Determining intracortical, corticospinal and alpha motoneurone excitability in athletes with patellar tendinopathy compared to asymptomatic controls

Lower capacity to generate knee extension maximal voluntary force (MVF) has been observed in individuals affected with patellar tendinopathy (PT) compared to asymptomatic controls. This MVF deficit is hypothesized to emanate from alterations in corticospinal excitability (CSE). The modulation of CSE is intricately linked to the excitability levels at multiple sites, encompassing neurones within the corticospinal tract (CST), intracortical neurones within the primary motor cortex (M1), and the alpha motoneurone. The aim of this investigation was to examine the excitability of intracortical neurones, CST neurones, and the alpha motoneurone, and compare these between volleyball and basketball athletes with PT and matched asymptomatic controls.


| INTRODUCTION
Patellar tendinopathy (PT) is a common condition, affecting 14% of elite athletes and 9% of recreational athletes. 1t is most frequently observed in athletes who participate in sports that require explosive jumping activity, for example volleyball or basketball where as many as 45% and 32% of athletes, respectively, are affected. 1A prominent feature of athletes with PT is reduced knee extension maximal voluntary force production (MVF). 2 The mechanism driving reduced MVF remains unclear, however, it likely involves subtle changes in the excitability and synaptic efficacy between intracortical neurones within the primary motor cortex (M1), corticospinal tract (CST) neurones and alpha motoneurones innervating motor units.This is supported by investigations utilizing the interpolated twitch technique (ITT) in PT, 3,4 which measures the extent of motor unit activation 5 through electrical stimulation of a target muscle that is already under maximal voluntary activation.Reduced corticospinal excitability (CSE) would limit MVF, and consequently electrical stimulation may recruit additional and previously unengaged motor units.Both investigations utilizing the ITT in PT found a deficit in MVF compared to the force generated when the electrical stimulation was superimposed, 3,4 indicating the presence of altered excitability at site/s that contribute to corticospinal excitability (CSE), 6 that could underlie MVF deficits.Therefore, it is indicated to explore excitability at these sites, to gain a detailed understanding of the contributions of each site to altered CSE.
Different sites that contribute to CSE can be studied non-invasively using electrophysiological techniques, such as transcranial magnetic stimulation (TMS) [7][8][9] and peripheral nerve electrical stimulation. 10,11While a small number of studies have utilized neurophysiological techniques in PT, these have focused on CSE, 3,4,12 or intracortical inhibition in isolation. 13None have systematically investigated excitability across the different sites that influence CSE in athletes with PT, which may provide important insight for mechanisms underlying reductions to knee extension MVF.For example, single-pulse TMS can be used to assess the excitability of CST neurones.The amplitudes of motor-evoked potentials (MEPs), which are the primary outcome obtained through single-pulse TMS, serve as a general measure of CSE.Notably, a study by Rio et al. 12 reported greater CSE (hyper-excitability) in people with PT.It is possible that this was an adaptation to enhance the drive to the motoneurone pool, to preserve motor function.However, the amplitude of MEPs obtained through TMS do not offer detailed information about the specific neural circuits or pathways involved, and variations could arise from modifications in the excitability of different sites within the CST, including intracortical neurones within the M1 and CST neurones themselves, and alpha motoneurones. 14tilization of additional outcomes can provide more detailed insight for modifications at specific sites when considered alongside MEP findings.When TMS is administered during ongoing muscle activity, a momentary disruption occurs in the continuous surface electromyography (sEMG) signal, signifying a pause in neural drive which is accompanied by a withdrawal of descending input to the motoneurone pool. 15,16This phenomenon is known as the silent period (SP), 17 and its duration can serve as an indicator of corticospinal inhibition regulated by the neurotransmitter gamma-aminobutyric acid-B (GABA-B ). 17A limitation of the study by Rio et al. 12 was the absence of SP duration assessment.Given that the efficiency of neural transmission along the CST and the degree of corticospinal inhibition appear to be important for activating the motoneurone pool, changes in both CSE and inhibition could be mechanisms altering MVF in athletes with PT.At present, it remains unclear whether the elevated CSE reported by Rio et al. 12 served as a compensatory response to a prolonged SP.
Paired-pulse TMS can be used to facilitate the assessment of distinct intracortical circuits in the M1. 7pecifically, it enables the investigation of short interval intracortical inhibitory pathways (SICI).A lower SICI ratio indicates increased GABAergic activity, especially GABA-A , which results in heightened synaptic efficacy between inhibitory intracortical neurones and CST neurones, leading to the inhibition of CSE and a reduction in motoneurone activation. 18Rio et al. 13 investigated SICI in a population with PT and reported diminished inhibition after isometric strength training.However, a limitation of this study is that it remains unclear whether intracortical inhibition is a symptom of PT, as SICI was not compared to controls without PT.This distinction is crucial to establish, as the assessment of both SP and SICI could potentially unveil whether neurotransmitters like GABA-A or GABA-B are predominantly affected in populations with PT.This insight could provide a mechanism that explains a deficit in knee extension MVF.
Finally, techniques exist that utilize peripheral nerve electrical stimulation to study excitability of the alpha motoneurone and lumbar spine nerve roots. 10,11In contrast to aforementioned ITT, which measures CSE, electrical stimulation while resting can determine excitability of subcortical components.In particular, techniques are well established that allow stimulation of the alpha motoneurone to measure excitability of peripheral components, 11 while novel techniques also exist that allow investigation of spinal motoneurone excitability. 10To date, no study has included electrical stimulation outcomes to investigate subcortical excitability in a population with PT.
Our primary aim was to evaluate whether alterations in excitability exist in intracortical neurones within the M1, CST neurones, or alpha motoneurones, between volleyball and basketball athletes with PT and asymptomatic controls.Additionally, we aimed to investigate if there are any associations between MVF, neurophysiological properties, or demographic or activity outcomes.

