Sage Journals: Discover world-class research

Abstract

Background:

Knee proprioception is believed to be deficient after anterior cruciate ligament (ACL) injury. Tests of joint position sense (JPS) are commonly used to assess knee proprioception, but their psychometric properties (PMPs) are largely unknown.

Purpose:

To evaluate the PMPs (reliability, validity, and responsiveness) of existing knee JPS tests targeting individuals with ACL injury.

Study Design:

Systematic review; Level of evidence, 4.

Methods:

PubMed, Allied and Complementary Medicine, CINAHL, SPORTDiscus, Web of Science, Scopus, CENTRAL, and ProQuest databases were searched to identify studies that assessed PMPs of knee JPS tests in individuals with ACL injury. The risk of bias for each included study was assessed and rated at the outcome level for each knee JPS test. Overall quality and levels of evidence for each PMP were rated according to established criteria. Meta-analyses with mean differences were conducted using random effects models when adequate data were available.

Results:

Included were 80 studies covering 119 versions of knee JPS tests. Meta-analyses indicated sufficient quality for known-groups and discriminative validity (ACL-injured knees vs knees of asymptomatic controls and contralateral noninjured knees, respectively), owing to significantly greater absolute errors for ACL-injured knees based on a strong level of evidence. A meta-analysis showed insufficient quality for responsiveness, which was attributed to a lack of significant change over time after diverse interventions with a moderate level of evidence. Statistical heterogeneity (I ² > 40%) was evident in the majority of meta-analyses. All remaining PMPs (reliability, measurement error, criterion validity, convergent validity, and other PMPs related to responsiveness) were assessed qualitatively, and they failed to achieve a sufficient quality rating. This was a result of either the study outcomes not agreeing with the statistical cutoff values/hypotheses or the level of evidence being rated as conflicting/unknown or based on only a single study.

Conclusion:

Knee JPS tests appear to have sufficient validity in differentiating ACL-injured knees from asymptomatic knees. Further evidence of high methodologic quality is required to ascertain the reliability, responsiveness, and other types of validity assessed here. We recommend investigations that compare the modifiable methodologic components of knee JPS tests on their PMPs to develop standardized evidence-based tests.

Keywords

proprioception reliability validity

Anterior cruciate ligament (ACL) rupture is a common injury, particularly in sports, with an annual incidence of 68.6 isolated tears per 100,000 person-years in the United States.⁹³ Treatment involves extensive rehabilitation with or without reconstructive surgery. Despite treatment, approximately 1 in 4 young athletes who return to high-risk sports, such as football and skiing, will experience a secondary ACL injury.¹⁰⁷ It is widely believed that ACL injury compromises neuromuscular control and proprioceptive acuity of the knee and may contribute to the increased risk for secondary injury.^48,72

Proprioceptive receptors in muscles, skin, and joints convey afferent information to the brain regarding sensations including tension, effort, force, movement, and position.^82,91 This information is subsequently interpreted by the central nervous system, which in turn conveys efferent signals to guide behavioral motor responses and avoid injury.^13,75 The presence of mechanoreceptors in the ACL has been well documented in histologic studies,^{20,40,41,95,112} and the damage or loss of these receptors may negatively influence knee proprioception.¹ Proprioceptive acuity tests of the knee are therefore prevalent in the literature.⁴⁹

Assessing knee proprioception involves testing 1 specific submodality of proprioception, most often either joint position sense (JPS) or threshold to detect passive motion.⁴⁹ Differentiating submodalities of proprioception is important given the lack of correlation among tests that attempt to quantify them.⁶⁴ Meta-analyses have found significantly poorer knee proprioception among ACL-injured knees as compared with the contralateral noninjured knees of the same individuals^58,87 and with asymptomatic control groups,⁸⁷ with knee JPS tests showing greater differences than movement discrimination tests showed. However, the absolute differences in degrees of knee flexion were <1°, and differences versus asymptomatic controls are not always evident.⁷³ A major reason for the inconsistent results is likely the lack of standardized test procedures.

The psychometric properties (PMPs; eg, reliability, validity, and responsiveness) of proprioception assessment methods have not been established across joints or tests, and 2 systematic reviews have emphasized the need for further research in this area.^46,49 Specifically, common tests of knee proprioception lack sufficient evidence to draw reliable and valid conclusions based on their outcome measures.⁸⁷ A systematic review by Smith et al¹⁰⁰ found varying evidence for the reliability of knee JPS tests across pathologic and asymptomatic populations; the authors recommended additional examination of such tests among those with ACL injury. This conflicting evidence may be in part due to the many modifiable components of knee JPS tests, such as joint angles, reproduction procedure, and body position. A plethora of knee JPS tests therefore exists, and this systematic review of the updated literature assimilates the current evidence pertaining to their PMPs. It is intended to provide guidance for those seeking to evaluate knee JPS, as well as to inform future researchers aiming to develop improved methods of testing.

This article forms part of a systematic review series investigating the PMPs of all knee proprioception tests targeting individuals with ACL injury and asymptomatic controls.⁸ The aim of the current systematic review is to corroborate the evidence for the PMPs of all knee JPS tests that have been applied in the scientific literature targeting those with ACL injury. In line with previous meta-analyses, we hypothesized that knee JPS tests would show sufficient validity regarding differentiation between ACL-injured and asymptomatic knees. We also hypothesized that a lack of evidence would make it difficult to sufficiently ascertain the quality of all remaining PMPs of knee JPS tests among individuals with ACL injury.

Methods

We adhered to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines for conducting this review,^69,97 as stated in our protocol,⁸ which was registered in the International Prospective Register of Systematic Reviews (PROSPERO; registration CRD42018108014).

Eligibility Criteria

Eligibility criteria for inclusion and exclusion of studies are detailed in Table 1. Briefly, studies were to include individuals ≥10 years of age with ACL injury managed with or without using surgical reconstruction who performed at least 1 test of knee JPS. Studies must have investigated at least 1 PMP (reliability, validity, or responsiveness) relating to the outcome measures of the JPS tests.

Table 1

Eligibility Criteria for Studies

Category	Details
Inclusion criteria
Participants	Age ≥10 y with anterior cruciate ligament injury managed with or without using surgical reconstruction
Construct	At least 1 specific method of measuring knee joint position sense
Equipment	Any equipment that is capable of quantifying knee joint position sense error
Setting	The test can be performed in any setting, including a laboratory or a clinic
Outcome measures	Studies designed to investigate at least 1 of the following psychometric properties: reliability, measurement error, criterion validity (concurrent or predictive), hypothesis testing (convergent, known-groups, or discriminative validity), and responsiveness
Study type	(1) The primary or sole aim of investigating at least 1 psychometric property of a knee joint position sense test; (2) reliability, validity, or responsiveness reported as secondary or additional findings on the condition that sufficient details are included to rate the methodologic quality/risk of bias; (3) studies that have included data separately for individuals with anterior cruciate ligament injury, other lower limb disorders, and knee-healthy controls; (4) peer-reviewed observational studies, cross-sectional studies, randomized controlled clinical trials, or quasi-experimental studies
Language	English language only
Access	Full-text publications retrievable via electronic database or manual search
Exclusion criteria
Construct	(1) Validation of self-reported knee function and/or physical activity levels without addressing specific knee joint position sense tests and/or (2) validation of proprioception-related function, such as threshold to detect passive motion, not specifically assessing knee joint position sense
Equipment	Validation of measurement instruments not designed to assess knee joint position sense
Outcome measures	Measures not addressing any psychometric properties of a knee joint position sense test
Study type	Pilot studies, abstracts, systematic reviews and meta-analyses, narrative reviews, book reviews, case series/reports, commentaries, editorials, letters to the editor, patient education handouts, consensus statements, clinical practice guidelines, theses/dissertations, or unpublished literature

Search Strategy and Study Selection

A systematic search was performed by the first author (A.S.) using the following databases between their inception and March 3, 2020: PubMed, Allied and Complementary Medicine (via EBSCO), CINAHL (via EBSCO), SPORTDiscus (via EBSCO), Web of Science, Scopus, the Cochrane Central Register of Controlled Trials (CENTRAL), and Physical Education Index (via ProQuest). Search terms and strategies are elaborated in the published review protocol.⁸ In addition, searches were performed via Google Scholar using relevant search terms.

All identified articles were imported to EndNote X9.3.1 (Clarivate Analytics Software) and then exported to Covidence systematic review software (Veritas Health Innovation) where duplicates were removed. Two authors (A.S., E.T.) independently screened all titles and abstracts for suitability according to the predetermined eligibility criteria. The full texts of the potentially eligible articles were retrieved and screened independently by the same 2 authors. Disagreements were resolved either by consensus or by consulting a third author (A.A.). Reference lists of the included articles were screened for any potential additional studies not found in the systematic search.

Data Extraction and Risk-of-Bias Assessment

Data were extracted from all eligible studies that reported PMPs of knee JPS tests in individuals with ACL injury as primary or secondary (additional) findings. Extracted study information included the following: authors, year of publication, study design, population characteristics (number of participants, age, sex distribution, height, weight, treatment type, time since injury/surgery, activity level), testing procedures, statistical analyses, results relating to knee JPS PMP outcome measures, and any other data required for risk-of-bias (RoB) assessment. If data were missing or inadequate, an email was sent to the authors requesting the required information.

