Anterior cruciate ligament (ACL) injury is a common problem with consequences ranging from chronic joint instability to early development of osteoarthritis. Recent studies suggest that changes in brain activity (i.e., functional neuroplasticity) may be related to ACL injury. The purpose of this article is to summarize the available evidence of functional brain plasticity after an ACL injury. A scoping review was conducted following the guidelines of the Joanna Briggs Institute and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses. The terms “brain,” “activity,” “neuroplasticity,” “ACL,” “injury,” and “reconstruction” were used in an electronic search of articles in PubMed, PEDro, CINAHL, and SPORTDiscus databases. Eligible studies included the following criteria: (a) population with ACL injury, (b) a measure of brain activity, and (c) a comparison to the ACL-injured limb (contralateral leg or healthy controls). The search yielded 184 articles from which 24 were included in this review. The effect size of differences in brain activity ranged from small (0.05, ACL-injured
An anterior cruciate ligament (ACL) rupture is a traumatic knee injury typically affecting young and active people [
Optimising neuromuscular function is considered a key aspect in both prevention [
Recent publications support the notion of functional neuroplasticity in the brain of people with ACL injury [
Another area with serious knowledge gaps concerns the role of brain function as a potential target of rehabilitation following ACL injury. To our knowledge, no study has focused on brain activity changes throughout the rehabilitation process and how these are related to different outcome metrics. Nevertheless, some authors have advanced recommendations for therapies to be implemented with these patients based on observations on brain activity differences. Examples of these interventions range from transcutaneous electrical nerve stimulation (TENS) and cryotherapy [
The study of neuroplastic changes following ligament injuries is a fairly recent topic with a limited number of studies. Moreover, there is a considerable variety of methods and outcome measures in the study of brain activity which makes it difficult to pool data in a systematic review and meta-analysis format. Thus, a scoping review seems to be the most appropriate method to answer the following research questions: (a) what is the current evidence of differences in brain activity following an ACL injury? (b) What are the potential confounding variables that may influence brain activity following an ACL injury?
According to the guidelines from the Joanna Briggs Institute, the most appropriate method to address a research question involving emerging evidence is to conduct a scoping review [
Following the recommendations of the Joanna Briggs Institute for scoping reviews, different sources of information were considered. Hence, included studies could be from any type of primary research (randomized clinical trials, quasiexperimental studies, cohort studies, case-series studies, cross-sectional studies, case-control studies, and case studies), using human participants with ACL injury and an outcome of brain activity, as measured by electroencephalography (EEG), transcranial magnetic stimulation (TMS), or magnetic resonance imaging (MRI). Letters, commentaries, conference abstracts, and reviews were excluded. Eligible studies needed to include a population with ACL injury, with or without reconstruction; to include a measure of brain activity; and to include any form of comparison to the affected limb (e.g., a healthy control group or the contralateral limb).
The selected studies had to include participants with unilateral ACL injury, with or without reconstruction, males or females, and from any age category. Participants could present an isolated, primary or recurrent, ACL injury or have concomitant injuries (e.g., menisci or collateral ligament injuries), in addition to the ACL injury. No minimal or maximal time since injury was defined, and all levels of physical activity and knee function were accepted. Studies using people with isolated knee osteoarthritis or using artificially induced knee effusions were excluded.
For the purpose of this review, neuroplasticity was defined as the ability of the CNS to adapt and reorganize following a lesion or environmental change [
An electronic search was conducted in PubMed, CINAHL, SPORTDiscus, and Cochrane Central Register of Controlled Trials, with no restriction on language or dates of publications, between the months of May 2018 and August 2018. Two independent authors (TN and TS) conducted the search by screening the studies for eligibility after eliminating duplicates. Studies were first selected based on the title and abstract, and only afterwards the inclusion criteria were applied based on full text. The following keywords, and associations between them, were used during the search: brain (or cortical), activation (or activity), neuroplasticity, ACL, and injury (or reconstruction). A detailed syntax of the search can be found in the appendix. Results of the search were organised and presented according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Figure
Flow chart of the search strategy and results.
