Moreover, the absence of bilateral deficit (reduction of the produced force from a single leg during maximal bilateral contraction) could be explained by the same two factors (lack of body position stabilization/fixation and division of attention during unilateral measurements). The sum of peak torques for the unilaterally measured left and right leg was lower than peak torques for the bilaterally measured test in all three sessions (e.g., for the adduction in the first session: 2.09 Nm/kg for bilateral test, >0.77 Nm/kg for the right leg, + 0.77 Nm/kg for the left leg). To our knowledge, this is the first study in which both the reliability of bilateral and unilateral contractions was tested on this type of externally fixated dynamometer. It would be interesting to see if the participants would achieve comparable results in terms of bilateral deficit on similarly designed dynamometers (e.g., GroinBar device).
An advantage of the MuscleBoard ® compared to other recently introduced portable, externally fixed dynamometers, is its ability to measure the strength of knee extensors and plantarflexors. Specifically, knee extension strength has been repeatedly shown to be essential in monitoring the rehabilitation process following anterior cruciate ligament reconstruction [28,29], while plantarflexor strength is associated with mid-portion Achilles tendinopathy [30]. The current study showed moderate to high intra-and inter-session reliability of testing strength of these muscle groups, suggesting that MuscleBoard ® has the potential to be used in clinical settings related to knee and Achilles tendon injuries.
Inter-session reliability was generally lower compared with intra-session reliability, with a higher number of measured parameters showing a statistically significant systematic bias. This was particularly evident for the NHE and HFLX tests. The reasons could be in participants' motor learning and task/measurement procedure familiarization, as they had already performed six repetitions before third measurement session started. It is, therefore, suggested that each participant perform one familiarization session before the first official measurement of these muscle groups with the MuscleBoard ® device.
Some limitations of the present study need to be acknowledged. The measurement for each muscle group was performed in only one body position. Different body positions could significantly influence produced peak torques or the reliability of the measurements. For example, HADD maximal strength can be evaluated with different hip and knee flexion angles, which could have a significant impact on peak HADD torques and peak muscle activity [31]. Further studies are needed to test the reliability of the MuscleBoard ® dynamometer for different body positions within the same muscle group. Moreover, a certain degree of learning effect, which could have notably influenced the reliability between the sessions, cannot be ruled out. Hence, a more extensive familiarization (particularly for unilateral test) and, possibly, better stabilization with the arms is warranted when individuals or groups are being tested with a MuscleBoard ® device for the first time.

Conclusions
Our results indicate a high to excellent intra-and inter-session reliability of the new, portable, and externally fixated dynamometer MuscleBoard ® for assessing bilateral hip and lower limb muscle strength and agonist:antagonist strength ratios. Lower (i.e., moderate to good) intra-and inter-session reliability was generally shown for unilateral hip and lower limb muscle strength. These findings can be a good basis for further use of the described dynamometer in research-and clinical-based environments, where objective, easy to use, and reliable measurement of hip and lower limb strength is both important and necessary. Future studies are needed to evaluate the MuscleBoard ® device in sports and clinical settings.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Written informed consent to publish this paper was obtained from the patient(s). Data Availability Statement: Data is contained within the article.
Conflicts of Interest: G.M. and N.Š. are co-owners of the company that developed MuscleBoard ® and is planning its commercialization. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.