| Study design and setting
This cross-sectional study is reported in line with STROBE recommendations for cross-sectional case-control studies.The study was approved by the Monash University Human Ethics Committee (MUHEC project ID: 19892) and we collected written informed consent from all participants.
We recruited participants between June 2020 and January 2023, and collected data across the same period of time.Participants attended a single, 1.5-h testing session at the Monash Exercise Neuroplasticity Research Unit research laboratory (Monash University Peninsula Campus, Melbourne, Victoria, Australia).

| Recruitment and screening
Participants were recruited from the community and subelite basketball and volleyball organizations in Melbourne, Australia, through Monash University departmental and social media advertisements.We used identical strategy to recruit both the PT and asymptomatic control participants.We screened initial eligibility of potential participants via telephone, followed by a physical examination performed by a qualified physiotherapist (PV) prior to the testing session.

| Eligibility criteria
We recruited basketball and volleyball athletes aged 18 years or older to participate in this study.The PT group had current PT symptoms based on pain localized to the inferior pole of the patella upon palpation (both reported and observed through palpation), along with the following participant reported features: (i) pain during knee-extension loading activities (e.g., running, jumping, walking down stairs), specifically at the inferior pole of the patella; (ii) gradual onset of symptoms, and a minimum symptom duration of 3 months; and (iii) increased pain intensity after periods of inactivity, such as sitting or upon waking.Ultrasound imaging (Mindray M7, Shenzhen, China) was performed by the trained physiotherapist (PV), to confirm the presence of typical signs of tendinopathy including hypoechoic regions and thickening (>0.40 mm) at the proximal patellar tendon.Participants in both groups were required to be actively participating in basketball or volleyball at the time of testing.
We excluded participants from either group if they had sustained an injury to any body region at the time of testing (except for the PT in the PT group), or within the previous 6 months if the injury had lasted longer than 1 week or required consultation with a healthcare professional.Additional exclusion criteria included a history of patellar tendon rupture or surgery, recent injection of corticosteroids, platelet-rich plasma, or other pharmaceutical agents into the patellar tendon or its surrounding area within the last 6 months, or a diagnosis of inherited connective tissue disorders or type I or II diabetes.People with neurological conditions, a history of severe headaches or migraines, or the presence of metal in the cranial region were also excluded.Participants were also required to complete the TMS safety questionnaire and fulfill all specified criteria. 192.3 | Procedure and outcome measures

