Task‐specific strength increases after lower‐limb compound resistance training occurred in the absence of corticospinal changes in vastus lateralis

What is the central question of the study? Are corticospinal responses to acute and short‐term squat resistance training task‐specific? What is the main finding and its importance? A single bout of resistance training increased spinal excitability, but no changes in corticospinal responses were noted following 4 weeks of squat training despite task‐specific increases in strength. The present data suggest that processes along the corticospinal pathway of the knee extensors play a limited role in the task‐specific increase in strength following resistance training.

firing rates and decreased recruitment thresholds of motor units (Del-Vecchio et al., 2019). In recent years, transcranial magnetic stimulation (TMS) of the primary motor cortex and electrical stimulation of the corticospinal tract at subcortical levels have been performed to assess these adaptations (for reviews see Mason et al., 2019;Siddique et al., 2020), with the change in evoked electromyographical (EMG) responses used as indices of adaptation.
The neural adaptations to resistance training are considered, by some, to be a form of motor learning, as the individual learns to produce specific patterns of muscle recruitment (Carroll et al., 2001).
Indeed, motor learning and resistance training share similar patterns of adaptation, such as a reduction in motor cortex inhibition and an increase in corticospinal excitability (Leung, Rantalainen, Teo, & Kidgell, 2017;Ljubisavljevic, 2006;Weier et al., 2012), and it is well established that the adaptations to motor training are specific to the trained task (Beck et al., 2007;Schubert et al., 2008;Taube, Gollhofer, & Lauber, 2020). Specific to resistance training, when two distinct tasks are employed (ballistic vs. sustained contractions), neural adaptations are only demonstrated when the corticospinal tract is stimulated during the trained task (Giboin, Weiss, Thomas, & Gruber, 2018). The notion of utilising a task-specific testing task has been echoed throughout the past decade, with researchers highlighting the need to assess neurophysiological variables during the motor task used as in the intervention (Avela & Gruber, 2011;Kalmar, 2018;Sidhu, Cresswell, & Carroll, 2013). Despite the requirement for task-specific neural assessment, adaptation in response to lower-limb compound resistance training has not been assessed in a task-specific manner. For instance , Weier et al. (2012) assessed corticospinal responses in a single-limb isometric task following 4 weeks of squat resistance training. Similarly, following an acute bout of squat training, Thomas et al. (2018) demonstrated no changes in measures of corticospinal function when assessed in a single-limb isometric task. Thus, a common approach in the literature is to assess corticospinal adaptations to squat training by evoking responses in the knee extensors during a single-limb isometric testing task, with considerably different neuromechanical characteristics to the squat exercise. Indeed, our laboratory recently demonstrated poor agreement between measures of short-interval intracortical inhibition (SICI) and corticospinal excitability when measured during isometric squat (IS) and knee extension (KE) tasks at the same relative intensity (Brownstein et al., 2018a). Given the disparities in neural activity between the two tasks, it is possible that neural adaptations to squat resistance training could be masked when measuring responses during isometric KE. As such, investigating neural changes to squat exercise using a task that more closely replicates the squat is warranted.
Considering that improving bilateral lower-limb force production is a goal of neurorehabilitation and athletic training programmes (Baker & Nance, 1999;Carr & Shepherd, 2010;Ng & Shepherd, 2000), understanding the mechanisms of neural adaptations in response to lower-limb compound interventions is necessary to inform exercise prescription in a range of populations. Consequently, determining appropriate testing methodologies in order to capture these neural

New Findings
• What is the central question of the study?
Are corticospinal responses to acute and short-term squat resistance training task-specific?
• What is the main finding and its importance?
A single bout of resistance training increased spinal excitability, but no changes in corticospinal responses were noted following 4 weeks of squat training despite task-specific increases in strength. The present data suggest that processes along the corticospinal pathway of the knee extensors play a limited role in the task-specific increase in strength following resistance training.
adaptations is imperative in obtaining valid results. Therefore, the present two-part study aimed to quantify the corticospinal responses to acute (Study A) and short-term (4 weeks; Study B) squat resistance training in a task-specific IS (Brownstein et al., 2018a)  (3) in both Study A and B, the IS task would demonstrate greater changes in evoked corticospinal responses, due to the task-specific nature of assessment.

