Hirayama disease (HD), also known as juvenile muscular atrophy of the distal upper extremity, is a benign neurological disease that was first reported by Hirayama in 1959.1 Based on the characteristic “loss of attachment” (abnormal forward movement of the posterior dura) on cervical-flexion magnetic resonance imaging (MRI),2–5 the main hypothesis regarding the etiology of HD implicates chronic ischemia of cervical motor neurons caused by cervical flexion.6 Therefore, some physicians have performed anterior cervical fusion (ACF) procedures for HD patients to limit abnormal cervical flexion.7–10 Although these studies reported satisfying outcomes for most HD patients, procedures were performed in a small number of cases because not all HD patients can obtain a good surgical prognosis. A previous study demonstrated that approximately 10% of HD patients experience disease progression after surgery, potentially due to inadequate cervical fusion segment length.10
As a characteristic abnormality on MRI, loss of attachment is closely related to the pathogenesis of HD, and it has been suggested that the extent of loss of attachment may correlate with the severity of HD.11 More importantly, a recently published study demonstrated that residual loss of attachment at adjacent segments of the surgical site is related to the poor surgical outcome in HD.10 Therefore, loss of attachment may be one of the most important factors affecting the surgical prognosis for HD.
In this study, the authors sought to identify the change of abnormal loss of attachment in patients with HD after ACF. They also analyzed the influence of the measurements of loss of attachment on the surgical outcome in HD.
Materials and Methods
A total of 42 patients were included in this study (Table 1). All patients were diagnosed as having HD and admitted to the Department of Orthopedics, Huashan Hospital, Fudan University, Shanghai, China, from February 2014 to November 2016. All patients underwent ACF for two segments with the most severe loss of attachment.
Patients' Clinical Information
The criteria for inclusion were (1) insidious disease onset before 25 years of age; (2) unilateral or asymmetric weakness and amyotrophy of the distal upper extremities without sensory dysfunction or lower limb involvement, with denervation limited to the unilateral or bilateral upper extremities and identified by electrophysiological methods, with normal sensory nerve function; (3) gradually slower clinical progression after a relatively rapid initial progression; (4) clinical symptoms such as “cold paralysis,” “fasciculation,” or “tremor of the upper extremities,” which may or may not be present; and (5) cervical-flexion MRI showing lower cervical compression resulting from forward shifting of the posterior dura and a crescent abnormality posterior to the dura.
The criteria for exclusion were (1) a history of syringomyelia, spinal cord tumor, or abnormalities of the cervical vertebrae; (2) focal or multifocal neuropathy; (3) brachial plexus injury; (4) congenital muscular dystrophy; (5) cervical spondylotic amyotrophy; (6) primary or concomitant disorder of the neuromuscular junction; or (7) concomitant trauma, inflammation, or infection.
Dynamic MRI Detection
All patients underwent neutral and cervical-flexion MRI before surgical fusion, with the flexion MRI repeated at the 3-month postoperative visit. A triangular pillow was used to keep the patient's neck in the flexed position during the cervicalflexion MRI.12 The measurements of loss of attachment were performed on both cross-sectional and sagittal MRI views.11,13
On the cross-sectional MRI view, the following measurements were taken at the cervical segment that displayed maximum forward shifting (MFS) of the cervical cord on the cervical-flexion scan: the distance between the posterior edge of the spinal cord and the posterior wall of the spinal canal (X), the anteroposterior diameter of the spinal canal (Y), the anteroposterior diameter of the cervical cord (A), and the transverse diameter of the cervical cord (B) (Figure 1). The cervical-flexion X/Y was used to indicate the MFS degree in the cervical cord, the cervical-flexion A/B indicated the relative morphological changes of the cervical cord, and the neutral A/B indicated the degree of diminution. On the sagittal MRI view, the longitudinal separation range (LSR) of the loss of attachment was measured in the flexed position. Imaging software (Centricity Enterprise web 3.0; United Imaging) was used for these measurements.
On cross-sectional magnetic resonance imaging, measurement of the cervical spine in neutral position (A, B) and flexion position (C, D) in patients with Hirayama disease. A, the anteroposterior diameter of the cervical cord; B, the transverse diameter of the cervical cord; X, the distance between the posterior edge of the spinal cord and the posterior wall of the spinal canal; Y, the anteroposterior diameter of the spinal canal.