The RoB was assessed and rated using the Consensus-Based Standards for the Selection of Health Measurement Instruments (COSMIN) RoB checklist.⁷⁰ Thus, the methodologic quality of each study was rated according to a 4-point rating scale: inadequate (high RoB), doubtful, adequate, and very good (low RoB). Studies rated inadequate were not included in evidence syntheses. The following PMPs were then examined at the outcome level:

Reliability

Measurement error

Criterion validity: concurrent or predictive

Hypothesis testing for construct validity:

Convergent versus other outcome measures

Discriminative, which was operationally defined as ACL-injured knees versus the contralateral asymptomatic knees of the same individuals

Known groups, which was operationally defined as ACL-injured knees versus either asymptomatic knees of other individuals or ACL-injured knees of otherwise asymptomatic individuals

Responsiveness: criterion and construct approach

Data extraction and RoB evaluation were piloted using 2 studies chosen to include what was assumed to be the most common general PMPs: hypothesis testing⁵³ and responsiveness.¹⁴ Piloting was achieved by all authors independently performing data extraction and rating the studies for RoB before meeting to discuss any concerns and to arrive at a consensus. Data from all articles were then extracted independently by an assigned author (A.S., 38%; E.T., 17%; A.A., 15%; U.R., 15%; C.K.H., 14%) and verified by a second (A.S., 25%; E.T., 31%; A.A., 28%; C.K.H., 15%), ensuring that all pairs of authors worked together on at least 2 articles. A third author was assigned to each article in case of any disagreements, but all were resolved through consensus.

Within studies, JPS tests were considered separate if they differed regarding modifiable components of reproduction method, direction of movement, body position, measurement equipment, and test procedure but not if they differed for target angle (TA) because some studies used data from multiple TAs for the same outcome. In instances where an ACL group underwent reconstruction during the study, it was classified as having ACL reconstruction (ACLR) as opposed to being ACL deficient (ACLD). Also, when participants were followed up over time and known-groups validity or discriminative validity was examined, follow-up data for only the last test were extracted. This was done to provide more stable and homogeneous results. In studies where knee JPS was assessed during the same testing session with and without an intervention, such as Kinesio tape¹⁵ or vibration,⁷¹ data for known-groups and discriminative validity were extracted just for baseline. Knee JPS can be analyzed using absolute error (AE), constant error, or variable error. For this review, only data for AE were extracted because constant and variable errors were seldom reported.

Quality Assessment of PMPs

The quality of each PMP was evaluated according to our protocol⁸ using a criterion list adapted from Prinsen et al⁸¹ (see Appendix Table A1, available online as Supplemental Material). Briefly, ratings of sufficient, indeterminate, or insufficient depended on whether the synthesized study outcomes were in agreement with the threshold cutoff values of relevant statistical parameters or hypotheses. Regarding hypothesis testing for construct validity and responsiveness, ratings of sufficient and insufficient required 75% of the findings to be in favor of the hypothesis (positive) or not (negative), respectively. If <75% of the findings were in agreement, the quality of the PMP was rated as inconsistent owing to the conflicting level of evidence. In instances when a PMP was evaluated only in studies with an RoB rating of inadequate or was not investigated at all, the quality of that PMP was denoted “not estimable” because of the unknown level of evidence.

Level of Evidence

The level of evidence for each PMP was ascertained according to a scale adapted by Kroman et al⁶⁰ (Appendix Table A2, available online). Rating criteria were as follows:

Strong: the test was evaluated in multiple studies of adequate RoB rating or at least 1 study of very good RoB rating (implying a low RoB).

Moderate: the test was evaluated in multiple studies of doubtful RoB rating or at least 1 study of adequate RoB rating.

Limited: the test was evaluated in 1 study of doubtful RoB rating.

Conflicting: the test was evaluated in studies with contradictory findings.

Unknown: the test was evaluated in studies of inadequate RoB rating or not investigated at all.

Meta-analysis

Meta-analyses for each PMP were performed using the Review Manager computer program (RevMan Version 5.3; Nordic Cochrane Centre, Cochrane Collaboration) when adequate (homogeneous) data were available from at least 3 studies with a doubtful, adequate, or very good RoB rating. A random effects model was applied to account for differences in the way that the studies were conducted in addition to the sampling error.¹⁸ The mean difference (MD) was used rather than the standardized MD, as only outcome measures in the same unit (AE) were included and for greater statistical power.¹⁰³ When the means and standard deviations (SDs) of the AE were not reported but median values with measures of dispersion were, means and SDs were estimated per the method of Wan et al.¹⁰⁵ Statistical heterogeneity was considered present if the I ² statistic exceeded 40%.²⁸ Funnel plots with trim-and-fill analyses²⁹ to help identify and correct for the presence of publication bias were considered appropriate only when the following criteria were met, as described by Ioannidis and Trikalinos⁵¹: the inclusion of at least 10 studies (with statistically significant results in at least 1 study), statistical heterogeneity <50%, and a ratio >4 for the maximal to minimal variance across studies. Subgroups were created within each meta-analysis according to each JPS test procedure, such as passive-active (passive movement to the TA and active movement to the reproduction angle), to investigate potential differences in outcome related to this factor. Sensitivity analyses were performed for study population and methodologic quality when appropriate by restricting the meta-analyses to ACLD or ACLR and studies with an adequate or very good ROB rating, respectively. Statistical significance was set at P < .05.

Control group data were obtained for the nondominant legs, when available, to provide a more stringent comparison with the injured legs of ACL groups to analyze known-groups validity. When information of side was provided without indicating dominance, the left leg was chosen to increase the likelihood of selecting the nondominant leg. For responsiveness, when knee JPS was tested over time, just the data comparing baseline and the last time point were used for analysis. In addition, when an intervention included a control condition and a placebo condition, such as sham Kinesio tape¹⁵ or placebo knee brace,¹⁰⁸ data were limited to comparisons between the control and intervention conditions to evaluate responsiveness.

Results

Search Results

The electronic database search yielded 2978 studies, and manual searches yielded an additional 63. After removal of duplicates, the titles and abstracts of 2527 articles were screened, of which 462 full-text articles were then screened for eligibility. Subsequently, 80 articles were included in this review (Figure 1).

Figure 1.

PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram illustrating inclusion of studies. ACL, anterior cruciate ligament.

Characteristics of the Studies

Study characteristics are provided in Appendix Table A3 (available online). To summarize, in the 80 studies, we identified 2475 participants with a unilateral ACL injury from 112 cohorts. Of these, 81 cohorts were classified as having ACLR (17 of which were ACLD at inclusion but underwent ACLR during the study), and 31 were classified as being ACLD. Sex distribution was indicated for 99 (88%) ACL populations, revealing that 30% were female. The mean weighted age of the ACL populations was 27.91 years, as based on the 104 ACL populations for which the mean age was reported. Three studies^1,43,66 cited median age only, and 4 studies^2,10,30,54 did not report age.

A total of 119 knee JPS tests were identified. The most common procedure was passive-active (54 tests), followed by active-active (25 tests), passive-passive (19 tests), passive–visual estimation (13 tests), and active-passive (1 test). The procedure was not reported or was unclear for 7 tests. In total, 74 tests were performed with participants seated; 16, side lying; 14, supine lying; 8, standing; 3, reclined; and 1, prone lying; and for 3 tests, this information was not indicated. The most common equipment used to measure the knee angle was a dynamometer (48 tests). The most common starting angle was 90° of knee flexion (47 tests), followed by 0° (23 tests), 60° (18 tests), and 30° (10 tests). Three tests used a specified range, 6 tests used >1 starting angle, and starting angle was not reported for 14 tests. The most common TA was 30° (38 tests), followed by 15° (24 tests), 60° (23 tests), and 45° (21 tests). More than 1 TA was used in 39 tests, unspecified TAs within a range or multiple ranges were used in 25 tests, and TA was not cited for 15 tests. The reproduction method was the ipsilateral leg in 101 tests, with a form of visual estimation in 13 tests and with the contralateral leg in 4 tests. The method was not reported for 1 test. Movement direction (toward the TA and reproduction angle) was extension for 61 tests, flexion for 33 tests, and a mixture of flexion and extension for 12 tests (eg, flexion toward TA but extension toward reproduction angle). Direction was not noted for 13 tests.

Assessment of RoB was performed at the outcome level (Appendix Table A4, available online). However, the RoB standards (ie, the questions that are scored) differ among PMPs to better reflect the specific study designs required; thus, some overall ratings differed for the same studies among PMPs. For example, 1 study¹⁰⁴ received an RoB rating of doubtful for known-groups validity but a lower rating of inadequate for convergent validity in part because the comparator outcome measures were not performed on the same day as were the knee JPS tests. One standard that was included for rating all PMPs was related to other important flaws in the design or statistical methods of the study. Considerations for this standard included but were not limited to the number of knee JPS test trials performed (a minimum of 5 trials was considered appropriate), sample size, description of the participant characteristics, and methodologic procedures. Thus, because the COSMIN checklist operates on a worst-score-counts principle, many of the studies received their lowest score for this standard; subsequently, the overall RoB rating was often equal to the score for this standard.