| Force recording and MVF
A summary of the experimental procedures is provided in Figure 1.All participants were positioned in a custom chair designed for assessing isometric knee extension force.The knee and hip joints were positioned to angles of 60° and 90°, respectively.The affected/most affected limb was securely fastened to a metal support, positioned approximately 25% of the tibial length above the talocrural joint (measured from the lateral malleolus to the tibial tuberosity).In the control group, the dominant lower limb was consistently tested instead.A calibrated S-Type load cell (Sparkfun, Niwot, Colorado, USA) was attached the test limb to the chair, posterior and perpendicular to the tibia to resist knee extension force.
The force signal was sampled at a rate of 2048 Hz and captured using a PowerLab 4/26, which was connected to a single-channel Bridge Amplifier (ADInstruments; Bella Vista, Australia).This setup facilitated real-time force feedback display to the participants, ensuring synchronization of recordings for subsequent measurements of M1, CST and alpha motoneurone excitability.
Participants practiced knee extension MVF twice before performing a minimum of two recorded trials.We allowed a minimum rest of 30 s between each MVF attempt, and participants could rest longer if they reported they were not fully recovered.If the MVF force differed by more than 5%, participants were required to perform an additional MVF trial, and this was repeated until two trials were within 5% (with one of these being the maximum).We provided identical instructions prior to each MVF (both practice and recorded trials), encouraging participants to extend their knee 'hard and fast', and to sustain maximum knee extension force for 3 s or until instructed to relax.We also provided strong verbal encouragement during every trial.We recorded the highest knee extension force (in kilograms [Kg]) achieved across the recorded trials as the MVF.

| Surface electromyography (sEMG)
We shaved the designated electrode placement area to remove fine hair, and then treated this with abrasive skin gel to eliminate dead skin cells.Subsequently, we sanitized the area using 70% isopropyl alcohol.Surface electromyography (sEMG) recordings were performed on the Rectus Femoris (RF) muscle using bipolar Ag-AgCl electrodes which we placed over the RF muscle following identification using SENIAM guidelines, 20 with the ground electrode positioned over the patella.The sEMG signals were amplified by a factor of 1000, underwent bandpass filtering (high pass at 13 Hz, low pass at 1000 Hz), and were digitized in real-time at a rate of 2 kHz.These signals were then recorded for a duration of 1 s and subsequently analyzed using Power Lab 4/26 software (ADInstruments, Bella Vista, Australia).

| Transcranial magnetic stimulation
We administered both single-pulse and paired-pulse TMS targeting the M1 using a concave double-cone coil connected to a Magstim 200 2 magnetic stimulator (Magstim Co., Ltd, Whitland, UK).This approach enabled us to assess the excitability of the intrinsic circuits of the M1 and CST. 7The M1 'hotspot,' generating the highest MEP amplitude in the RF muscle, was determined using the Rossini-Rothwell method. 21This hotspot was then consistently selected as the stimulation site for subsequent TMS assessments.To determine the active motor threshold (AMT), we identified the lowest magnetic stimulator output (MSO) intensity that induced a MEP greater than 200 μV, exhibiting a distinct latency and SP, within the RF muscle during 20% of knee extension MVF.To identify this, we started at 70% MSO, and increased or decreased by 2% MSO until we found the lowest MSO that achieved this criterion for at least 6 out of 10 consecutive stimuli. 21tilizing single-pulse TMS, CSE and inhibition SICI were assessed.Participants performed an isometric knee extension task at 20% of their MVF, during which we administered 10 stimuli to the predetermined M1 hotspot at 130%, 150% and 170% of AMT.Real-time force feedback was provided by displaying the target force (20% MVF) on a monitor, facilitated through Lab Chart software connected to a PowerLab 4/26 device (ADInstruments, Bella Vista, Australia), positioned approximately two meters away from the participant at eye level.We quantified CSE by calculating the peak-to-peak amplitude of MEPs elicited at each AMT intensity, which was averaged and normalized to the mean of the M MAX .To assess corticospinal inhibition (SP), the duration of sEMG inactivity following each MEP was determined, spanning from the onset of the MEP to the resumption of voluntary sEMG, 22 which was visually inspected to determine.This was measured for all 10 single-pulse stimuli at 130%, 150% and 170% AMT, and the mean calculated for each respective AMT intensity stimuli.
For the assessment of intracortical inhibition, we employed short-interval cortical inhibition (SICI).Using pulse TMS, we applied a subthreshold conditioning stimulus (80% of AMT), followed by a test stimulus (130% of AMT) with a 3 ms inter-stimulus interval. 23This procedure was performed while participants performed isometric knee extension at 20% MVF, and the sequence was repeated 10 times.SICI was calculated using the formula SICI = ((conditioned MEP)/(test MEP)) × 100. 24In total, including single and paired pulse protocols, 40 MEP were elicited (in addition to a moderate number required to locate the M1 hotspot and identify the AMT).