Clinical Motor Function Measures
All 42 patients had further muscle strength examination, underwent handgrip strength (HGS) testing (Jamar hydraulic hand dynamometer; Sammons Preston), and completed the Disabilities of the Arm, Shoulder and Hand (DASH) questionnaire before surgery and at the 1-year postoperative visit. Muscle strength was assessed by manual testing and summarized as the Medical Research Council (MRC) score in the abductor digiti mini-mi and abductor pollicis brevis bilaterally. The MRC sum-scores were calculated by adding the standards previously described in the literature.14 All patients performed a maximum of three attempts for each HGS measurement, with 1 minute of rest between attempts. The mean value of the three attempts was recorded as the HGS measurement.
The normal values of the measurements on both dynamic MRI and HGS testing were defined based on recommendations from previous studies.11 Two orthopedists (Q.Y., S.T.) performed all clinical examinations and imaging measurements. These procedures were further verified by another experienced orthopedist (X.J.). All were blinded to the patients' clinical information.
The statistical analysis was performed using SPSS, version 22.0, software (IBM Corp). The Kolmogorov–Smirnov test was used for the normal distribution test. The difference between healthy control and HD patients was analyzed by t test, and the difference before and after surgery was analyzed by paired-samples t test. The Wilcoxon rank-sum test was used when the statistics did not satisfy the normal distribution. Pearson correlation coefficient analysis was used to analyze the correlation between imaging abnormality and clinical outcomes. Furthermore, the parameters with statistical difference (P<.05) on the independent-samples t test were further analyzed by logistic regression. P<.05 indicated significance.
The measurements of MRI detection and clinical function for all 42 patients are listed in Table A, available in the online version of the article.
The pre-operative and post-operative MRI parameters and upper extremity function
Compared with healthy controls, increased cervical-flexion X/Y and decreased cervical-flexion A/B were present in HD patients (P<.05). There was no statistical difference between the X/Y and A/B of neutral position, but the X/Y of neutral position was increased (t=1.761, P=.084). Bilateral HGS and MRC scores of HD patients were lower than those of healthy controls (P<.05). There was a significant correlation between DASH scores and the measurements of loss of attachment (LSR: r=0.798, P<.01; cervical-flexion X/Y: r=0.442, P<.01; cervical-flexion A/B: r=−0.352, P<.05).
At a mean of 94.17±8.65 days (range, 75–110 days) postoperative, both the cervical-flexion X/Y and the LSR were significantly decreased (P<.01), whereas the cervical-flexion A/B was increased (P<.01) (Table A; Figure 2). The characteristic loss of attachment completely resolved at the fused segments postoperatively in all 42 HD patients (Figure 3). However, in 7 (16.7%) of the 42 patients, residual loss of attachment during cervical flexion remained at the segment adjacent to the fused levels (Figure 4). A significantly longer LSR was found preoperatively in these 7 patients compared with the others (P<.05).
Measurement for the “loss of attachment” of patients with Hirayama disease. No obvious abnormality at neutral position on cervical magnetic resonance imaging (MRI) (A). The loss of attachment centered on C5-6 and affected 4 segments in all at the flexion position on cervical MRI (B). The loss of attachment appeared at T2-T3 but was still mild at neutral position on cervical MRI (C). The loss of attachment centered on C5-6 and C6-7, although the loss of attachment could be found under T3. The authors calculated the segments with loss of attachment as 6 because the loss of attachment did not change obviously under T3 (D). Arrows indicate the range of loss of attachment.
On sagittal magnetic resonance imaging (MRI) for this 23-year-old man, the “loss of attachment” at the flexion position could be found from C4 to T1. C5-C7 was fixed in the surgery, and the loss of attachment disappeared. Cross-sectional and sagittal cervical MRI at neutral (A, D) and flexion (B, E) positions before surgery. Cross-sectional and sagittal cervical MRI at flexion position after surgery (C, F).
On sagittal magnetic resonance imaging at flexion position, “loss of attachment” could be found in more than 5 segments of this 18-year-old man with Hirayama disease before surgery (A, C). Loss of attachment disappeared at the surgical site (B, D), while residual loss of attachment and forward movement of the cervical cord could be found at the adjacent cervical segments of the surgical site (B, E).
At a mean of 387.26±32.67 days (range, 332–470 days) postoperative, bilateral HGS and MRC scores of HD patients were not different from preoperative (P>.05). There was no significant difference between preoperative and postoperative DASH scores of HD patients (P>.05).