Assessment of Quality and Level of Evidence of the Studies

Reliability and Measurement Error

Test-retest reliability of knee JPS tests was analyzed among participants with ACLR in 3 studies,^6,62,68 in which measurement error was reported in 2 studies^6,68 (Table 2). However, all studies had an inadequate RoB rating (Appendix Table A4, available online). Thus, the quality of test-retest reliability and related measurement error was not estimable because of the unknown level of evidence.

Table 2

Reliability and Measurement Error ^a

	Population		JPS Test Details			Test-Retest Reliability			Measurement Error
Study (Year)	ACLD / ACLR	No.	Position	Procedure	TA, deg	ICC (95% CI)	PMP Quality ^b	COSMIN RoB	SEM (SDD), deg	PMP Quality ^b	COSMIN RoB
Angoules (2011)⁶	ACLR	40	Seated	P-A			NA	Inadequate		NA	Inadequate
					15	0.96 (0.86-0.99)			0.24 (0.40)
					45	0.99 (0.95-1.00)			0.20 (0.32)
					75	0.95 (0.81-0.99)			0.35 (0.57)
Lee (2015)⁶²	ACLR	76	Seated	P-P	45	0.78 (NR)	NA	Inadequate	NA	NA	NA
Moezy (2008)⁶⁸	ACLR	NR	Seated	A-A			NA	Inadequate		NA	Inadequate
					30	0.99 (0.95-1.00)			0.76
					60	0.93 (0.77-0.98)			2.09
• Pooled results for PMP quality: 6 NA
• Level of evidence: unknown
Blackburn (2020)¹⁶	ACLR	72	Reclined	A-A	20/25/30	0.06 (NR) ^c	– ^c	Very good ^c	2.7	?	Very good
• Pooled results for PMP quality: insufficient (intrasession reliability), indeterminate (measurement error)
• Level of evidence: strong (intrasession reliability), strong (measurement error)

^a A, active; ACLD, anterior cruciate ligament deficient; ACLR, anterior cruciate ligament reconstruction; COSMIN, Consensus-Based Standards for the Selection of Health Measurement Instruments; ICC, intraclass correlation coefficient; JPS, joint position sense; NR, not reported; P, passive; PMP, psychometric property; RoB, risk of bias; SDD, smallest detectable difference; TA, target angle.

^b Ratings: +, sufficient; ?, indeterminate; –, insufficient; NA, not applicable (owing to rating of inadequate risk of bias).

^c For intrasession reliability.

Intrasession reliability and associated measurement error were evaluated for 1 active knee JPS test in 1 study,¹⁶ which had an RoB rating of very good (Appendix Table A4, available online). Thus, the intraclass correlation coefficient (2,1) of 0.06 (95% CI not reported) and SEM of 2.7° (without reference to minimal important change) (Table 2) indicated insufficient and indeterminate quality for intrasession reliability and associated measurement error, respectively, based on a strong level of evidence.

Criterion Validity

No gold standard knee JPS test exists; therefore, no studies that performed such a comparison were identified. In addition, no studies that investigated the predictive validity of knee JPS tests were identified. Concurrent validity (ie, comparison among knee JPS tests) was analyzed in 1 study,³⁰ which had an RoB rating of inadequate (Appendix Table A4, available online). Thus, the quality of criterion (concurrent) validity for knee JPS tests among individuals with ACL injury was not estimable because of an unknown level of evidence.

Convergent Validity

Convergent validity (ie, correlation between knee JPS and other outcome measures) was analyzed in 16 studies^∥ (Appendix Table A5, available online). In case of multiple correlations in the same study, all were given the same RoB rating: very good for 1 study,⁴⁷ doubtful for 9 studies,^¶ and inadequate for 6 studies^{10,21,26,63,77,104} (Appendix Table A4, available online). The most commonly examined outcome measures were quadriceps strength,^{21,62,74,104,111} hamstring strength,^21,62,74,104 hamstring to quadriceps ratio,^26,62,74,111 and knee laxity.^1,10,32,35 A meta-analysis was precluded because set criteria were not met. As 91.2% of the findings did not show correlations between the relevant knee JPS tests and the other outcome measures, we found insufficient quality for convergent validity of knee JPS tests. The level of evidence was determined to be moderate because of the multiple diverse outcome measures, and despite 1 study having an RoB rating of very good, the remaining 9 studies had an RoB rating of doubtful.

Hypothesis Testing

Known-Groups Validity

Known-groups validity was evaluated in 181 group comparisons across 52 studies^＃ (Appendix Table A6, available online). All multiple within-study comparisons received the same overall RoB rating: very good for 3,^34,44,86 adequate for 4,^47,79,85,88 doubtful for 32,^** and inadequate for 13 studies^†† (Appendix Table A4, available online). Meta-analyses including, when appropriate, subgroup comparisons and sensitivity analyses were thus performed as detailed in the following sections. The criteria for funnel plots with trim-and-fill analyses were not met by the studies.

ACL-Injured Knees vs Asymptomatic Controls

A meta-analysis included 59 comparisons between ACL groups and asymptomatic controls across 18 studies^‡‡ (Figure 2). ACL-injured knees showed significantly greater JPS error than did those of asymptomatic controls (MD, 0.70°; 95% CI, 0.40°-1.01°; P < .00001; I ² = 91%). A sensitivity analysis for study quality was performed by including only studies with an adequate or very good RoB rating (Figure 3) (9 comparisons across 4 such studies^34,79,85,86). ACL-injured knees produced significantly greater JPS error than did asymptomatic controls (MD, 2.97°; 95% CI, 1.57°-4.37°; P = .00001; I ² = 97%).

Figure 2.

Meta-analysis comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament (ACL)–injured knees and the knees of healthy controls (known-groups validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. ACLD, anterior cruciate ligament deficient; ACLR, anterior cruciate ligament reconstruction; BPTP, bone–patellar tendon–bone; CTRL, control; ext, extension; flex, flexion; HT, hamstring tendon; IV, inverse variance; TA, target angle.

Figure 3.

Meta-analysis including only studies with a risk of bias rating of adequate or very good comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament (ACL)–injured knees and the knees of healthy controls (known-groups validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. ACLD, anterior cruciate ligament deficient; ACLR, anterior cruciate ligament reconstruction; CTRL, control; ext, extension; flex, flexion; IV, inverse variance; TA, target angle.

Subgroups were created in the meta-analysis based on the type of JPS test: passive-active (29 comparisons across 11 studies^§§), passive-passive (10 comparisons across 5 studies^{7,34,65,79,111}), active-active (14 comparisons across 4 studies^19,67,71,92), and passive–visual estimation (6 comparisons across 2 studies^89,90). Because of an insufficient number of studies, 1 meta-analysis subgroup (passive–visual estimation) was precluded, and only the other 3 subgroup meta-analyses were considered relevant (Figure 2). The JPS error was significantly greater for ACL-injured knees as compared with asymptomatic controls for passive-passive procedures (MD, 1.59°; 95% CI, 0.96° to 2.22°; P < .00001; I ² = 95%) and was close to being significantly greater for ACL-injured knees for passive-active procedures (MD, 0.61°; 95% CI, –0.03° to 1.25°; P = .06; I ² = 87%) and active-active procedures (MD, 0.26°; 95% CI, 0.00° to 0.51°; P = .05; I ² = 60%). Sensitivity analyses comparing ACLD and ACLR separately against asymptomatic controls are detailed hereafter.

ACLD Knees vs Asymptomatic Controls

Twenty-four comparisons between ACLD knees and asymptomatic controls were identified across 8 studies.^{7,32,34,79,80,86,90,114} ACLD knees produced significantly greater JPS errors than did asymptomatic controls (MD, 0.94°; 95% CI, 0.48°-1.39°; P = .0001; I ² = 89%) (Figure 4). A sensitivity analysis limited to studies with only adequate or very good RoB ratings^34,79,86 revealed similar findings (MD, 3.56°; 95% CI, 2.84°-4.27°; P < .00001; I ² = 50%) (Figure 5).

Figure 4.

Meta-analysis comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament deficient (ACLD) knees (treated without using surgical reconstruction) and the knees of healthy controls (known-groups validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. CTRL, control; ext, extension; flex, flexion; IV, inverse variance; TA, target angle.

Figure 5.

Meta-analysis including only studies with a risk of bias rating of adequate or very good comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament deficient (ACLD) knees (treated without using surgical reconstruction) and the knees of healthy controls (known-groups validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. CTRL, control; IV, inverse variance; TA, target angle.

ACLR Knees vs Asymptomatic Controls

After 35 comparisons between ACLR knees and asymptomatic controls across 13 studies,^∥∥ ACLR knees showed significantly greater JPS error as compared with asymptomatic controls (MD, 0.48°; 95% CI, 0.18°-0.78°; P = .002; I ² = 80%) (Figure 6). A sensitivity analysis was precluded owing to an insufficient number of studies with RoB ratings of adequate or very good.

Figure 6.

Meta-analysis comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament reconstruction (ACLR) knees and the knees of healthy controls. A random effects model with mean difference was used. Subgroups are shown per test procedure. BPTP, bone–patellar tendon–bone; CTRL, control; ext, extension; flex, flexion; HT, hamstring tendon; IV, inverse variance; TA, target angle.