| Maximum compound muscle action potential
Electrical stimuli were administered to the femoral nerve using a constant-current stimulator (Digitimer DS7AH) with a pulse duration of 1 ms.This stimulation was facilitated through self-adhesive surface electrodes (CF3 200; Valutrode Electrodes, Melbourne, Australia).The cathode was situated over the femoral nerve within the femoral triangle, while the anode was positioned between the greater trochanter and iliac crest.Gradually increasing the stimulation intensity in 20 mA increments, we reached an intensity where M MAX (maximum muscle response) reached a plateau.Subsequently, the intensity was increased by an additional 20% to ensure that the stimulation remained supramaximal. 11The largest sEMG response produced across three supramaximal applications was used as the M MAX .

| Central activation ratio
To assess central activation ratio (CAR), we applied supramaximal electrical stimulation (matched to the intensity used to assess M MAX ) with a pulse width of 200 μs (DS7A, Digitimer, Hertfordshire, UK) to the femoral nerve (as per M MAX protocol) while participants attempted knee extension MVF.This electrical stimulation was applied immediately upon plateau of force, within close proximity to MVF. 25 This procedure was repeated three times.
We calculated CAR by measuring the MVF achieved and then comparing it to the force achieved via supramaximal electrical stimulation (superimposed force).CAR was calculated using the following formula 25 : CAR (%) = (MVF/superimposed force) × 100.

| Electrical stimulation of the spinal tract
To assess motoneurone excitability at the spinal level, we investigated LEPs by administering electrical stimulation to the lumbar spine while participants were in a resting state.Using self-adhesive electrode pads measuring 50 mm × 50 mm (TENS Machines Australia, West End, Queensland, Australia), we positioned the cathode over the spinous processes of the first lumbar vertebra and placed the anode over the eighth thoracic vertebra.Starting with a submaximal intensity, we gradually increased the electrical stimulation until the peak-to-peak amplitude of the RF sEMG reached a level within 5-10% of the M MAX .Once this threshold was achieved, we repeated the stimulation three times at this intensity to ensure the consistency and reliability of the collected data. 26As stimulator intensity was increased, the onset latency of the LEP was monitored for potential activation of ventral roots, 26 as increasing stimulus intensity would be expected to result in shorter onset latency. 26The recorded stimulator intensity (in mA) was measured as the LEP, which served as a measure of spinal motoneurone excitability. 10

| PT pain severity
Participants in the PT group completed standardized single leg decline squat (SLDS), as region specific painprovocative tendon-loading activity is recommended to measure the severity of pain symptoms in tendinopathy conditions. 27Standing on a 30° decline board on the affected/most affected limb, participants completed two repetitions of a single leg squat to a depth of 90° knee flexion, whilst maintaining an upright trunk throughout the task. 28We asked participants to rate the worst pain experienced at 90° knee flexion, on the 11-point numerical pain rating scale (0 = no pain and 10 = the worst pain imaginable).

| Questionnaire outcomes
We evaluated the severity of PT using the Victorian Institute of Sport Assessment-Patellar questionnaire (VISA-P), which employs a scale ranging from 0 to 100.A score of 100 indicates the absence of patellar tendon pain or functional limitations.The VISA-P is a reliable and condition-specific tool designed to assess the severity of PT-related issues. 29o examine participants' activity history, we used the 7-Day Physical Activity Recall survey.This survey required participants to recall and report their engagement in moderate, hard, and very hard activities during the week preceding their participation. 30We used this collected information to calculate metabolic equivalents (METs).

| Sample size calculation
We calculated the required sample size based on CAR data from Davi et al. 3 This study was used as it provided the most conservative estimate for between group difference (Cohen's d = 1.1) for a neurophysiological outcome included in the current study, that compared people with PT to asymptomatic controls.To detect a similar between group difference in CAR (Cohen's d = 1), with power of 80% and the alpha set at 5%, we required a total of n = 17 case and n = 17 control participants.