Twenty (47.6%) of the 42 patients' DASH scores had decreased at the 1-year postoperative visit, and 9 (45%) of these 20 patients presented with improved MRC scores in at least one of the tested muscles. Twelve (28.6%) of the 42 patients had no changes in DASH scores or MRC scores. The remaining 10 patients (23.8%) had mildly increased DASH scores without an increase in MRC scores, and 6 of these 10 patients had residual loss of attachment at the segments adjacent to the fused ones.
Measurements Between HD Patients With and Without Postoperative Functional Improvement
The 42 HD patients were divided into two groups based on whether their DASH scores had improved. Twenty (47.62%) of the 42 patients had a lower DASH score. Compared with the 22 patients (52.4%) who did not, the HD patients who had lower DASH scores had relatively shorter LSR, lower cervical-flexion X/Y, and higher cervical-flexion A/B preoperatively (P<.01; Table B, available in the online version of the article). According to the logistic regression analysis, LSR and cervical-flexion X/Y were related to the surgical outcomes. Receiver operating characteristic curve analysis was performed to identify both area under the curve and cutoff values of these two parameters for the prognosis of HD. The LSR had area under the curve, sensitivity, specificity, and cutoff values of 0.959, 0.909, 0.850, and 4.5, respectively. The MFS degree in the cervical cord had area under the curve, sensitivity, specificity, and cutoff values of 0.782, 0.818, 0.700, and 0.215, respectively.
The univariate analysis on the basis of whether the Dash score decrease or not
The results of this study demonstrated that ACF could improve the loss of attachment of HD patients. Further, obvious correlations existed between the abnormal findings on cervical-flexion MRI and the surgical outcomes. All of these findings collectively support that the extent of loss of attachment might be an important risk factor for the outcomes.
In this study, as in a previous study,11 significant correlation was observed between the measurements of loss of attachment and motor dysfunction in HD patients. More importantly, LSR, compared with the deformation of the spinal cord, presented a closer correlation with motor dysfunction in HD patients. It indicated that the ischemia is more likely to be the main reason for anterior horn cell injury. That the HD patients with residual loss of attachment at adjacent segments postoperatively had a relatively poorer prognosis further supported this.
The authors found that all of the HD patients had a significant improvement in abnormal loss of attachment on MRI postoperatively. Several previous studies have demonstrated that loss of attachment is a specific imaging manifestation of HD and is closely related to the pathogenic process of HD.15–18 A recent study proved that this imaging abnormality would not improve during the progression of the disease.11 Therefore, the improved findings on cervical-flexion MRI at the 3-month postoperative visit and in clinical function at the 1-year postoperative visit indicated that surgical fusion can effectively prevent or delay the aggravation of dyskinesia. Although not every HD patient who underwent surgical fusion had a lower DASH score or MRC score at the 1-year postoperative visit, this is significant.
Although several previous studies7–10 and the current study demonstrated that ACF procedures for HD could have a satisfying outcome, 10 patients in this study presented with mildly increased DASH scores 1 year after surgery. Six of these 10 patients had residual loss of attachment at the adjacent segments, which is likely the reason for the poorer prognosis. A similar result was found in a previous study,10 and the short surgical fusion segment was considered to be the reason for residual loss of attachment. This speculation was supported by the receiver operating characteristic curve analysis in the current study. When the LSR of loss of attachment is 5 or more cervical vertebra segments, a longer surgical fusion (>2 segments) may be required for a better surgical outcome.
In addition to the LSR, the MFS degree in the cervical spinal cord is a risk factor in the surgical outcome. Previous studies speculated that the blood supply of the spinal cord would partly flow into the space between the posterior edge of the spinal cord and the posterior wall of the spinal canal at flexion position in patients with HD.6,19,20 Therefore, ischemia may be much more severe in patients with larger cervical-flexion X/Y on MRI. In this study, the authors found a significant negative correlation between cervicalflexion X/Y and motor function, suggesting that severe ischemia of motor neurons often caused loss of motor neurons, which resulted in irreversible motor dysfunction. The cervical-flexion X/Y (MFS) of 0.215 or more may indicate a poor prognosis after surgical treatment.
In a previous study, Xu et al21 reported that HD patients had significantly increased ranges of flexed motion in cervical segments compared with healthy subjects, and that this cervical segmental instability may lead to the abnormal loss of attachment. Furthermore, relatively larger pre-operative neutral X/Y was observed in the HD patients with cervical kyphotic deformity, suggesting that the forward-shifting degree of the cervical cord at flexion position may also be affected by abnormality of cervical sagittal alignment. Therefore, HD patients with larger cervical-flexion X/Y may have potential cervical segmental instabilities and/or sagittal alignment abnormalities at the same time, and multilevel cervical fusion of more than 2 segments may be required to overcome these abnormalities.