ACLD vs ACLR Knees

Based on 10 comparisons across 3 studies,^32,59,80 a meta-analysis found significantly greater JPS error for ACLD knees than ACLR knees (MD, 0.80°; 95% CI, 0.29°-1.32°; P < .002; I ² = 34%) (Figure 7). A sensitivity analysis was precluded owing to a lack of studies meeting the defined criteria.

Figure 7.

Meta-analysis comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament deficient (ACLD) knees (treated without using surgical reconstruction) and anterior cruciate ligament reconstruction (ACLR) knees (known-groups validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. ext, extension; flex, flexion; IV, inverse variance; Std, standardized; TA, target angle.

Overall Quality of Known-Groups Validity and Level of Evidence

To summarize, knee JPS tests elicited significantly greater errors for ACL-injured knees as compared with healthy controls and for ACLD knees as compared with ACLR knees. Thus, the known-groups validity of knee JPS tests targeting individuals with ACL injury was sufficient. The sensitivity analysis supported these findings; thus, the level of evidence was rated as strong.

Discriminative Validity

Discriminative validity was analyzed in 146 comparisons across 44 studies^¶¶ (Appendix Table A7, available online). All relevant within-study comparisons received the same RoB rating, which was very good for 3 studies,^33,34,86 adequate for 3 studies,^47,79,85 doubtful for 26 studies,^## and inadequate for 12 studies^a (Appendix Table A4, available online). Meta-analyses were performed with subgroup comparisons and sensitivity analyses when appropriate, as described hereafter. Funnel plots and trim-and-fill analyses were precluded owing to the aforementioned criteria not being met.

ACL-Injured vs Contralateral Noninjured Knees

A meta-analysis comparing knee JPS AE between ACL-injured knees (ACLD or ACLR) and the contralateral noninjured knees of the same persons included 57 comparisons across 22 studies.^b ACL-injured knees showed significantly greater knee JPS error than did the contralateral knees (MD, 0.69°; 95% CI, 0.31° to 1.06°; P = .0003; I ² = 87%) (Figure 8). A sensitivity analysis of 16 comparisons across 6 studies was performed.^{33,34,47,79,85,86} Significantly greater JPS error was again seen for ACL-injured knees as compared with the contralateral knees of the same individuals (MD, 1.55°; 95% CI, 0.57° to 2.53°; P = .002; I ² = 96%) (Figure 9). In addition, within the main analysis, subgroups based on test procedure were created for the following: passive-active (29 comparisons across 14 studies^c), passive-passive (12 comparisons across 6 studies^{31,34,53,62,74,79}), active-active (16 comparisons across 8 studies^{16,27,33,43,50,53,67,71}), passive–visual estimation (2 comparisons in 1 study,⁸⁹ thus precluding meta-analysis), and not reported (1 comparison in 1 study¹⁵). Significantly greater errors were seen for the ACL-injured knees than the contralateral knees for test procedures that were passive-active (MD, 1.10°; 95% CI, 0.36° to 1.83°; P = .003; I ² = 92%) and passive-passive (MD, 0.78°; 95% CI, 0.27° to 1.29°; P = .003; I ² = 72%) but not active-active (MD, 0.11°; 95% CI, –0.33° to 0.55°; P = .63; I ² = 35%).

Figure 8.

Meta-analysis comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament (ACL)–injured knees (treated with or without using surgical reconstruction) and the contralateral healthy knees of the same individuals (discriminative validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. ACLD, anterior cruciate ligament deficient; ACLR, anterior cruciate ligament reconstruction; ext, extension; flex, flexion; IV, inverse variance; MJA, maximum joint amplitude; MVC, maximal voluntary contraction; P-F, poor-functioning; TA, target angle; W-F, well-functioning.

Figure 9.

Meta-analysis including only studies with a risk of bias rating of adequate or very good comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament (ACL)–injured knees (treated with or without using surgical reconstruction) and the contralateral healthy knees of the same individuals (discriminative validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. ACLD, anterior cruciate ligament deficient; ACLR, anterior cruciate ligament reconstruction; ext, extension; flex, flexion; IV, inverse variance; TA, target angle.

ACLD vs Contralateral Noninjured Knees

For 24 comparisons between the knees of individuals with unilateral ACLD across 11 studies,^d ACLD knees showed significantly greater knee JPS error (MD, 0.95°; 95% CI, 0.44° to 1.45°; P = .0003; I ² = 77%) (Figure 10). A sensitivity analysis (7 comparisons across 4 studies^33,34,79,86) also showed significantly greater JPS errors for ACLD knees than for the contralateral knees of the same individuals (MD, 1.81°; 95% CI, 0.09° to 3.54°; P = .04; I ² = 90%) (Figure 11). To elucidate the potential effects of test procedure on outcomes, the following subgroups of tests were created: passive-active (10 comparisons across 8 studies^{32
–34,43,53,56,80,86}), passive-passive (5 comparisons across 3 studies^34,53,79), active-active (8 comparisons across 4 studies^27,33,43,53), and not reported (1 comparison in 1 study¹⁵). Greater JPS errors were evident for ACLD knees than for the contralateral knees for procedures that were passive-active (MD, 1.06°; 95% CI, 0.23° to 1.89°; P = .01; I ² = 82%) and passive-passive (MD, 2.25°; 95% CI, 1.39° to 3.12°; P < .00001; I ² = 17%) but not active-active (MD, 0.49°; 95% CI, –0.27° to 1.24°; P = .21; I ² = 59%).

Figure 10.

Meta-analysis comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament deficient (ACLD) knees (treated without using surgical reconstruction) and the contralateral healthy knees of the same individuals (discriminative validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. ext, extension; flex, flexion; IV, inverse variance; MJA, maximum joint amplitude; MVC, maximal voluntary contraction; P-F, poor-functioning; TA, target angle; W-F, well-functioning.

Figure 11.

Meta-analysis including only studies with a risk of bias rating of adequate or very good comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament deficient (ACLD) knees (treated without using surgical reconstruction) and the contralateral healthy knees of the same individuals (discriminative validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. IV, inverse variance; TA, target angle.

ACLR vs Contralateral Noninjured Knees

For 36 comparisons between the knees of individuals with unilateral ACLR across 15 studies,^e no statistically significant difference in knee JPS errors were found (MD, 0.47°; 95% CI, –0.03° to 0.96°; P = .06; I ² = 89%) (Figure 12). A sensitivity analysis (9 comparisons across 3 studies^47,79,85) conversely revealed significantly greater knee JPS error for ACLR knees than for the healthy contralateral knees (MD, 1.36°; 95% CI, 0.08° to 2.65°; P = .04; I ² = 97%) (Figure 13). The following subgroups of tests were also analyzed: passive-active (19 comparisons across 8 studies^{31,32,47,55,80,85,89,99}), passive-passive (7 comparisons across 4 studies^31,62,74,79), active-active (8 comparisons across 4 studies^16,50,67,71), and passive–visual estimation (separate meta-analysis precluded owing to 2 comparisons in only 1 study⁸⁹). JPS error was significantly greater for the ACLR knee as compared with the contralateral knee for passive-passive procedures (MD, 0.22°; 95% CI, 0.05° to 0.40°; P = .01; I ² = 0%) and close to significant for passive-active procedures (MD, 0.95°; 95% CI, 0.01° to 1.89°; P = .05; I ² = 93%); yet, no significant differences were seen for active-active (MD, –0.34°; 95% CI, –0.82° to 0.13°; P = .15; I ² = 0%).

Figure 12.

Meta-analysis comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament reconstruction (ACLR) knees and the contralateral healthy knees of the same individuals (discriminative validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. ext, extension; flex, flexion; IV, inverse variance; TA, target angle.

Figure 13.

Meta-analysis including only studies with a risk of bias rating of adequate or very good comparing knee joint position sense absolute error group means in degrees of flexion between anterior cruciate ligament reconstruction (ACLR) knees and the contralateral healthy knees of the same individuals (discriminative validity). A random effects model with mean difference was used. Subgroups are shown per test procedure. ext, extension; flex, flexion; IV, inverse variance; TA, target angle.

Overall Quality of Discriminative Validity and Level of Evidence

To summarize, knee JPS tests generally elicited significantly greater errors for ACL-injured knees as compared with the healthy contralateral knees of the same individuals. Thus, the discriminative validity of knee JPS tests targeting individuals with ACL injury was rated as sufficient. The sensitivity analysis found even greater differences in agreement with the hypothesis; therefore, the level of evidence was rated as strong.

Responsiveness

Criterion Approach

Responsiveness regarding comparisons with a gold standard or other knee JPS test over time was not examined in any study.

Construct Approach Among Other Outcome Measurements

One study analyzed correlations between a passive-active knee JPS test and somatosensory evoked potentials for 2 ACLR groups over time,⁷⁷ but this received an RoB rating of inadequate (Appendix Table A4, available online). Thus, the quality of responsiveness among other outcome measurements for knee JPS tests among individuals with ACL injury was not estimable owing to an unknown level of evidence.