Ethical approval
The study received institutional ethical approval from the

Northumbria University Health and Life Sciences Research Ethics
Committee (submission reference: 9610) and was conducted according to all aspects of the Declaration of Helsinki, apart from registration in a database. Participants provided written, informed consent to volunteer for the study.

Participants
Based on the effect sizes reported in Weier et al. (2012)

Experimental design
Participants in the training group visited the laboratory 15 times in total, including a familiarisation visit, pre-training assessment, 12 training sessions (separated by a minimum of 24 h), and a posttraining assessment (see Figure 1 for details

Study B: responses to short-term training
The baseline visit began with a resting ultrasound assessment to discern VL thickness (see 'Ultrasound'). Subsequently, the neuromuscular assessments (in both IS and KE tasks, pseudorandomised order) were performed with 20 min rest between the two in order to negate the influence of fatigue (Carroll, Taylor, & Gandevia, 2017).
Participants performed three MVCs separated by 30 s. The greatest instantaneous force of the three was used to set a target guideline at 10% MVC, whereby all subsequent stimulations were performed.
Next, SICI was assessed from 20 conditioned and 20 unconditioned TMS pulses, of which the unconditioned pulses were also used as an index of corticospinal excitability (expressed relative to M max ), and the TMS silent periods were used as an index of neural inhibition. LICI was then assessed using 20 conditioned and 20 unconditioned pulses, and 10 LEPs were evoked to assess spinal motoneuron excitability.
Following the neuromuscular assessments, participants were given 20 min rest, then performed a warm-up followed by a dynamic one repetition maximum (1RM) squat. The warm-up consisted of 5 min cycling at 1.5 W (kg body mass) −1 , followed by warm-up sets of one to five repetitions of back squats (high bar position), beginning with an unloaded barbell and progressing to 50, 70, 80 and 90% of their estimated 1RM. The load on the bar was then incremented by 2-5% until participants could not complete one repetition. A maximum of three attempts at each weight were permitted, and participants were required to descend to a depth corresponding to 90 deg of knee flexion. Squat depth was verified by tracking the position of the barbell (GymAware, Kinetic Performance, Canberra, Australia). Participants performed a testing visit 2-4 days following the final training session to permit recovery of exercise-induced neuromuscular dysfunction (Howatson, Brandon, & Hunter, 2016) and TMS-evoked responses (Škarabot et al., 2019c). Post-training assessments were performed at absolute (10% of pre-training MVC) and relative (10% of post-training MVC) intensities.

Training protocol
The training protocol for Study B was similar to that used in Weier et al. sets. This training protocol was previously shown to be effective in eliciting maximum strength improvements (Weier et al., 2012). The velocity of each repetition was controlled using an audible electronic metronome and visual feedback of bar displacement (GymAware) to ensure a 3 s eccentric phase, 3 s concentric phase, and adequate squat depth (90 deg knee flexion; Weier et al., 2012). The metronome paced approach to strength training was utilised as previous research has demonstated that corticospinal adaptations are apparent only after externally paced, and not self-paced, strength training (Ackerley, Stinear, & Byblow, 2011;Leung et al., 2017). Once participants could complete four sets of eight repetitions at the target load, the load was increased by 5%, whereas the load was maintained if participants were unable to complete all repetitions (see Figure 4a for depiction of progression in training load). In Study A, five participants were able to successfully perform four sets of eight repetitions, and five participants performed three sets of eight and one set of six. All participants performed the 12 training sessions in Study B.

Isometric knee extension
Isometric knee extension force (N) was measured using a calibrated load cell (MuscleLab force sensor 300; Ergotest Technology, Prosgrunn, Norway). The load cell was fixed to a custom-built chair and strapped with a non-compliant cuff to the participant's right leg, superior to the ankle malleoli. Hip and knee angles were set at 90 deg flexion measured using a goniometer at the beginning of the trial and visually inspected by the investigators throughout the trial to ensure consistency. Participants were instructed to maintain the same posture throughout trials. Verbal encouragement was provided by the investigators during MVCs, and real-time force feedback was provided to the participants on a computer screen directly in front of them (Spike2 v8; Cambridge Electronic Design (CED), Cambridge, UK).
During stimulation procedures, a horizontal line corresponding to 10% MVC was provided on the screen and participants were instructed to try to match the line as closely as possible.