When reviewing these results, it is critical to consider that one potential limitation was the lack of a conservative treatment group for comparison. A previous study indicated that collar therapy slowed progression in most cases22; however, some HD patients cannot tolerate wearing a neck collar continuously.23 A clinician-led guideline24 supports the opinions mentioned above. A previous study demonstrated that the main measurements of imaging abnormalities would not improve as the disease advanced.11
Another limitation was that the authors did not include the research regarding the posterior approach for fusion. There were several reasons why. First, the surgery aimed to limit the range of motion of the cervical spine, and the decompression is not necessary, so the ACF procedure is relatively safe. Second, previous studies suggested that HD was more likely attributed to the abnormality of the spine cord, instead of the abnormal ligament structure. In their previous study, the authors confirmed that ACF for 2 segments had a satisfying outcome; thus, the anterior approach was likely superior to the posterior approach. However, in a future study, the authors will compare the two approaches to determine which one is better.
Another potential limitation of this study was the small number of cases. More significant findings involving the prognosis for the longer surgical segments of HD patients with multiple segments with the loss of attachment might be achieved with more cases in the future. The authors will continuously follow up with these HD patients to confirm the outcome of ACF for more than 5 years.
Anterior cervical fusion procedures can effectively improve the abnormal loss of attachment and prevent further progress of disease in patients with HD. The LSR is an important risk factor for the prognosis. Longer fused segments may be required when the LSR is 5 segments or more.
- Hirayama K, Tomonaga M, Kitano K, Yamada T, Kojima S, Arai K. Focal cervical poliopathy causing juvenile muscular atrophy of distal upper extremity: a pathological study. J Neurol Neurosurg Psychiatry. 1987;50(3):285–290. doi:10.1136/jnnp.50.3.285 [CrossRef] PMID:3559609
- Tan CT. Juvenile muscular atrophy of distal upper extremities. J Neurol Neurosurg Psychiatry. 1985;48(3):285–286. doi:10.1136/jnnp.48.3.285 [CrossRef] PMID:3981204
- Schröder R, Keller E, Flacke S, et al. MRI findings in Hirayama's disease: flexion-induced cervical myelopathy or intrinsic motor neuron disease?J Neurol.1999;246(11):1069–1074. doi:10.1007/s004150050514 [CrossRef] PMID:10631640
- Sobue I, Saito N, Iida M, Ando K. Juvenile type of distal and segmental muscular atrophy of upper extremities. Ann Neurol. 1978;3(5):429–432. doi:10.1002/ana.410030512 [CrossRef] PMID:727722
- Zhou B, Chen L, Fan D, Zhou D. Clinical features of Hirayama disease in mainland China. Amyotroph Lateral Scler. 2010;11(1–2):133–139. doi:10.3109/17482960902912407 [CrossRef] PMID:19412815
- Hirayama K. [Juvenile muscular atrophy of unilateral upper extremity (Hirayama disease): half-century progress and establishment since its discovery]. Brain Nerve. 2008;60(1):17–29. PMID:18232329
- Imamura H, Matsumoto S, Hayase M, Oda Y, Kikuchi H, Takano M. [A case of Hirayama's disease successfully treated by anterior cervical decompression and fusion]. No To Shinkei. 2001;53(11):1033–1038. PMID:11761912
- Guo X, Lu M, Xie N, Guo Q, Ni B. Multilevel anterior cervical discectomy and fusion with plate fixation for juvenile unilateral muscular atrophy of the distal upper extremity accompanied by cervical kyphosis. J Spinal Disord Tech. 2014;27(7):E241–E246. doi:10.1097/BSD.0000000000000098 [CrossRef] PMID:24686334