Construct Approach Among Subgroups

Responsiveness among subgroups was evaluated in 18 comparisons across 10 studies^f (Appendix Table A8, available online). Eight of these studies compared change in knee JPS over time among ACL-injured groups that received different interventions^{6,9,23,38,68,98,106,113}; 1 compared change between male and female patients with ACLR³; and 1 compared change in knee JPS after one 36-minute exercise protocol between an ACLR group and a control group.⁴² All multiple comparisons within studies received the same RoB rating: adequate for 1 study,³⁸ doubtful for 3 studies,^6,9,42 and inadequate for 6 studies^{3,23,68,98,106,113} (Appendix Table A4, available online). A meta-analysis was precluded because data were missing from the relevant studies. Qualitatively, none of the findings from the studies rated as having a doubtful or adequate RoB agreed with the hypothesis that the experimental ACL group would show significantly greater improvements in knee JPS as compared with the subgroups. The quality of the PMP responsiveness among subgroups for knee JPS tests among individuals with ACL injury was therefore estimated as insufficient, and the level of evidence was rated moderate.

Construct Approach Before and After Intervention

Responsiveness of knee JPS tests to the effects of interventions for ACL-injured knees was assessed for 64 before-versus-after intervention comparisons across 22 studies^g (Appendix Table A9, available online). All multiple within-study comparisons achieved the same RoB ratings: very good for 1,⁷⁶ adequate for 1,⁷⁹ doubtful for 11,^h and inadequate for 9 studiesⁱ (Appendix Table A4, available online). A meta-analysis (12 comparisons across 6 studies^{15,38,71,79,108,113}) found no statistically significant differences in JPS error for ACL-injured knees after interventions (Figure 14) (MD, 0.67°; 95% CI, –0.46 to 1.81; P = .25; I ² = 89%). A sensitivity analysis and a funnel plot with trim-and-fill analysis were precluded because the previously stated criteria were not met. Thus, the quality of responsiveness to intervention for knee JPS tests was rated insufficient. The level of evidence for these findings was rated moderate owing to the studies in the meta-analysis consisting of only 1 study with an RoB rating of adequate and the remaining 11 having an RoB rating of doubtful.

Figure 14.

Meta-analysis comparing knee joint position sense absolute error group means in degrees of flexion between baseline (preintervention) and postintervention assessments (responsiveness to intervention) among anterior cruciate ligament–injured knees (treated with or without using reconstruction). A random effects model with mean difference was used. Subgroups are shown per test procedure. IV, inverse variance; TA, target angle.

Discussion

The purpose of this systematic review was to substantiate the evidence for the PMPs of knee JPS tests targeting individuals with ACL injury. The main findings based on meta-analyses were that ACL-injured knees produced significantly greater AEs for knee JPS tests as compared with the knees of asymptomatic controls (known-groups validity) as well as the contralateral healthy knees of the same individuals (discriminative validity). Thus, the quality of these properties was rated as sufficient, and the level of evidence for both was rated as strong. However, a meta-analysis for responsiveness to interventions did not show statistically significant results. As a result, the quality of this property was rated as insufficient, and the level of evidence was rated moderate, as was responsiveness in comparison with subgroups based on a qualitative synthesis. Caution should nonetheless be taken with these findings, as the presence of statistical heterogeneity in the majority of meta-analyses and publication bias may have affected the results. Reliability and associated measurement error of an active-active JPS test procedure for an intrasession design were the only properties to show a strong level of evidence for the qualitative syntheses, but this was based on just 1 study, the findings of which indicated insufficient and indeterminate quality for each property, respectively.¹⁶ The quality of the remaining PMPs—reliability and measurement error for test-retest design, criterion validity, responsiveness from a criterion approach, and responsiveness in comparison with other outcome measures—could not be estimated because of an unknown level of evidence.

Reliability and measurement error were investigated in only 4 studies. The 3 studies that evaluated test-retest reliability^6,62,68 found high reliability, but all had an RoB rating of inadequate; however, the 1 study with an intrasession design¹⁶ received an RoB rating of very good but indicated poor intrasession reliability. The low intrasession intraclass correlation coefficient of 0.06 and relatively large accompanying SEM of 2.7° indicated large within-patient variability among trials within the same session. These findings support calculating a mean value from multiple trials to avoid erroneous data. A previous study of knee JPS tests among individuals with patellofemoral pain syndrome determined that 5 trials were required for stable estimates for active procedures, whereas 6 trials were required for a passive knee JPS test.⁹⁶ A previous systematic review investigating the reliability of knee JPS tests using pathologic and asymptomatic cases found mixed evidence from 18 studies, largely dependent on the type of test methods.¹⁰⁰ The authors cited a lack of studies investigating reliability of the test among pathologic populations such as those with ACL injury. Thus, studies with a low RoB are required to establish the quality of reliability and measurement error for knee JPS tests among individuals with ACL injury and to improve the level of evidence.

Criterion (concurrent) validity was examined in only 1 study³⁰ with an RoB rating of inadequate, highlighting the need for further research in this area. This is of interest given the differences in results seen in the current review for the subgroup analyses of known-groups validity (see Figure 2), which showed mixed results for different test procedures. As there is no gold standard available to compare JPS tests, the criterion validity relating to the gold standard remains uncertain; however, additional studies are warranted on concurrent and predictive validity of JPS tests in individuals with an ACL injury.

For convergent validity, knee JPS test procedures were compared with many other diverse outcome measures, such as self-reported satisfaction and biomechanical measures of gait, making it difficult to draw conclusions regarding this PMP. However, >90% of findings did not find a correlation with the specific knee JPS tests and their comparator outcome measures, indicating a lack of correlation with at least those tested in this review. Thus, further studies are required to provide evidence for the convergent validity of knee JPS tests targeting individuals with ACL injury.

Our results for known-groups validity are in line with 1 previous meta-analysis⁸⁷ but in contrast to a systematic review without meta-analysis,⁷³ finding that ACL-injured knees produced significantly greater AE (MD, 0.70°) (see Figure 2) as compared with asymptomatic controls, regardless of whether ACLD and ACLR groups were combined or evaluated separately. In addition, our sensitivity analysis, which was limited to studies with an adequate or very good rating for RoB, showed even greater differences in errors (MD, 2.97°) (see Figure 3). However, subgroup analyses indicated that these results may depend on the test procedure, with only the knee JPS tests involving passive-passive test procedures producing statistically significant results; yet, passive-active (P = .06) and active-active (P = .05) procedures showed a trend toward statistical significance. It is possible that active procedures may facilitate compensation for potential loss of proprioceptors (Ruffini nerve endings, Pacini receptors, and Golgi tendon organ–like endings) in the injured/reconstructed ligament by relying more on muscle spindles. Further studies with a low RoB (very good rating) are required to substantiate these findings.

Our results for discriminative validity were in agreement with 2 previous meta-analyses,^58,87 finding significantly greater knee JPS errors for ACL-injured knees than for the contralateral noninjured knees of the same individuals (MD, 0.67°) (see Figure 8). When analyzed separately, however, only ACLD knees (MD, 0.95°) (Figure 10), not ACLR knees, showed significantly greater errors. Sensitivity analyses (ie, studies with a very good or adequate RoB rating) did nevertheless reveal a statistically significant difference for the separate analysis of ACLR groups (MD, 1.36°) (Figure 13) as well as even greater differences for the combined analysis (MD, 1.55°) (Figure 9) and for ACLD groups (MD, 1.81°) (Figure 11). Similar to known-groups validity, active-active test procedures did not differentiate between knees, supporting the hypothesis of compensatory mechanisms during active movements.

Responsiveness as compared with a gold standard or other knee JPS test over time was not assessed in any study, and responsiveness versus other outcome measurement instruments was examined in only 1 study of inadequate quality.⁷⁷ Thus, these PMPs require further research to provide evidence for their quality. Despite responsiveness among subgroups being analyzed in 10 studies, a meta-analysis was not possible owing to the studies having an RoB rating of inadequate or a lack of available data.^{3,6,9,23,38,42,68,98,106,113} Although a meta-analysis was performed for responsiveness to interventions, the lack of statistically significant differences suggests that the knee JPS test methods might not be sensitive enough to detect change over time among the ACL populations. However, differences among JPS tests, types of interventions, and study populations make it difficult to generalize conclusions from these findings. Additional research investigating the responsiveness of knee JPS tests targeting individuals with ACL injury is therefore required.

Limitations of this review should be acknowledged. Recommending specific knee JPS tests based on our results is difficult, given the many paradigms used for assessment and their often poorly reported methods and outcomes. Although we differentiated among tests by performing subgroup analyses based on test procedures, it would be of value to investigate the effects on outcome measures of other modifiable components, not least TA and movement direction. It should be noted that AE was the sole outcome measure considered in this review as it was by far the most commonly reported. However, constant and variable error can provide valuable information regarding JPS and may provide different results when compared with AE. Specifically, AE provides the absolute difference between the TA and reproduction angle without consideration for the direction of the error, whereas constant error provides a measure of systematic bias by considering the direction of the error (ie, whether the individual under- or overshoots the TA). In contrast, variable error indicates the variance of the errors regardless of the proximity to the TA and can be considered a measure of precision.⁹⁴ To increase the generalizability of our results, we included a range of ACL populations. We also differentiated between populations in our meta-analyses by performing separate analyses for ACLD and ACLR groups when appropriate. Nevertheless, ACL groups differed with regard to, for example, age, sex, time since injury, activity level, and body composition, as well as time since and type of surgery in the case of the ACLR groups. Further investigations focusing on the potential effects of these factors would be of clinical and scientific value.