Isometric squat
A detailed procedure for the isometric squat assessment task has been previously published (Brownstein et al., 2018a). A force plate placed directly under the right foot of the participants (Type 9286B; Kistler Group, Winterthur, Switzerland) was used to measure isometric squat force, with a sampling frequency of 5 kHz. Participants were seated on a bench placed directly under a fixed barbell to provide support during isometric contraction. The barbell height was adjusted according to individual torso length and positioned on the shoulders.
Knee and hip angle were kept at 90 deg as measured by the goniometer at the beginning of the trial. Participants were instructed to keep their feet at hips width apart with the toes pointing forwards. The position of the foot was marked on the force plate with tape to ensure consistent placement throughout the trials and maintenance of knee and hip joint angles. Additionally, the position of the hip and knee were visually inspected by the investigators throughout the trial to ensure consistency. This knee and hip position also ensured similar VL muscle length in both IS and KE tasks and thus avoided the muscle length-related differences in neural recruitment (Doguet et al., 2017).
Participants had freedom in choosing their hand position on the bar, but were instructed to keep it consistent throughout the protocol.
Participants were instructed to keep the neck in an anatomical zero (neutral) position and orient their gaze on the screen in front of them where force feedback was provided. This indirectly ensured consistency of head position throughout TMS trials. During contractions, participants were instructed to exert force upwards against the bar using their whole body (Bishop et al., 2017). Similar to KE, verbal encouragement was provided by the investigators and real-time force feedback was provided to the participants on a computer screen directly in front of them, including a horizontal guideline at 10% MVC for stimulations (Spike2 v8).

Electromyography
Surface bipolar EMG activity was recorded using self-adhesive surface  and KE (167 ± 86 mA). were chosen based on modelling studies that showed the greatest electrical field magnitude was induced between T 10 and T 12 spinous processes as the electric field is highest between the stimulating electrodes (Kuck, Stegeman, & van Asseldonk, 2017). As a result, the site of the greatest spinal cord activation is likely to occur between L 1 and L 5 spinal cord segments, corresponding to the motoneuron pool of the quadriceps (Sayenko et al., 2015). This stimulating site has been shown to activate corticospinal axons at the level of lumbar spinal segments (Škarabot et al., 2019b). Latency of the response was constantly monitored for an abrupt change with increases in stimulus intensity, to minimise the possibility of dorsal roots being activated (Taylor & Gandevia, 2004). To ensure ventral roots were not activated, a change in LEP size with increased contraction strength was ensured before the testing protocol (Martin, Butler, Gandevia, & Taylor, 2008). The electrodes remained in place throughout Study A ensuring consistency of stimulating site. In Study B, the electrodes were repositioned following 4 weeks of resistance training using anatomical landmarks as reference points (Nuzzo et al., 2017). The intensity of stimulation was standardised to ∼15-25% M max evoked at 10% MVC, and remained constant from PRE to POST45 in Study A (IS:

Transcranial magnetic stimulation
Single-and paired-pulse TMS were delivered over the motor cortex via a concave double-cone coil using a Magstim 200 2 magnetic stimulator (Magstim Co., Ltd, Whitland, UK). Initially, the junction of the doublecone coil was placed 1-2 cm left of the vertex and oriented to induce posterior-to-anterior cortical current. After that, the optimal location ('hotspot') was determined by locating the coil position that elicited the greatest MEP amplitude in the VL muscle at 50% stimulator output during a 10% MVC and was subsequently marked with indelible ink.
After that, active motor threshold (AMT) was determined during a 10% MVC and defined as the intensity that elicited a MEP amplitude in the VL muscle of >200 µV in three out of five trials (Kidgell, Stokes, Castricum, & Pearce, 2010
Muscle thickness has been shown to be highly associated with resistance training-induced changes in anatomical cross-sectional area (Franchi et al., 2018). Prior to the measurement of VL thickness, participants lay supine for 20 min to allow for fluid distribution to equilibrate (Berg, Tender, & Tesch, 1993). With the participant laid supine and their non-dominant leg fully extended, the distal and proximal insertions sites and the medial and lateral borders of the VL were identified using an ultrasound probe (7.5 MHz linear array probe, 55 mm wide). Muscle length and width were measured using an anthropometric tape measure, and muscle thickness was then measured at 50% of VL length and width. The distance between the superficial and deep aponeurosis was used for muscle thickness.
Digitizing software (ImageJ 1.45, National Institutes of Health, Bethesda, MD, USA) was used for image analysis, with the average of three measurements taken across the width of the image recorded.