Data extraction of the selected studies was performed by two authors. The following data were extracted from the studies: (a) authors and year of publication; (b) type of study and level of evidence; (c) characteristics of population (e.g., gender, age, and side affected); (d) time since injury/surgery; (e) type of surgery (i.e., for people subjected to ACL reconstruction); (f) presence of other injuries (e.g., menisci, cartilage, or collateral ligaments); (g) physical activity and knee function level; (h) type of rehabilitation performed; (i) methods used to measure brain activity (e.g., EEG, TMS, and fMRI); and (j) task performed during brain activity measurements. In accordance with the Manual of JBI Scoping Reviews, the results of a scoping review should be presented in a diagrammatic or tabular form. This recommendation was followed, and data were presented in tables summarizing information on points (a), (b), (c), (d), (e), (h), (i), and (j). The remaining information was described in the text, in the Results section.
Cohen’s
A summary of the search strategy is depicted in Figure
Table
Summary of the techniques used to measure brain activity and clinical interpretations in ACL injury.
Technique | Measurement | Interpretation | |
---|---|---|---|
EEG | Somatosensory-evoked potentials (SEPs) | Peaks of activity are measured by electroencephalography (EEG) electrodes in the somatosensory cortex after an external stimulus is delivered to the common peroneal nerve or to the ACL (i.e., via arthroscopy) | The ascending stimulus to the somatosensory cortex, following common peroneal nerve stimulation, is detected as P27 component which provides information about the afferent system. Literature shows contradictory information regarding the ability to reproduce SEPs in people with ACL deficiency or reconstruction |
Spectral analysis | EEG signals are measured during a movement (i.e., joint angle or force reproduction). The mean absolute EEG spectral power is divided into different frequencies: delta (0–4 Hz), theta (4.75–6.75 Hz), alpha 1 (7–9.5 Hz), alpha 2 (9.75–12.5 Hz), beta (12.75–18.5 Hz), and gamma (30–80 Hz), corresponding to different levels of activity in different areas of the cortex | It has been suggested that differences in theta power in the frontal cortex may be linked to differences in working memory and focused attention, whereas alpha power is typically inversely related to the neuronal activation. As such, increased alpha power recorded over parietal cortical areas may be interpreted as a deactivation of the somatosensory cortical areas | |
TMS | Motor-evoked potentials (MEPs) | Muscle response (measured by electromyography), following a transcranial magnetic stimulus (TMS) delivered at the motor cortex travelling down the motor pathways | Decreased MEPs represent less information travelling in the motor pathways to the target muscle |
Motor threshold | Minimum transcranial magnetic stimulation (TMS) intensity necessary to cause a response (MEP) in the target muscle—it is a measure of motor cortex excitability and can be measured at rest (i.e., resting motor threshold), or during an activity (i.e., active motor threshold (AMT)) | Motor threshold is inversely related to motor cortex excitability, meaning that people with reduced corticomotor excitability would have a higher motor threshold. A reduction in motor cortex excitability may affect motor output | |
Intracortical inhibition (SICI and LICI) | Paired TMS pulses (first, a conditioning subthreshold pulse, followed by a suprathreshold testing pulse) are delivered with varying interstimulus intervals. Short intervals (<5 ms) produce short-interval intracortical inhibition (SICI), while longer intervals (>50 ms) produce long-interval intracortical inhibition (LICI) | SICI is associated with GABAa activity, while LICI is associated with GABAb activity. Higher levels of intracortical inhibition may be associated to lower cortical excitability | |
Cortical silent period (CSP) | The cortical silent period (CSP) corresponds to an interruption in voluntary electromyography (time from MEP onset to EMG activity resumption) in the target muscle following TMS. CSP is mediated by GABAb activity at a cortical level | Longer CSP represent higher levels of inhibition, which may lead to muscle inhibition. However, a link between CSP and MEP changes has not been established | |
Intracortical facilitation | Similar to intracortical inhibition measurements, paired TMS pulses are used for measuring intracortical facilitation. In this case, a 7 to 30 ms interval between the conditioning and testing pulses is used | Cortical facilitation is mediated by neurotransmitter glutamate onto non-N-methyl-D-aspartate receptors. There is conflicting evidence on whether ICF is changed in people with ACL injury or reconstruction | |
MRI | Functional MRI during a task | The blood oxygen level-dependent signal is quantified through the blood hemodynamics during a specific task (e.g., knee flexion-extension cycles) | An increased BOLD signal is associated to a higher activity of the respective brain area, which may be associated to reduced efficiency of these cortical regions, in people with ACL injury |
Summary of included EEG and fMRI studies (effect size is presented for between or within-group comparisons).