Studies that had received an RoB rating of inadequate were precluded from qualitative and quantitative examinations of quality. This was done to provide a more accurate evaluation of the quality of each PMP as opposed to including studies regardless of their methodologic quality. For qualitative syntheses, however, a post hoc sensitivity analysis of studies with an RoB rating of inadequate showed little effect on the overall quality ratings of the respective PMPs and their levels of evidence. To assess RoB, we used the COSMIN RoB checklist⁷⁰ but adhered to alternative guidelines, as stated in our protocol.⁸ The latest COSMIN guidelines (2018) now cautiously recommend the inclusion of studies with an RoB rating of inadequate in analysis syntheses, as long as their findings are in line with those of the studies with a better RoB rating; thus, future studies may consider whether this is appropriate.

The lack of standardized reporting of knee JPS methods among studies, as well as unclear or missing information, made compiling evidence difficult. For example, we provided specific starting angles and TAs whenever this was indicated; however, 26 tests used unspecific angles within ranges of motion, and 15 tests failed to cite 1 or both angles. Finally, statistical heterogeneity was present in the majority of meta-analyses in this systematic review; thus, according to our stated criteria, no funnel plots with trim-and-fill analyses were produced to identify and correct for potential publication bias. There is therefore a possibility that publication bias might have been present in our findings, and caution should be taken while interpreting the results. Nevertheless, we searched 8 databases between their inception and March 2020 and included as many published studies (in English) from peer-reviewed journals as available in this area.

Future studies implementing knee JPS tests should publish clear and detailed testing methods. Reporting of outcomes should include a minimum of group mean and SD for AE with correlation, change scores, or both when relevant and also for constant and variable error. In addition, to further differentiate among ACL populations, it may be pertinent to control for and report the following: time since injury, time to surgery in the case of individuals with ACLR, preinjury and current activity level, leg dominance, side of injury, other concomitant injuries, demographic factors (age and sex), and anthropometric factors (body height and mass).

Conclusion

This systematic review and meta-analysis identified specific gaps in knowledge and provided evidence that can help to guide future investigations. Meta-analyses indicated sufficient quality and a strong level of evidence for known-groups and discriminative validity of knee JPS tests, given the significantly greater AEs for ACL-injured knees as compared with asymptomatic controls and the asymptomatic contralateral knees of the same individuals. Furthermore, passive test procedures appeared to elucidate greater differences between groups than did active test procedures. However, responsiveness to intervention was rated insufficient with a moderate level of evidence, owing to a lack of statistically significant change in AE over time after various interventions and limited studies of low RoB, respectively. The remaining PMPs in this review—reliability, criterion validity (concurrent and predictive), convergent validity, and responsiveness (criterion approach and construct approach between other outcome measurements as well as subgroups)—were assessed only qualitatively and require more studies with a low RoB to increase their level of evidence to better evaluate their quality.

Supplemental material for this article is available at http://journals.sagepub.com/doi/suppl/10.1177/23259671211007878.

Supplemental Material

Supplemental Material, sj-docx-1-ojs-10.1177_23259671211007878 - Properties of Knee Joint Position Sense Tests for Anterior Cruciate Ligament Injury: A Systematic Review and Meta-analysis

Supplemental Material, sj-docx-1-ojs-10.1177_23259671211007878 for Properties of Knee Joint Position Sense Tests for Anterior Cruciate Ligament Injury: A Systematic Review and Meta-analysis by Andrew Strong, Ashokan Arumugam, Eva Tengman, Ulrik Röijezon and Charlotte K. Häger in Orthopaedic Journal of Sports Medicine

Footnotes

Acknowledgment

The authors thank Anna Cronström for her critical review of the manuscript.

Final revision submitted November 9, 2020; accepted January 1, 2021.

One or more of the authors has declared the following potential conflict of interest or source of funding: This work was supported by the Swedish Research Council (grant 2017-00892); Region Västerbotten (grant ALF 7003575 VLL838421, strategic funding VLL-358901, project No. 7002795); the Swedish Research Council for Sports Science (grant CIF P2019-0068), Umeå School of Sport Science, Umeå University; and King Gustaf V and Queen Victoria’s Foundation of Freemasons 2019 (Dr Häger). The funders did not have any role in the study design. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.

Notes

References

Adachi

Ochi

Uchio

, et al. Mechanoreceptors in the anterior cruciate ligament contribute to the joint position sense. Acta Orthop Scand. 2002;73(3):330–334.

Ahmad

Shubin Stein

Jeshuran

Nercessian

Henry

. Anterior cruciate ligament function after tibial eminence fracture in skeletally mature patients. Am J Sports Med. 2001;29(3):339–345.

Park

Lee

. Effects of acceleration training 24 weeks after anterior cruciate ligament reconstruction on proprioceptive and dynamic balancing functions. J Phys Ther Sci. 2015;27(9):2825–2828.

Anders

Venbrocks

Weinberg

. Proprioceptive skills and functional outcome after anterior cruciate ligament reconstruction with a bone-tendon-bone graft. Int Orthop. 2008;32(5):627–633.

Andonovski

Topuzovska

Samardziski

, et al. The influence of anterior cruciate ligament remnant on postoperative clinical results in patients with remnant preserving anterior cruciate ligament reconstruction. Open Access Maced J Med Sci. 2017;5(5):624–629.

Angoules

Mavrogenis

Dimitriou

, et al. Knee proprioception following ACL reconstruction: a prospective trial comparing hamstrings with bone--patellar tendon—bone autograft. Knee. 2011;18(2):76–82.

Arockiaraj

Korula

Oommen

, et al. Proprioceptive changes in the contralateral knee joint following anterior cruciate injury. Bone Joint J. 2013;95-B(2):188–191.

Arumugam

Strong

Tengman

Röijezon

Häger

. Psychometric properties of knee proprioception tests targeting healthy individuals and those with anterior cruciate ligament injury managed with or without reconstruction: a systematic review protocol. BMJ Open. 2019;9(4):e027241.

Baltaci

Harput

Haksever

Ulusoy

Özer

. Comparison between Nintendo Wii Fit and conventional rehabilitation on functional performance outcomes after hamstring anterior cruciate ligament reconstruction: prospective, randomized, controlled, double-blind clinical trial. Knee Surg Sports Traumatol Arthrosc. 2013;21(4):880–887.

10.

Barrett

. Proprioception and function after anterior cruciate reconstruction. J Bone Joint Surg Br. 1991;73(5):833–837.

11.

Başar

Aybar

Kurtan

Başar

. The effects of partial meniscectomy and meniscal repair on the knee proprioception and function. J Orthop Surg (Hong Kong). 2020;28(1):2309499019894915.

12.

Baumeister

Reinecke

Weiss

. Changed cortical activity after anterior cruciate ligament reconstruction in a joint position paradigm: an EEG study. Scand J Med Sci Sports. 2008;18(4):473–484.

13.

Benjaminse

Otten

. ACL injury prevention, more effective with a different way of motor learning? Knee Surg Sports Traumatol Arthrosc. 2011;19(4):622–627.

14.

Birmingham

Kramer

Kirkley

, et al. Knee bracing after ACL reconstruction: effects on postural control and proprioception. Med Sci Sports Exerc. 2001;33(8):1253–1258.

15.

Bischoff

Babisch

, et al. Effects on proprioception by Kinesio taping of the knee after anterior cruciate ligament rupture. Eur J Orthop Surg Traumatol. 2018;28(6):1157–1164.

16.

Blackburn

Pietrosimone

Spang

Goodwin

Johnston

. Somatosensory function influences aberrant gait biomechanics following anterior cruciate ligament reconstruction. J Orthop Res. 2020;38(3):620–628.

17.

Bonfim

Jansen Paccola

Barela

. Proprioceptive and behavior impairments in individuals with anterior cruciate ligament reconstructed knees. Arch Phys Med Rehabil. 2003;84(8):1217–1223.

18.

Borenstein

Hedges

Higgins

Rothstein

. A basic introduction to fixed-effect and random-effects models for meta-analysis. Res Synth Methods. 2010;1(2):97–111.

19.

Bűyűkafşar

Başar

Kanatli

. Proprioception following the anterior cruciate ligament reconstruction with tibialis anterior tendon allograft. J Knee Surg. 2020;33(7):722–727.

20.

Çabuk

Kuşku Çabuk

. Mechanoreceptors of the ligaments and tendons around the knee. Clin Anat. 2016;29(6):789–795.

21.

Carter

Jenkinson

Wilson

Jones

Torode

. Joint position sense and rehabilitation in the anterior cruciate ligament deficient knee. Br J Sports Med. 1997;31(3):209–212.

22.

Chen

Wang

. Anterior cruciate ligament reconstruction using an anterior cruciate ligament stump. Wideochir Inne Tech Maloinwazyjne. 2019;14(3):461–467.

23.

Cho

Bae

Gak

. Effects of closed kinetic chain exercises on proprioception and functional scores of the knee after anterior cruciate ligament reconstruction. J Phys Ther Sci. 2013;25(10):1239–1241.

24.

Chouteau

Testa

Viste

Moyen

. Knee rotational laxity and proprioceptive function 2 years after partial ACL reconstruction. Knee Surg Sports Traumatol Arthrosc. 2012;20(4):762–766.

25.