Data analysis
All analyses were performed offline using Spike2 software. The

Statistical analysis
All data are reported as means ± standard deviation. Normality and sphericity of the data were assessed using Shapiro-Wilks and Mauchly's test, respectively. All data were normally distributed.
Within-(Study A) and between-session (Study B) test-retest reliability was calculated from control group data using multiple indexes (Atkinson & Nevill, 1998;Hopkins, 2000), including bias (using repeated measures ANOVA for Study A and paired-samples Student's t test for Study B) and within-participant variation as typical error (standard deviation of the mean differences divided by the square root For Study B, a three-way (2 × 2 × 2; group (training and control), task (IS and KE), and time (PRE, POST)) ANOVA was used to assess whether responses to short-term squat training were task-specific.
To assess training induced changes in 1RM, a two-way (2 × 2; group (training and control) and time (PRE and POST)) ANOVA was used.
If significant interactions or main effects were found, analysis was continued using pairwise comparisons with Bonferroni correction.
Analyses were performed on both absolute and relative intensity posttraining data, but the results were similar for both, and therefore for simplicity only the relative data are reported. Statistical significance was determined as an of 0.05. Hedge's g with correction for small sample sizes was calculated to estimate effect sizes of between-group differences (<0.2 is a small, 0.2-0.8 is a medium, >0.8 is a large effect).  No change in neural inhibition in response to squat exercise was observed in either task (Figure 4)

Study B: responses to short-term training
There were no between-group differences in 1RM at baseline The stimulus intensity at AMT did not differ in either task or between the training and control groups, with no time × group (F 1,16 = 0.05, P = 0.820, g = 0.10) and time × task × group interaction (F 1,16 = 0.11, P = 0.745, g = 0.14) observed. No change in corticospinal excitability was observed following short-term squat training measured in either task, or between the training and control groups

DISCUSSION
The present study aimed to assess whether the neuromuscular responses to acute (Study A) and short-term (Study B) squat resistance training were task-specific, using a comprehensive assessment of the

Responses to acute resistance training
The voluntary force production and evoked potentials (Taylor & Martin, 2009). Therefore, the increase in LEP amplitude following acute squat training could also be due to improved efficacy of corticospinalmotoneuronal synapses. Indeed, LEPs exhibited a delayed facilitation following exercise (only observed from POST15 onwards), which is the hallmark of improved efficacy of corticospinal-motoneuronal synapses (Taylor & Martin, 2009), and is in agreement with Nuzzo et al. (2016).
Alternatively, the delayed facilitation could be due to the dissipation of the neuromodulatory effects of exercise-induced fatigue, which has F I G U R E 5 Squat load throughout 12 training sessions across 4 weeks (a) and squat one repetition maximum pre-and post-4 weeks of squat resistance training (b). The blue line with circles in (a) denotes the sample mean, whilst each black line represents an individual participant; in (b) *P < 0.001 compared to 'PRE' previously been shown to reduce spinal motoneuron excitability (Finn, Rouffet, Kennedy, Green, & Taylor, 2018). Nevertheless, the present data corroborate previous findings in other muscle groups showing increased spinal motoneuron excitability following an acute bout of resistance training (Nuzzo et al., 2016).
The indices of inhibition (SICI, LICI, SP) remained unchanged following an acute bout of lower-limb compound resistance training. This is in agreement with similar data following a bout of heavy strength training in the elbow flexors that demonstrated no change in SICI or LICI (Latella et al., 2016), but in contrast to Latella et al. (2018), who reported reduced SP duration in RF up to 1 h following acute resistance training in the knee extensors. Similar to the discrepancies with acute changes in MEP amplitude, no change in SP could be due to the muscle(s) trained, the training protocol and the assessment task. Furthermore, measures of SP might be significantly constrained by differences in the methods employed to evoke this measure (Škarabot, Mesquita, Brownstein, & Ansdell, 2019d). Using a similar squat exercise protocol to the present study, Thomas et al. (2018) showed no change in SICI following an acute bout of squatting when assessed in the non-specific KE. Using a task-specific assessment, this study provided further evidence that lower-limb compound resistance training does not induce immediate adjustments in neural inhibition when measured in VL.