Study | Level of evidence | Group ( |
Type of surgery; time from injury/surgery | Equipment, outcomes | Task | Results | Effect size, Cohen’s |
---|---|---|---|---|---|---|---|
Baumeister et al. [ |
Case-control, 3b | ACLR ( |
All hamstrings; |
EEG, power spectral analysis | Knee extension force reproduction (50% of MVIC) | Significantly higher frontal theta power in ACLR | ACLR vs. healthy, |
Ochi et al. [ |
Case-control, 3b | ACLD ( |
All hamstrings; >13 months after surgery in 38 ACLR participants | EEG—SEP of the ACL | Direct mechanical stimulation of the ACL during arthroscopy (under general anaesthesia) | Mechanically reproduced SEPs were observed in 58% of ACLD, 86% of ACLR, and 100% of healthy ACL |
ACLD vs. ACLR, |
Ochi et al. [ |
Case-control, 3b | ACLD ( |
Hamstring graft in 22 patients and 1 allogeneic fascia lata graft; >18 months after surgery | EEG—SEP of the ACL | Electrical stimulation of the ACL during arthroscopy (under general anaesthesia) | Reproducible SEPs in 47% of ACLD, 96% of ACLR, and 100% of healthy ACL |
ACLD vs. ACLR, |
Miao et al. [ |
Case-control, 3b | ACLD ( |
EEG, power spectral analysis | EEG was recording during the following: |
The ACLD group showed a significant increase in band power of all frequencies, during all tasks | ACLD vs. healthy | |
Valeriani et al. [ |
Case-control, 3b | ACLD ( |
Between 12 and 96 months after injury | EEG—SEP of the common peroneal nerve and posterior tibial nerve | Patients relaxed in supine | Seven subjects from the ACLD group showed SEP abnormalities (loss of P27) after common peroneal nerve stimulation | Unable to determine |
Valeriana et al. [ |
Case-series, 4 | ACLR ( |
All patellar tendon; time from surgery/injury unknown | EEG—SEP of the common peroneal nerve | Patients relaxed in supine | Absence of cortical P27 response in the injured limb before, and after, ACL reconstruction surgery | Unable to determine |
Baumeister et al. [ |
Case-control, 3b | ACLR ( |
All hamstrings; |
EEG, power spectral analysis | Reproduce a given knee angle of 40° | Significantly higher theta and alpha 2 power in ACLR | Unable to determine |
Courtney et al. [ |
Case-control, 3b | 17 ACLD patients (7M, 10F), divided in the following: |
Overall |
EEG—SEP of the common peroneal nerve | Patients relaxed in supine | The adapter group showed normal SEPs, 75% of noncopers had normal SEPs, and all copers had altered SEPs | Unable to determine |
Courtney et al. [ |
Case-control, 3b | 15 ACLD patients (5M, 10F, |
Overall mean = 67 months after injury: noncopers: 85 months, adapters: 63 months, and copers: 69 months | EEG—SEP of the common peroneal nerve | Patients relaxed in supine | The adapter group showed normal SEPs, 75% of noncopers had normal SEPs, and all copers had altered SEPs | Unable to determine |
Lavender et al. [ |
Case-control, 3b | 11 patients: 4 with intact ACL, 6 with complete rupture, and 1 with partial rupture. No information on sex and age | 28 months ( |
EEG—SEP of the ACL | Electrical stimulation of the ACL during arthroscopy | All intact ACLs (and the partially ruptured) showed reproducible SEPs; ruptured ACL did not show reproducible SEPs | Unable to determine |
Kapreli et al. [ |
Case-control, 3b | ACLD ( |
fMRI | Cycles of 45° knee extension/flexion (1.2 Hz), during 25 s, positioned in supine inside the scanner | ACLD showed less activation of thalamus, PP, PM, cerebellum, iSM1, cSM1, BG GPe, and CMA and showed higher activation of pre-SMA, SIIp, and pITG | Unable to determine | |