- Lu F, Wang H, Jiang J, et al. Efficacy of anterior cervical decompression and fusion procedures for monomelic amyotrophy treatment: a prospective randomized controlled trial. Clinical article. J Neurosurg Spine. 2013;19(4):412–419. doi:10.3171/2013.4.SPINE12575 [CrossRef] PMID:23930718
- Zheng C, Nie C, Lei W, et al. Can anterior cervical fusion procedures prevent the progression of the natural course of Hirayama disease? An ambispective cohort analysis. Clin Neurophysiol. 2018;129(11):2341–2349. doi:10.1016/j.clinph.2018.08.024 [CrossRef] PMID:30248624
- Shao M, Yin J, Lu F, Zheng C, Wang H, Jiang J. The quantitative assessment of imaging features for the study of Hirayama disease progression. BioMed Res Int. 2015;2015:803148. doi:10.1155/2015/803148 [CrossRef] PMID:26558283
- Huang YC, Ro LS, Chang HS, et al. A clinical study of Hirayama disease in Taiwan. Muscle Nerve. 2008;37(5):576–582. doi:10.1002/mus.20980 [CrossRef]
- Lai V, Wong YC, Poon WL, Yuen MK, Fu YP, Wong OW. Forward shifting of posterior dural sac during flexion cervical magnetic resonance imaging in Hirayama disease: an initial study on normal subjects compared to patients with Hirayama disease. Eur J Radiol. 2011;80(3):724–728. doi:10.1016/j.ejrad.2010.07.021 [CrossRef] PMID:20727701
- Paternostro-Sluga T, Grim-Stieger M, Posch M, et al. Reliability and validity of the Medical Research Council (MRC) scale and a modified scale for testing muscle strength in patients with radial palsy. J Rehabil Med.2008;40(8):665–671. doi:10.2340/16501977-0235 [CrossRef] PMID:19020701
- Ghosh PS, Moodley M, Friedman NR, Rothner AD, Ghosh D. Hirayama disease in children from North America. J Child Neurol. 2011;26(12):1542–1547. doi:10.1177/0883073811409226 [CrossRef] PMID:22114244
- Yang G, Yang X, Zhang M, et al. Hirayama disease in children from mainland of China. J Child Neurol. 2014;29(4):509–513. doi:10.1177/0883073813482770 [CrossRef] PMID:23549046
- Hassan KM, Sahni H. Nosology of juvenile muscular atrophy of distal upper extremity: from monomelic amyotrophy to Hirayama disease—Indian perspective. BioMed Res Int. 2013;2013:478516. doi:10.1155/2013/478516 [CrossRef] PMID:24063005
- Elsheikh B, Kissel JT, Christoforidis G, et al. Spinal angiography and epidural venography in juvenile muscular atrophy of the distal arm “Hirayama disease”. Muscle Nerve. 2009;40(2):206–212. doi:10.1002/mus.21307 [CrossRef] PMID:19609908
- Kitagawa M, Tashiro K, Kikuchi S, Matsuura T. [Correlation between clinical features and neuroradiological findings in juvenile muscular atrophy of unilateral upper extremity (Hirayama disease)—with and without “tight dural canal in flexion”]. Rinsho Shinkeigaku. 1992;32(5):479–482. PMID:1458724
- Toma S, Shiozawa Z. Amyotrophic cervical myelopathy in adolescence. J Neurol Neurosurg Psychiatry. 1995;58(1):56–64. doi:10.1136/jnnp.58.1.56 [CrossRef] PMID:7823068
- Xu X, Han H, Gao H, et al. The increased range of cervical flexed motion detected by radiographs in Hirayama disease. Eur J Radiol. 2011;78(1):82–86. doi:10.1016/j.ejrad.2010.08.012 [CrossRef] PMID:20801593
- Hassan KM, Sahni H, Jha A. Clinical and radiological profile of Hirayama disease: a flexion myelopathy due to tight cervical dural canal amenable to collar therapy. Ann Indian Acad Neurol. 2012;15(2):106–112. doi:10.4103/0972-2327.94993 [CrossRef] PMID:22566723