| Statistical analysis
Statistical analyses were conducted using GraphPad Prism (Version 9.4.1,GraphPad Software, Boston, USA).The normality of distribution for all collected data was assessed utilizing the Shapiro-Wilk test.Demographic, activity (METs), MVF, and neurophysiological data were examined for between-group differences.Unpaired t-tests were applied to normally distributed data, while Mann-Whitney tests were used for non-normally distributed data (age, M MAX electrical stimulator output, CAR, and METs).Additionally, Fisher's exact test was used to examine dichotomous variables (sport [volleyball or basketball] and sex [female or male]) for between group difference. 31orrelations were run to explore associations for outcomes observed to be different between groups, to neurophysiological, demographic, activity and MVF data, which was pooled from both groups.Additionally, we ran associations to investigation links with CAR, as it served as the primary outcome.Pearson's correlation coefficient was used for normally distributed data, and Spearman's Rho when data were non-normally distributed, with thresholds set to negligible (r < 0.3), low (0.3-0.5), moderate (0.5-0.7), high (0.7-0.9) and very high (>0.9). 32tatistical significance was defined as p < 0.05 for all analysis.All data is presented as individual data plots, while group data are presented as mean (standard deviation) when normally distributed and median (interquartile range) when non-normally distributed, with betweengroup differences as mean difference (95% confidence intervals [95% CI]).Standardized mean difference (SMD) for between group differences was calculated using the formula for Cohen's d = (mean 1 − mean 2)/√((SD 1 2 + SD 2 2 )/2), with its 95% confidence interval calculated as d ± 1.96 × SEM (35).

| RESULTS
Nineteen basketball or volleyball athletes with PT and 18 asymptomatic control athletes participated.There were no reported adverse effects among the participants due to their involvement.In the PT group, participants experienced symptoms for a mean (SD) duration of 49.3 (40.9) months, and n = 2 experienced unilateral symptoms while n = 17 experienced bilateral symptoms.Mean (SD) pain (NRS; 0-10) during SLDS was 5.6 (1.9), and PT-specific disability (VISA-P; 0-100) was 59.6 (10.1).Table 1 displays the demographic information and activity levels (METs), while Figure 1 displays maximal MVF data normalized to body weight (Kg/Kg).
When data for groups were pooled, we observed a moderate negative association between BMI and MVF (r = −0.571,p < 0.001), and a low positive association between BMI and the electrical simulator intensity required to elicit M MAX (r = 0.390, p = 0.017).and positively associated with MEP measured at 130% AMT (r = 0.541, p < 0.001).We also observed a moderate positive association between SICI and CAR (r = 0.544, p = 0.013).

| Motoneurone excitability and central activation ratio
Supplementary appendix 3 presents the mean data for electrical stimulation outcomes, with the mean differences between the two groups.The M MAX electrical stimulator output was 61 (95% CI 6-116; p = 0.016; d = 0.74 [95% CI 0.07-1.40])mA greater in the PT group compared to the asymptomatic control group (Figure 4).No significant differences were observed between the groups for CAR (p = 0.374); Figure 5 data for groups were pooled, we observed a low negative association between the electrical simulator intensity required to elicit M MAX and MVF (r = −0.464,p = 0.004), and a low positive association between M MAX amplitude and MVF (r = 0.400, p = 0.014).There was also a moderate positive association between CAR and MVF (r = 0.601, p < 0.001), a low positive association between the electrical simulator intensity required to elicit M MAX and BMI (r = 0.390, p = 0.017), and a moderate negative association between the electrical simulator intensity required to elicit M MAX and M MAX amplitude (r = −0.455,p = 0.005).