Skinner

Cannon

. Effect of reconstruction of the anterior cruciate ligament on proprioception of the knee and the heel strike transient. J Orthop Res. 1993;11(5):696–704.

26.

Corrigan

Cashman

Brady

. Proprioception in the cruciate deficient knee. J Bone Joint Surg Br. 1992;74(2):247–250.

27.

Cossich

Mallrich

Titonelli

, et al. Proprioceptive deficit in individuals with unilateral tearing of the anterior cruciate ligament after active evaluation of the sense of joint position. Rev Bras Ortop. 2014;49(6):607–612.

28.

Deeks

Higgins

JPT

Altman

. Analysing data and undertaking meta-analyses. In: Higgins

JPT

Thomas

Chandler

, et al., eds. Cochrane Handbook for Systematic Reviews of Interventions Version 6.0. Updated July 2019. Cochrane; 2019:chap 10. http://www.training.cochrane.org/handbook

29.

Duval

Tweedie

. Trim and fill: a simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics. 2000;56(2):455–463.

30.

Dvir

Koren

Halperin

. Knee joint position sense following reconstruction of the anterior cruciate ligament. J Orthop Sports Phys Ther. 1988;10(4):117–120.

31.

Faggal

Abdelsalam

Adel Elhakk

Mahmoud

. Proprioceptive training after ACL reconstruction: standard versus stump preservation technique. Physiother Pract Res. 2019;40(1):69–75.

32.

Fischer-Rasmussen

Jensen

. Proprioceptive sensitivity and performance in anterior cruciate ligament–deficient knee joints. Scand J Med Sci Sports. 2000;10(2):85–89.

33.

Fischer-Rasmussen

Jensen

TØ

Kjær

, et al.

Is proprioception altered during loaded knee extension shortly after ACL rupture?

Int J Sports Med. 2001;22(5):385–391.

34.

Fonseca

Ocarino

Silva

, et al. Proprioception in individuals with ACL-deficient knee and good muscular and functional performance. Res Sports Med. 2005;13(1):47–61.

35.

Fremerey

Lobenhoffer

Zeichen

, et al. Proprioception after rehabilitation and reconstruction in knees with deficiency of the anterior cruciate ligament: a prospective, longitudinal study. J Bone Joint Surg Br. 2000;82(6):801–806.

36.

Fridén

Roberts

Zätterström

Lindstrand

Moritz

. Proprioception after an acute knee ligament injury: a longitudinal study on 16 consecutive patients. J Orthop Res. 1997;15(5):637–644.

37.

Fridén

Roberts

Zätterström

Lindstrand

Moritz

. Proprioception in the nearly extended knee: measurements of position and movement in healthy individuals and in symptomatic anterior cruciate ligament injured patients. Knee Surg Sports Traumatol Arthrosc. 1996;4(4):217–224.

38.

Yung

Law

, et al. The effect of early whole-body vibration therapy on neuromuscular control after anterior cruciate ligament reconstruction: a randomized controlled trial. Am J Sports Med. 2013;41(4):804–814.

39.

Furlanetto

Peyré-Tartaruga

do Pinho

Bernardes Eda

Zaro

. Proprioception, body balance and functionality in individuals with ACL reconstruction. Acta Ortop Bras. 2016;24(2):67–72.

40.

Gao

Zhou

, et al. A morphologic and quantitative study of mechanoreceptors in the remnant stump of the human anterior cruciate ligament. Arthroscopy. 2016;32(2):273–280.

41.

Georgoulis

Pappa

Moebius

, et al. The presence of proprioceptive mechanoreceptors in the remnants of the ruptured ACL as a possible source of re-innervation of the ACL autograft. Knee Surg Sports Traumatol Arthrosc. 2001;9(6):364–368.

42.

Goetschius

Kuenze

Saliba

Hart

. Reposition acuity and postural control after exercise in anterior cruciate ligament reconstructed knees. Med Sci Sports Exerc. 2013;45(12):2314–2321.

43.

Good

Roos

Gottlieb

Renström

Beynnon

. Joint position sense is not changed after acute disruption of the anterior cruciate ligament. Acta Orthop Scand. 1999;70(2):194–198.

44.

Guney-Deniz

Harput

Kaya

Nyland

Doral

. Quadriceps tendon autograft ACL reconstructed subjects overshoot target knee extension angle during active proprioception testing. Knee Surg Sports Traumatol Arthrosc. 2020;28(2):645–652.

45.

Gupta

Aggarwal

Garg

Kumar

. Preserved insertions of the semitendinosus and gracilis tendons (STG) in ACL reconstruction: a new surgical technique with preliminary results. Curr Orthop Pract. 2010;21(4):409–414.

46.

Han

Waddington

Adams

Anson

Liu

. Assessing proprioception: a critical review of methods. J Sport Health Sci. 2016;5(1):80–90.

47.

Harter

Osternig

Singer

. Knee joint proprioception following anterior cruciate ligament reconstruction. J Sport Rehab il. 1992;1(2):103–110.

48.

Hewett

Paterno

Myer

. Strategies for enhancing proprioception and neuromuscular control of the knee. Clin Orthop Relat Res. 2002;402:76–94.

49.

Hillier

Immink

Thewlis

. Assessing proprioception: a systematic review of possibilities. Neurorehabil Neural Repair. 2015;29(10):933–949.

50.

Hopper

Creagh

Formby

, et al. Functional measurement of knee joint position sense after anterior cruciate ligament reconstruction. Arch Phys Med Rehabil. 2003;84(6):868–872.

51.

Ioannidis

Trikalinos

. The appropriateness of asymmetry tests for publication bias in meta-analyses: a large survey. CMAJ. 2007;176(8):1091–1096.

52.

Iwasa

Ochi

Adachi

, et al. Proprioceptive improvement in knees with anterior cruciate ligament reconstruction. Clin Orthop Relat Res. 2000;381:168–176.

53.

Jensen

TØ

Fischer-Rasmussen

Kjær

Magnusson

. Proprioception in poor- and well-functioning anterior cruciate ligament deficient patients. J Rehabil Med. 2002;34(3):141–149.

54.

Kannabiran

. Variation in joint position sense in the contralateral knee following unilateral ACL injury. MOJ Orthop Rheumatol. 2016;4(3):138.

55.

Karasel

Akpinar

Gülbahar

, et al. Clinical and functional outcomes and proprioception after a modified accelerated rehabilitation program following anterior cruciate ligament reconstruction with patellar tendon autograft. Acta Orthop Traumatol Turc. 2010;44(3):220–228.

56.

Katayama

Higuchi

Kimura

, et al. Proprioception and performance after anterior cruciate ligament rupture. Int Orthop. 2004;28(5):278–281.

57.

Kaya

Guney-Deniz

Sayaca

Calik

Doral

. Effects on lower extremity neuromuscular control exercises on knee proprioception, muscle strength, and functional level in patients with ACL reconstruction. Biomed Res Int. 2019;2019:1694695.

58.

Kim

Lee

. Proprioception in patients with anterior cruciate ligament tears: a meta-analysis comparing injured and uninjured limbs. Am J Sports Med. 2017;45(12):2916–2922.

59.

Kovalak

Atay

Çetin

Atay

Serbest

. Is ACL reconstruction a prerequisite for the patients having recreational sporting activities? Acta Orthop Traumatol Turc. 2018;52(1):37–43.

60.

Kroman

Roos

Bennell

Hinman

Dobson

. Measurement properties of performance-based outcome measures to assess physical function in young and middle-aged people known to be at high risk of hip and/or knee osteoarthritis: a systematic review. Osteoarthritis Cartilage. 2014;22(1):26–39.

61.

Lee

Kwon

Kim

Choi

Min

. Comparison of clinical results according to amount of preserved remnant in arthroscopic anterior cruciate ligament reconstruction using quadrupled hamstring graft. Arthroscopy. 2008;24(5):560–568.

62.

Lee

Ahn

Park

. Effect of time after anterior cruciate ligament tears on proprioception and postural stability. PLoS One. 2015;10(9):e0139038.

63.

Lee

Cheng

Liau

. Correlation between proprioception, muscle strength, knee laxity, and dynamic standing balance in patients with chronic anterior cruciate ligament deficiency. Knee. 2009;16(5):387–391.

64.

Liu

. Correlation study of knee joint proprioception test results using common test methods. J Phys Ther Sci. 2016;28(2):478–482.

65.

Littmann

Iguchi

Madhavan

Kolarik

Shields

. Dynamic-position-sense impairment’s independence of perceived knee function in women with ACL reconstruction. J Sport Rehabil. 2012;21(1):44–53.

66.

Liu

Qian

Gao

Ruan

. Effects of Kinesio taping of the knee on proprioception, balance, and functional performance in patients with anterior cruciate ligament rupture: a retrospective case series. Medicine (Baltimore). 2019;98(48):e17956.

67.

Mir

Hadian

Talebian

Nasseri

. Functional assessment of knee joint position sense following anterior cruciate ligament reconstruction. Br J Sports Med. 2008;42(4):300–303.

68.

Moezy

Olyaei

Hadian

Razi

Faghihzadeh

. A comparative study of whole body vibration training and conventional training on knee proprioception and postural stability after anterior cruciate ligament reconstruction. Br J Sports Med. 2008;42(5):373–378.

69.

Moher

Shamseer

Clarke

, et al. Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4(1):1.

70.