Responses to short-term lower-limb compound resistance training
The present training protocol improved the 1RM squat of participants in the training group (+35%). This increase in 1RM was only reflected in the MVC in the IS (+49%) and not the KE (+1%) task. Due to no change in muscle thickness, the changes in strength cannot be explained by adaptations in muscle cross-sectional area (Franchi et al., 2018). The lack of change in muscle thickness is in agreement with previous findings (Weier et al., 2012), which showed no change in muscle thickness and isometric KE strength following 4 weeks of squat resistance training. It is important to note, however, that ultrasound measures were confined to a single site on the VL and thus potential changes in hypertrophy at other, e.g. distal (Häkkinen et al., 2001), muscle sites cannot be entirely excluded without additional measuring sites or the use of more sensitive measures (e.g. magnetic resonance imaging). Although no changes in muscle thickness suggest that the strength adaptations might be underpinned by alterations within the CNS, the present study found no effect of short-term squat training on corticospinal excitability or intracortical inhibition when measured during either KE or the more task-specific IS, at absolute or relative contraction intensities. These results are in contrast to that of Weier et al. (2012), who demonstrated a substantial increase in corticospinal excitability (+116%) and reduction in intracortical inhibition (−32%) following the same training protocol and similar experimental design to that of the present study. While the results differ from that of Weier et al. (2012), discrepancies between studies in this area have been previously highlighted (Kidgell, Bonanno, Frazer, Howatson, & Pearce, 2017) reported changes in intracortical inhibition following strength training (Siddique et al., 2020), but in agreement with numerous other studies (Carroll, Barton, Hsu, & Lee, 2009;Christie & Kamen, 2014;Coombs et al., 2016), and the conclusions of Kidgell et al. (2017), who in their systematic review suggested that the change in MEP amplitude following strength training is negligible.
To investigate any potential alterations in the corticospinal tract at a segmental level, the present study compared responses to subcortical stimulation before and after the training protocol. No difference was observed following the 4 weeks of training, which agrees with the only other study to employ this type of investigation (Nuzzo et al., 2017). While Nuzzo et al. (2017) elicited CMEPs in the elbow flexors at rest to avoid differences in muscle activity, the present study aimed to recreate the training task as best as possible to maximise aspects of task specificity (i.e. posture, joint angles, bilateral force production), and because resistance traininginduced changes have been shown to be only detectable in an active muscle (Siddique et al., 2020) Mason et al., 2017). Indeed, individual muscle groups involved in the movement might not be controlled by distinct areas within the motor cortex, but are more likely interconnected by intrinsic collaterals involved in the integrated control of muscle synergies (Capaday et al., 2013). Regarding the IS, the knee extensors act as the primary agonists, but these muscles are supported by other agonist and synergist muscles, including spine stabilising muscles (e.g. rectus abdominis, the obliques and erector spinae; Nuzzo, McCaulley, Cormie, Cavill, & McBride, 2008;Willardson, Fontana, & Bressel, 2009). Thus,

Further considerations
given the overlapping and intertwined nature of muscle representation in the motor cortex (Devanne et al., 2006), it is plausible that the changes in activation patterns of synergists could have contributed to the task-specific expression of strength in the present study. This is consistent with the lack of change in isometric knee extension strength, whereby the quadriceps act as an agonist without significant contribution of synergist muscles.
Furthermore, the present study demonstrated that LEP size was increased following the acute session, but remained unchanged following the 4-week training period. One possible explanation for this is that the acute increase in LEP was a result of 'reactive plasticity' , in which general, possibly compensatory, changes at the spinal cord occur due to a change in the activity of surrounding networks directly implicated in the task, i.e. those involved in 'primary plasticity' (Giboin, Tokuno, Kramer, Henry, & Gruber, 2020;Wolpaw, 2010). In turn, this could induce secondary changes in spinal pathways not directly implicated in the task, in our case leading to an increase in LEP amplitude during both the IS and KE. Subsequently, time-dependent, task-specific neural reorganisation might have occurred throughout the training period, meaning that spinal alterations could only be observed during the task itself (Giboin et al., 2020). Despite the more task-specific nature of the IS, the differences in characteristics of the dynamic squats involved in the task (see 'limitations' section) could have precluded the detection of these neuroplastic changes in response to the intervention.
Future research should also consider potential alterations in other descending tracts. For example, the reticulospinal tract is implicated in force generation during gross and forceful motor tasks (Baker & Perez, 2017;Zaaimi, Edgley, Soteropoulos, & Baker, 2012), and its neurons have been shown to synapse onto -motoneurons in primates (Riddle, Edgley, & Baker, 2009), and are activated bilaterally within the spinal cord (Davidson, Schieber, & Buford, 2007). It is conceivable that this descending tract is implicated in force production during the squat, a gross bilateral motor task, but any adaptations within this tract in the present study would have gone undetected with the methodology employed.