| DISCUSSION
Consistent with previous reports, [2][3][4] we found knee extension MVF was 25% lower in our PT group, while we observed altered neurophysiological properties at specific sites.Specifically, in the PT group, the electrical stimulation output intensity required to elicit M MAX was greater, and associated with lower MVF, while SP duration was longer.This suggests there may be subtle adaptations along the CST that influence neural drive to the motoneurone pool, and consequently impact knee extension MVF in basketball and volleyball athletes with PT.
The electrical stimulation output intensity required to achieve M MAX was 34% greater in our PT group.This result, assessed via supramaximal electrical stimulation of the femoral nerve, indicates changes to excitability between the site of electrical stimulation application and the site of recording. 11Specifically, this could be due to reduced excitability of the alpha motoneurone, or the sacrolemmal membrane, which would result in an increase recruitment threshold. 11Otherwise, synaptic transmission could be altered at the neuromuscular junction. 11respective of the exact subcortical driver, this novel finding would mechanistically reduce CSE.Notably, in our study, electrical stimulation output intensity was associated with MVF, whereby greater inhibition was related to lower MVF.This provides a plausible mechanism explaining lower knee extension MVF in jumping athletes with PT, whereby modification to neural drive directed toward the spinal motoneurone pool mechanistically leads to a reduction in MVF.
As far as the authors are aware, to date no studies have investigated M MAX amplitude or the electrical stimulation output intensity required to elicit M MAX , in any persistent pathology localized to the knee.A reduction in spinal reflex excitability has been reported in people with patellofemoral pain 33 and, consistent with our M MAX outcomes, this is considered to measure subcortical excitability. 34Interestingly, comparison can also made to a study that used an acute-pain experimental model, whereby pain was induced at the knee via injection, in healthy participants.That study demonstrated reduced subcortical excitability, also by assessing spinal reflexes. 35Based on these findings, the reduction to subcortical excitability we observed may be a typical neurophysiological response to painful pathology at the knee, whether this be acute or persistent.
Silent period duration measured using a test stimulus intensity of 130% AMT was greater (12 ms longer) in our PT group, which would mechanistically reduce CSE.The duration of the silent period is modulated by inhibitory mechanisms within M1, with spinal inhibitory processes, such as Renshaw cell inhibition, contributing during the initial 50-60 ms. 36Considering the duration SP was 86 ms in our PT group, this suggests that the increases in inhibition we observed are localized to the cortical level, which would be regulated by GABA-Bmediated inhibition. 36This indicates that PT specifically targets distinct intracortical inhibitory neurones which, through increased synaptic efficiency, reduce the excitability of CST neurones and consequently CSE.Whilst this modification to neural drive could mechanistically reduce MVF, we did not observe any association between SP and MVF.Therefore, it appears alterations to subcortical excitability may be more important to MVF in jumping athletes with PT, compared to supraspinal adaptations.
Similar increases in SP duration have been observed previously in a study that included people with chronic ankle instability, 37 consistent with our observations in athletes with PT.Interestingly, in people with knee osteoarthritis, one study observed a reduction to SP duration. 38When considered together, these findings highlight there may be different intracortical or corticospinal responses to persistent lower limb pathology.did observe a between group in when measured using a test stimulus intensity of 150% AMT.Notably, this measure could not be obtained at the higher stimulus intensity in a number of participants, and consequently the sample size was reduced.Wide confidence intervals suggest this outcome was not appropriately powered to identify group differences if they were present, and this could explain our null finding.This is further supported by a high positive association between SP duration at 130% and 150% AMT, which indicates participants with longer SP duration at one intensity are likely to exhibit longer SP duration at the other intensity.
Our study did not detect differences in SICI, a measure of intracortical neurone function influenced by GABA-A receptors localized within the M1. 18It is possible that only intracortical inhibitory neurones within the M1 mediated by GABA-B are altered in PT.Rio et al. 13 previously reported a reduction in SICI after isometric strength training in volleyball athletes with PT.This observation is not surprising, given that SICI decreases linearly as force output incrementally increases. 39Therefore, high-intensity contractions inherently result in the release of inhibition.Our study is the first to compare SICI between a population with PT and asymptomatic controls, and indicates SICI is not affected in PT.Notably, the application of a 20% MVF background contraction while we measured this outcome could have played a significant role in reducing GABA-A -mediated intracortical inhibition, and might explain the lack of difference noted between the PT and control groups. 40Consequently, the impact of PT on the excitability of SICI circuits remains unclear.
We did not identify any between group differences for our MEP outcomes.In contrast, a study by Rio et al. 12 reported an increased peak slope of a MEP stimulus-response curve in people with PT, suggesting increased excitability of the CST that may be an adaptive response to enhance drive to the motoneurone pool to preserve motor function.This discrepancy is likely due to differences in how the outcomes were calculated.The stimulus response curve used by Rio et al. 12 used the Boltzmann equation to examine the slope.In comparison, our MEP outcomes were collected at 130% AMT and 150% AMT.A limitation of that study was the absence of SP duration assessment.When considered with our finding of increased SP cortically, and reduced excitability subcortically, in the PT group, it is possible this increase to CST excitability served as a compensatory response to preserve motor function.
We did not find any between group difference in CAR.In contrast, Davi et al. 3 reported lower CAR in recreationally active individuals with PT (84.44%) compared to asymptomatic controls (97.93%).It is possible conflicting findings between the two studies are due to differences in the methodology utilized.The current study tested participants in a position of 60° knee flexion, while Davi et al. used 90°.Importantly, knee extension MVF is joint angle dependent, with greater muscle activation of knee extensor muscles at smaller angles. 41Alternatively, as we required participants to be actively participating in basketball or volleyball sports, it is possible our participants represent a sub-population that develop effective neurophysiological compensatory strategies to maintain drive to the motoneurone pool.Given our finding of higher electrical stimulation output intensity required to elicit M MAX , which was associated with lower knee extension MVF, the former appears more likely.Alternate central nervous system adaptations could be important to the deficits in knee extension MVF we observed in athletes with PT.For example, future research may focus on investigating the reticulospinal tract, which could provide insights for mechanisms underlying altered excitability that could consequently impact motor performance. 42Beyond MVF, the subtle neurophysiological adaptations we observed along the CST could also impact the efficiency of submaximal force motor control.Parallels can be drawn between our findings, and those reported by Ward et al. 43 in a population who had experienced anterior cruciate ligament injury.Notably, their case group exhibited longer silent period duration, while CSE was not affected, 43 and in a follow up study Ward et al., found altered force control. 44Future research in populations with PT may focus on the assessment of knee extension force control to determine if this is affected.We did not match groups for MVF, and this could have provided further insight for the nature of associations between the presence of PT, MVF and neurophysiological outcomes.At present, it is not clear whether the presence of PT is responsible for altered neurophysiological properties that reduce MVF or, alternatively, if people who are weaker have altered neurophysiological properties.A longitudinal study design could delineate the temporal relationship between these factors.
There was a between group difference for BMI, and this could have impacted results.For example, subcutaneous tissue thickness (which is largely subcutaneous fat), has been demonstrated to impact voluntary sEMG amplitude. 45However, we followed recommended procedure to account for this by normalizing MEP data to M MAX amplitude, as this can reduce the influence of adiposity. 45We observed a low positive association between BMI and the electrical stimulation output intensity required to achieve M MAX .This is consistent with one previous study that reported higher BMI is linked to lower M MAX amplitude, in healthy participants. 46diposity has been identified as an intrinsic risk factor to tendinopathy, and found to be associated with of tendon health, 47 recovery with rehabilitation. 48It's possible the higher in our PT group influenced the finding of greater electrical stimulation output intensity required to achieve M MAX .Notably, we also observed a moderate negative relationship between BMI and knee extension MVF, indicating there might be an interaction between adiposity and subcortical excitability, and MVF.Further research is indicated to identify the nature of this relationship, as it possible body mass could be an important rehabilitation target, to improve subcortical excitability and maximal strength.