- Fu Y, Qin W, Sun QL, Fan DS. [Investigation of the compliance of cervical collar therapy in 73 patients with Hirayama disease]. Zhonghua Yi Xue Za Zhi. 2016;96(43): 3485–3488. doi:10.3760/cma.j.issn.0376-2491.2016.43.009 [CrossRef] PMID:27903343
- Lyu F, Zheng C, Wang H, et al. Establishment of a clinician-led guideline on the diagnosis and treatment of Hirayama disease using a modified Delphi technique. Clin Neurophysiol. 2020;131(6):1311–1319. doi:10.1016/j.clinph.2020.02.022 [CrossRef] PMID:32311591
Patients' Clinical Information
|Age, mean±SD, y||18.9±4.0|
|Height, mean±SD, cm||173.5±5.4|
|Course of disease, mean±SD, mo||24.7±20.2|
|Affected side, left:right, No.||27:15|
|Symptoms and signs, No. of patients|
| Amyotrophy||42 (100%)|
| Muscle weakness||42 (100%)|
| Muscle tremor||9 (21.4%)|
| Cold paralysis||39 (92.9%)|
| Active lower limb reflex||7 (16.7%)|
|Performance of magnetic resonance imaging, No. of patients|
| Loss of attachment||42 (100%)|
| Increased signal intensity||8 (19.0%)|
| Snake eyes appearance||4 (9.5%)|
|Operation sections, No. of patients|
| C4-5 and C5-6||32 (76.2%)|
| C5-6 and C6-7||10 (23.8%)|
The pre-operative and post-operative MRI parameters and upper extremity function
|HD patients group||Health group$|
|Number of people||42||11|
|Age (years)||18.9 ± 4.0||36.5 ± 9.1|
|Height (cm)||173.5 ± 5.4||170.1 ± 6.6|
|Course of disease(Months)||24.7 ± 20.2||/|
|Pre-operation||Post-operation||statistical value||P value||statistical value||P value|
|DASH Score||9.19 ± 5.11||8.99 ± 5.88||Z=−0.040||0.968||/||/||/|
|Grip strength of the severe side||25.30 ± 8.84||25.57 ± 8.77||t=−0.940||0.353||42.5 ± 4.4||t=−11.686||0.000|
|Grip strength of the mild side||31.00 ± 6.80||31.02 ± 6.60||t=−0.127||0.900||42.5 ± 4.4||t=−9.690||0.000|
|MRC score of severe side (ADM)||3.63 ± 0.39||3.68 ± 0.39||t=−1.534||0.133||/||/||/|
|MRC score of severe side (APB)||3.56 ± 0.41||3.64 ± 0.39||t=−1.959||0.057||/||/||/|
|MRC score of mild side (ADM)||4.01 ± 0.66||4.02 ± 0.66||t=−1.432||0.160||/||/||/|
|MRC score of mild side (APB)||3.97 ± 0.63||3.99 ± 0.63||t=−1.776||0.083||/||/||/|
|X/Y of neutral position||0.28 ± 0.06||/||/||/||0.24 ± 0.07||t=1.761||0.084|
|X/Y of flexion position||0.50 ± 0.08||0.31 ± 0.06||t=13.807||0.000||0.29 ± 0.07||t=7.708||0.000|
|A/B of neutral position||0.43 ± 0.05||/||/||/||0.44 ± 0.07||t=−0.262||0.798|
|A/B of flexion position||0.35 ± 0.06||0.43 ± 0.07||t=−7.636||0.000||0.44 ± 0.07||t=−4.383||0.000|
The univariate analysis on the basis of whether the Dash score decrease or not
|No improvement||Improvement||t value||P value|
|Number of patients||22||20||/||/|
|Age (years)||17.91 ± 2.76||20.00 ± 4.91||1.679||0.104|
|Height (cm)||174.32 ± 5.78||173.80 ± 6.02||0.287||0.776|
|Course of disease (months)||23.14 ± 19.63||26.50 ± 21.26||0.533||0.776|
|Segments with “loss of attachment”||6.14 ± 1.32||2.90 ± 1.17||−8.387||0.000*|
|Dash Scores||6.81 ± 4.36||7.67 ± 9.02||0.398||0.693|
|Grip strength of the severe side||22.23 ± 7.99||26.47 ± 9.77||0.818||0.418|
|Grip strength of the mild side||29.96 ± 6.01||32.15 ± 7.56||1.040||0.305|
|MRC score of severe side (ADM)||3.50 ± 0.40||3.60 ± 0.43||−0.416||0.680|
|MRC score of severe side (APB)||3.58 ± 0.39||3.53 ± 0.44||−0.340||0.736|
|MRC score of mild side (ADM)||4.10 ± 0.61||3.90 ± 0.72||−1.003||0.322|
|MRC score of mild side (APB)||4.06 ± 0.58||3.87 ± 0.68||−0.999||0.324|
|A/B of neutral position||0.43 ± 0.05||0.44 ± 0.05||0.543||0.590|
|A/B of flexion position||0.33 ± 0.05||0.37 ± 0.05||2.375||0.022*|
|Difference value of X/Y||0.27 ± 0.08||0.19 ± 0.09||−3.114||0.003*|