Limitations
As acknowledged previously (Brownstein et al., 2018a), whilst the present IS squat set-up provides means to assess neuromuscular responses in a task that more closely replicates the characteristics of the squat, it still exhibits differences compared to a conventional dynamic squat. These differences include, but are not limited to, in the future will be designing an experimental set-up that replicates the characteristics of the dynamic squat more accurately, whilst recreating the demands of training and working within the constraints of measuring stimulation responses during different contraction types and muscle lengths, and potential differences in neural activation patterns that can arise from those (Doguet et al., 2017;Škarabot et al., 2019a).
To assess spinal excitability following 4 weeks of resistance training, electrode position over the spinal cord was replicated and intensity of stimulation kept the same as during pre-training assessments. Whilst this approach has been used previously in a similar investigation utilising magnetic CMEPs (Nuzzo et al., 2017), it should be noted that factors other than adaptations within the spinal cord might have contributed to the response, namely subtle difference in electrode location (both stimulating and EMG), or changes in skin resistance.
Whilst subtle difference in EMG electrode location or changes in skin resistance were minimised with normalisation of the evoked responses to M max (Lanza, Balshaw, & Folland, 2017), any slight difference in stimulation electrode location might have resulted in subtle differences in the activation site of lumbar spinal segments and be responsible for greater variability of LEPs observed (Table 1). Presumably, however, these factors would have affected both the training and the control group, such that had there been any observable changes in spinal excitability following 4 weeks of training, they would have been greater than the variability observed by the control group.
The present study utilised a low contraction intensity during TMS and lumbar stimulation, using stimulus intensities which evoked responses of 15-25% M max in order to prevent fatigability induced by the measurements and because this stimulus paradigm has been shown to be sensitive to strength training-induced alterations in corticospinal excitability (Griffin & Cafarelli, 2007;Siddique et al., 2020). Nevertheless, given the low contraction/stimulus intensities used when evoking neural responses, it is likely that these measurements reflect the excitability of low-threshold motor units, and adaptations to high-threshold motor units could thus have gone undetected. However, there is currently little evidence to suggest that motor unit adaptations which occur during strength training are threshold-specific (Enoka, 2019), with studies displaying similar adaptations to both low-and high-threshold motor units following strength training interventions (Del-Vecchio et al., 2019;Van Cutsem, Duchateau, & Hainaut, 1998), though these interventions differed to our own. Thus, while we believe our methods were sufficient to detect corticospinal alterations in the VL had they occurred, further work is required to determine whether motor unit adaptations following lower-body compound resistance exercise are threshold-specific.
Future studies should also consider constructing the stimulusresponse curves using a range of stimulation intensities (Rosenkranz, Williamon, & Rothwell, 2007), as limited evidence suggests higher intensities might be required to detect changes in neural function following skill learning (Kleim, Kleim, & Cramer, 2007).

SUMMARY AND CONCLUSION
This study assessed the acute and short-term neural responses to whole body resistance training at multiple levels of the corticospinal pathway, in non-specific and, for the first time, task-specific tasks. We hypothesised measuring responses in a novel task-specific assessment task would allow for a more sensitive assessment of the corticospinal adaptation to resistance training, but this hypothesis was rejected. For the immediate response to resistance exercise, an increase in spinal excitability was demonstrated, but this was not task-specific. After a period of resistance training, there were marked increases in task-specific strength, but no change in muscle thickness.
This absence of increase in muscle thickness of VL in the presence of a task-specific strength increase implies that neural adaptation was responsible, but surprisingly there were no changes in intracortical, corticospinal, or spinal responses in both specific and nonspecific tasks. The results of the present study therefore suggest that alterations in the corticospinal tract, when measured in the VL, might not contribute to task-specific improvements in strength following lower-body compound resistance exercise. Further work is required, including concomitant assessment of synergist muscles and consideration for other descending tracts, to gain a more holistic understanding of CNS adaptations to lower-limb compound resistance training.