Mokkink

de Vet

HCW

Prinsen

CAC

, et al. COSMIN Risk of Bias checklist for systematic reviews of patient-reported outcome measures. Qual Life Res. 2018;27(5):1171–1179.

71.

Nagai

Bates

Hewett

Schilaty

. Effects of localized vibration on knee joint position sense in individuals with anterior cruciate ligament reconstruction. Clin Biomech (Bristol, Avon). 2018;55:40–44.

72.

Nagelli

Hewett

. Should return to sport be delayed until 2 years after anterior cruciate ligament reconstruction? Biological and functional considerations. Sports Med. 2017;47(2):221–232.

73.

Nakamae

Adachi

Ishikawa

Nakasa

Ochi

. No evidence of impaired proprioceptive function in subjects with anterior cruciate ligament reconstruction: a systematic review. J ISAKOS. 2017;2(4):191–199.

74.

Nishiwaki

Tanaka

Urabe

. The effect of muscle strength on proprioceptive function after anterior cruciate ligament reconstruction of the knee. Jpn J Phys Fit Sport Med. 2007;56(5):451–460.

75.

Nyland

Gamble

Franklin

Caborn

DNM

. Permanent knee sensorimotor system changes following ACL injury and surgery. Knee Surg Sports Traumatol Arthrosc. 2017;25(5):1461–1474.

76.

Ochi

Adachi

Uchio

, et al. A minimum 2-year follow-up after selective anteromedial or posterolateral bundle anterior cruciate ligament reconstruction. Arthroscopy. 2009;25(2):117–122.

77.

Ochi

Iwasa

Uchio

Adachi

Kawasaki

. Induction of somatosensory evoked potentials by mechanical stimulation in reconstructed anterior cruciate ligaments. J Bone Joint Surg Br. 2002;84(5):761–766.

78.

Ochi

Iwasa

Uchio

Adachi

Sumen

. The regeneration of sensory neurones in the reconstruction of the anterior cruciate ligament. J Bone Joint Surg Br. 1999;81(5):902–906.

79.

Ordahan

Kűçűkşen

Tuncay

Salli

Uğurlu

. The effect of proprioception exercises on functional status in patients with anterior cruciate ligament reconstruction. J Back Musculoskelet Rehabil. 2015;28(3):531–537.

80.

Ozenci

Inanmaz

Ozcanli

, et al. Proprioceptive comparison of allograft and autograft anterior cruciate ligament reconstructions. Knee Surg Sports Traumatol Arthrosc. 2007;15(12):1432–1437.

81.

Prinsen

CAC

Mokkink

Bouter

, et al. COSMIN guideline for systematic reviews of patient-reported outcome measures. Qual Life Res. 2018;27(5):1147–1157.

82.

Proske

Gandevia

. The proprioceptive senses: their roles in signaling body shape, body position and movement, and muscle force. Physiol Rev. 2012;92(4):1651–1697.

83.

Rehm

Llopis-Miro

Turner

. The relationship between proprioception in the knee and the need for ligament reconstruction in the anterior cruciate ligament deficient knee. Knee. 1998;5(3):199–202.

84.

Reider

Arcand

Diehl

, et al. Proprioception of the knee before and after anterior cruciate ligament reconstruction. Arthroscopy. 2003;19(1):2–12.

85.

Relph

Herrington

. Knee joint position sense ability in elite athletes who have returned to international level play following ACL reconstruction: a cross-sectional study. Knee. 2016;23(6):1029–1034.

86.

Relph

Herrington

. The effect of conservatively treated ACL injury on knee joint position sense. Int J Sports Phys Ther. 2016;11(4):536–543.

87.

Relph

Herrington

Tyson

. The effects of ACL injury on knee proprioception: a meta-analysis. Physiotherapy. 2014;100(3):187–195.

88.

Remedios

Morris

Bendrups

. Reduced static proprioception of the knee joint following anterior cruciate ligament reconstruction. Physiother Can. 1998;50(4):299–315.

89.

Roberts

Fridén

Stomberg

Lindstrand

Moritz

. Bilateral proprioceptive defects in patients with a unilateral anterior cruciate ligament reconstruction: a comparison between patients and healthy individuals. J Orthop Res. 2000;18(4):565–571.

90.

Roberts

Fridén

Zätterström

Lindstrand

Moritz

. Proprioception in people with anterior cruciate ligament–deficient knees: comparison of symptomatic and asymptomatic patients. J Orthop Sports Phys Ther. 1999;29(10):587–594.

91.

Röijezon

Clark

Treleaven

Proprioception in musculoskeletal rehabilitation, part 1: basic science and principles of assessment and clinical interventions. Man Ther. 2015;20(3):368–377.

92.

San Martin-Mohr

Cristi-Sánchez

Pincheira

, et al. Knee sensorimotor control following anterior cruciate ligament reconstruction: a comparison between reconstruction techniques. PLoS One. 2018;13(11):e0205658.

93.

Sanders

Maradit Kremers

Bryan

, et al. Incidence of anterior cruciate ligament tears and reconstruction: a 21-year population-based study. Am J Sports Med. 2016;44(6):1502–1507.

94.

Schmidt

Lee

Winstein

Wulf

, Zelaznik HN. Motor Control and Learning: A Behavioral Emphasis. Human Kinetics; 2018.

95.

Schultz

Miller

Kerr

Micheli

. Mechanoreceptors in human cruciate ligaments: a histological study. J Bone Joint Surg Am. 1984;66(7):1072–1076.

96.

Selfe

Callaghan

McHenry

Richards

Oldham

. An investigation into the effect of number of trials during proprioceptive testing in patients with patellofemoral pain syndrome. J Orthop Res. 2006;24(6):1218–1224.

97.

Shamseer

Moher

Clarke

, et al. Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) 2015: elaboration and explanation. BMJ. 2015;350:G7647.

98.

Shen

Che

, et al. Effects of backward walking on knee proprioception after ACL reconstruction. Physiother Theory Pract. Published online October 21, 2019. doi:10.1080/09593985.2019.1681040

99.

Silva

Ribeiro

Oliveira

. Effect of an accelerated ACL rehabilitation protocol on knee proprioception and muscle strength after anterior cruciate ligament reconstruction. Arch Exerc Health Dis. 2012;3(1–2):139–144.

100.

Smith

Davies

Hing

. A systematic review to determine the reliability of knee joint position sense assessment measures. Knee. 2013;20(3):162–169.

101.

Strutzenberger

Braig

Sell

Boes

Schwameder

. Effect of brace design on patients with ACL-ruptures. Int J Sports Med. 2012;33(11):934–939.

102.

Suarez

Laudani

Giombini

, et al. Comparison in joint-position sense and muscle coactivation between anterior cruciate ligament–deficient and healthy individuals. J Sport Rehabil. 2016;25(1):64–69.

103.

Takeshima

Sozu

Tajika

, et al.

Which is more generalizable, powerful and interpretable in meta-analyses, mean difference or standardized mean difference?

BMC Med Res Methodol. 2014;14(1):30.

104.

Takuma

Hiroki

Yasuaki

Kazuyuki

Torn

. Comparison of the position-matching and position-reproducing tasks to detect deficits in knee position sense after reconstruction of the anterior cruciate ligament. J Sport Rehabil. 2020;29(1):87–92.

105.

Wan

Wang

Liu

Tong

. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. 2014;14:135.

106.

Wang

. Immediate effects of neuromuscular joint facilitation intervention after anterior cruciate ligament reconstruction. J Phys Ther Sci. 2016;28(7):2084–2087.

107.

Wiggins

Grandhi

Schneider

, et al. Risk of secondary injury in younger athletes after anterior cruciate ligament reconstruction: a systematic review and meta-analysis. Am J Sports Med. 2016;44(7):1861–1876.

108.

Mak

. Effects of knee bracing on the sensorimotor function of subjects with anterior cruciate ligament reconstruction. Am J Sports Med. 2001;29(5):641–645.

109.

Yosmaoğlu

Baltaci

Kaya

Özer

Atay

. Comparison of functional outcomes of two anterior cruciate ligament reconstruction methods with hamstring tendon graft. Acta Orthop Traumatol Turc. 2011;45(4):240–247.

110.

Zandiyeh

Kupper

Mohtadi

NGH

Goldsmith

Ronsky

. Effect of stochastic resonance on proprioception and kinesthesia in anterior cruciate ligament reconstructed patients. J Biomech. 2019;84:52–57.

111.

Zhou

Chen

, et al. Factors affecting proprioceptive recovery after anterior cruciate ligament reconstruction. Chin Med J (Engl). 2008;121(22):2224–2228.

112.

Zimny

Schutte

Dabezies

. Mechanoreceptors in the human anterior cruciate ligament. Anat Rec. 1986;214(2):204–209.

113.

Zult

Gokeler

van Raay

, et al. Cross-education does not accelerate the rehabilitation of neuromuscular functions after ACL reconstruction: a randomized controlled clinical trial. Eur J Appl Physiol. 2018;118(8):1609–1623.

114.

Zult

Gokeler

van Raay

, et al. An anterior cruciate ligament injury does not affect the neuromuscular function of the non-injured leg except for dynamic balance and voluntary quadriceps activation. Knee Surg Sports Traumatol Arthrosc. 2017;25(1):172–183.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.21 MB

0.73 MB