Grooms et al. 2017 [ |
Case-control, 3b | ACLR ( |
13 hamstrings and 2 patellar tendons; |
fMRI | ACLR showed less activation of iMC and cerebellum and showed higher activation of cMC, lingual gyrus, and iSII |
ACLR = anterior cruciate ligament reconstruction; ACLD = anterior cruciate ligament deficiency; BG GPe = basal ganglia-external globus pallidus; CMA = cingulated motor area; cMC = contralateral motor cortex; cSM1 = contralateral primary sensorimotor area; EEG = electroencephalography; ES = effect size; F = females; fMRI = functional magnetic resonance imaging; iMC = ipsilateral motor cortex; iSM1 = ipsilateral primary sensorimotor area; iSII = ipsilateral secondary somatosensory area; M = males; pITG = posterior inferior temporal gyrus; PM = premotor cortex; PP = postparietal cortex; pre-SMA = presupplementary motor area; SII = secondary somatosensory area; SEPs = somatosensory-evoked potentials; SIIp = posterior secondary somatosensory area.
Summary of included TMS studies (effect size is presented for between or within-group comparisons).
Study | Level of evidence | Group ( |
Type of surgery; time from injury/surgery | Outcomes | Task | Results | Effect size, Cohen’s |
---|---|---|---|---|---|---|---|
Pietrosimone et al. [ |
Case-control 3b | ACLR ( |
14 hamstrings, 12 patellar tendons, 2 allografts; |
AMT | Vastus medialis contraction at 5% MVIC | AMT was significantly higher in the ACLR limb ( |
ACLR vs. uninvolved limb, |
Pietrosimone et al. [ |
Case series 4 | ACLR ( |
Unknown; |
AMT | Vastus medialis contraction at 5% MVIC | The ACLR limb presented average AMT values of |
Unable to determine |
Lepley et al. [ |
Case-control 3b | ACLR ( |
Unknown; |
AMT | Vastus medialis and lateralis contraction at 5% MVIC | The ACLR group showed higher values of AMT ( |
ACLR vs. healthy, |
Lepley et al. [ |
Case-control, 3b | ACLR ( |
Nine hamstrings, 11 patellar tendons |
AMT, MEP | Vastus medialis contraction at 5% MVIC | Both at presurgery and 2 weeks after surgery, there were no differences between groups in the AMT values |
For AMT: |
Ward et al. [ |
Case-series, 4 | ACLD ( |
Unknown; |
AMT | Vastus medialis contraction at 5% MVIC | There were no significant differences between the ACLD limb ( |
ACLD vs. uninvolved, |
Kuenze et al. [ |
Case-control, 3b | ACLR ( |
12 hamstrings, 10 patellar tendons; |
AMT | Vastus medialis contraction at 5% MVIC | The ACLR limb showed a significantly higher AMT ( |
ACLR vs. uninvolved, |
Luc-Harkey et al. [ |
Case-series, 4 | ACLR ( |
18 patellar tendons (remaining unknown); |
AMT, ICF, SICI, MEP | Vastus medialis contraction at 5% MVIC | No significant differences in AMT were observed between the ACLR ( |
For AMT: |
Norte et al. [ |
Case-control, 3b | ACLR ( |
34 hamstrings, 29 patellar tendons, 9 allografts; |
AMT | Vastus medialis contraction at 5% MVIC | AMT values between the ACLR limb ( |
ACLR vs. uninvolved, |
Norte et al. [ |
Case-control, 3b | ACLRearly ( |
29 hamstrings, 26 patellar tendons, 23 allografts; |
AMT | Vastus medialis contraction at 5% MVIC | No significant differences in AMT were found between the ACLR limb and the uninvolved limb, both at early ( |
All ACLR groups vs. uninvolved limb |
Zarzycki et al. [ |
Case-control, 3b | ACLR ( |
Eight hamstrings, 5 patellar tendons, 3 allografts; |
ICF, MEP, SICI, and RMT | Participant seated in dynamometer and relaxed | No significant differences were found in RMT between the ACLR limb ( |