| Perspective
Our findings highlight that there are alterations to excitability at specific sites in athletes with PT, that likely impact drive to the motoneurone pool.In particular, there is reduced excitability of the alpha motoneurone or sarcolemmal membrane, or reduced transmission at the neuromuscular junction, that is linked to lower knee extension MVF.Additionally, there is increased activity of GABA-B -mediated intracortical neurones.Our findings provide novel insight that improves knowledge for mechanisms that could contribute to lower knee extension MVF observed in PT.

F I G U R E 1
Displays an overview of the experimental procedure.

T A B L E 1
Demographic and activity characteristics for both groups, along with PT-specific data for the PT group (Mean [SD]).

F I G U R E 3 F I G U R E 4
, M MAX amplitude (p = 0.190), LEP F I G U R E 2 Displays both individual and mean values (SD) of knee extension MVF normalized to body weight (Kg/Kg) for the patellar tendinopathy (PT) group and the asymptomatic controls group.***Denotes significantly lower knee extension MVF in the PT group compared to control, p < 0.01.Displays both individual and mean values (SD) of the SP at 130% of the AMT for both the PT and asymptomatic control groups.*Denotes a significantly longer SP for the PT group compared to control group, p < 0.05.Displays both individual and median values (interquartile range) of the electrical simulator output intensity required to elicit M MAX in both the PT and asymptomatic control groups.* Denotes greater electrical stimulator output intensity to achieve M MAX for the PT group compared to asymptomatic control group, p < 0.05.output intensity (p = or LEP normalized to M MAX (p = 0.280).

F I G U R E 5
Displays individual and median (interquartile range) central activation ratio (CAR) in PT and asymptomatic controls groups.