For RMT: |
Ward et al. [ |
Case-control, 3b | ACLR ( |
Unspecified; |
AMT, CSP, ICF, LICI, MEP, and SICI | Rectus femoris contraction at 10% MVIC | Differences in AMT between the ACLR limb ( |
For AMT: |
Héroux and Tremblay [ |
Case-control, 3b | ACLD ( |
22 (range 4-108) months from injury | RMT and MEP | Quadriceps contraction for MEP recordings (details unspecified) | RMT values from the ACLD limb were significantly lower ( |
Unable to determine |
%T = percentage of 2.0 tesla; ACLD = anterior cruciate ligament deficiency; ACLR = anterior cruciate ligament reconstruction; AMT = active motor threshold; CSP = cortical silent period; ES = effect size; F = females; ICF = intracortical facilitation; LICI = long-interval intracortical inhibition; M = males; MEP = motor-evoked potential; MVIC = maximal voluntary isometric contraction; RMT = resting motor threshold; SICI = short-interval intracortical inhibition.
From the 24 studies, 14 assessed participants with ACL reconstruction (ACLR), six assessed participants with ACL deficiency (ACLD), and two studies included both individuals with ACLR and ACLD. A total of 629 participants with ACL injury were included, 47.3% male and 52.7% female participants—three studies [
The average time since surgery in individuals with ACLR was
The types of surgery for ACL reconstruction were as follows: (a) graft from the hamstring muscles [
In the selected studies, brain activity was measured by the following: (a) fMRI, during a dynamic task (i.e., cycles of knee flexion-extension) [
Physical activity of the participants with ACL injury at the time of brain activity measurements was assessed by the Tegner scale in 11 studies, ranging from 4.5 (representing a moderately heavy labor) [
The International Knee Documentation Committee questionnaire was the most commonly used assessment of knee function. Scores of knee function ranged from 77.2 out of 100 [
A specific description of the ACL injury and the presence of concomitant injuries can be found in 16 studies. Five studies [
Only 4 out of the 24 studies provide information about rehabilitation performed by people with ACL deficiency or reconstruction. The study of Lepley et al. [
The ES of differences in brain activity between limbs and/or populations could be determined in 15 out of the 24 studies included. Overall, ES ranged from small to large (see Tables
In 8 studies, all using TMS, it was possible to determine the ES of differences in brain activity between the ACLR and the uninvolved limb. The effect size ranged from small (
The ES of differences in brain activity between the ACLR limb and a matched healthy limb was determined in 12 studies (8 using TMS, 3 using EEG, and 1 using fMRI). In the TMS studies, the ES of differences in motor threshold ranged from small (
Only 1 study [
Two studies [
In this scoping review, we included and analysed articles to summarize the current evidence related to the differences in brain activity and corticospinal excitability shown by people with ACL injury. The results from this review (summarized in Figure
Infographic summarizing the evidence of brain activity changes based on three different measurement techniques in people with ACL injury (legend: ACL: anterior cruciate ligament; EEG: electroencephalography; fMRI: functional magnetic resonance imaging; SEP: somatosensory-evoked potential; TMS: transcranial magnetic stimulation; arrows represent “increase” or “decrease”).
Evidence from the included studies supports the notion of changes both at a sensory and at a motor level after ACL injury. Several studies using different techniques to measure brain activity showed higher levels of activity recorded in the motor areas [
Regarding the sensory areas of the cortex, the studies included in this review show evidence of functional neuroplasticity in participants with ACL injury or reconstruction. This may be explained by the disruption in the somatosensory information from the injured ACL [
Several variables were identified in this review (Tables
The outcome measures of brain activity were heterogeneous across the studies (with the exception of both fMRI studies [
One aspect that some studies failed to properly control concerns the characteristics of the population. Variables such as the gender and age of the participants, type of surgery performed in the ACLR patients (Tables
Another aspect that should be better described in future studies is related to knee function and if the person is classified as
Moreover, very little information was reported about rehabilitation protocols, which is concerning given that several studies [
The lack of information regarding the type of activities practiced on a regular basis and the level of motor skill prior to ACL injury was also another considerable limitation observed in the included studies. Previous research has demonstrated that after six weeks of whole body training [
Finally, the type of study design may also influence the conclusions about neuroplastic changes after ACL injury. Only one of the included studies followed a longitudinal design, measuring corticomotor excitability over three time points [
This scoping review is aimed at providing an in-depth analysis about the evidence of differences in brain activity associated with ACL injury. The discussion about confounding variables and their possible impact on the evidence found leads us to suggest the following recommendations for future studies in this area:
Participants’ demographic information needs to be clearly and completely presented, as well as details regarding the level and type of physical activity performed before the injury and at the time of the brain activity measurements Clinical information from participants with ACL injury detailing the injury mechanism, time since injury and surgery, and type of surgery should be thoroughly described Detailed information pertaining the rehabilitation protocol of participants with ACL injury should be reported, mainly regarding its duration, frequency, and the type of exercises most frequently performed; the rehabilitation performed by an ACL patient may help to shape the functional neuroplasticity process after the injury but is unfortunately very poorly described in the studies included in this scoping review
Regarding the rehabilitation after ACL injury, it would be relevant if future studies could determine the association between rehabilitation and brain activity in people with ACL injury. Some of the studies included in this review suggest the use of techniques such as TENS [
Another aspect that should be explored, specially by studies using TMS, concerns the corticomotor excitability involving the hamstring muscles. All studies that used TMS to measure motor thresholds and motor-evoked potentials selected the quadriceps (specifically the vastus medialis or lateralis) as the target muscle. Although it is widely accepted that the quadriceps muscle is crucial for proper knee function [
Finally, it is currently unknown if, and how, “baseline” brain activity differences are associated with an increased risk of sustaining a future ACL injury. This could be explored by including brain activity measurements in the routine assessment of different sports populations, which would allow to confirm a certain brain activity profile linked to an increased risk of future ACL injury (i.e., similar to the study undertaken by Diekfuss et al. [
This scoping review provides a summary of the current evidence associated to differences in brain activity after an ACL injury and/or reconstruction. Overall, results suggest evidence of functional neuroplasticity following ACL injury (with or without reconstruction), in both sensory (e.g., increased activity in secondary somatosensory area and lingual gyrus) and motor (e.g., lower corticomotor excitability) cortical areas. However, the heterogeneity in the measures of brain activity and corticospinal excitability seems to have an influence on the magnitude of the differences found (i.e., effect size). Moreover, many of the studies failed to control critical variables (e.g., time since injury), which may influence the observed effects. Therefore, it is recommended that future studies in this area should be aimed at minimizing the impact of confounding variables, thus increasing the level of evidence.
The following keywords, and different combinations between them, were used in the electronic search conducted in PubMed, PEDro, CINAHL, SPORTDiscus, and Cochrane Central Register of Controlled Trials databases:
Brain AND Neuroplasticity AND ACL injury
OR
(brain OR cortical) AND neuroplasticity AND ACL (injury OR rupture)
OR
(brain OR cortical) AND (neuroplasticity OR activity) AND ACL (rupture OR deficiency)
OR
(brain OR cortical) AND (neuroplasticity OR activation) AND ACL (rupture OR deficiency)
OR
(brain OR cortical) AND (neuroplasticity OR activation) AND ACL reconstruction.
OR
(brain OR cortical) AND (neuroplasticity OR activation) AND ACL (reconstruction OR repair)
OR
corticomotor AND (neuroplasticity OR excitability) AND ACL (injury OR reconstruction).
OR
(brain OR cortical) AND (neuroplasticity OR activity) AND knee ligament injury.
The authors declare that there is no conflict of interest regarding the publication of this article.