Journal of Physiotherapy & Physical Rehabilitation Dramatic Effects of New Home Exercises for Patients with Osteoarthritis of the Hip: Pericapsular Soft Tissue and Realignment Exercises

Objective: Many patients with hip Osteoarthritis (OA), even those with Harris Hip Scores (HHS) below 60 points, have motion pain resulting from joint contracture. In these patients, pain occurs on standing and at first steps but decreases during walking. Motion pain seems to improve with exercise therapy, whereas pain during walking caused by subchondral bone exposure on the joint surface may need surgery. The goal of this study was to investigate the effectiveness of Pericapsular Soft Tissue and Realignment (PSTR) exercises for patients with OA of the hip, including those with HHS below 60 points. Methods: This retrospective observational study included 1,077 outpatients with mild to severe symptomatic and radiographic OA of the hip who were treated with patient education and supervised PSTR exercises. Approval of the local committee was obtained. Of the 1,077 patients who performed PSTR exercises, 792 were excluded from analysis. The remaining 285 participants were divided into two groups: the Unilateral OA group (no pain [HHS pain score, 44 points] in the opposite hip) and the Bilateral OA group (HHS pain score below 40 points bilaterally). HHS, pain according to numerical rating scale, range of motion, opening angle of the hip according to modified Patrick’s test, maximum strength of the hip abductors, and SF-36 were evaluated. Results: Among the 285 patients, 154 had unilateral OA and 131 had bilateral OA. Both groups showed significant HHS improvement at 3 month and 1 year follow-ups, regardless of HHS at baseline. Significant improvements in pain on the numerical rating scale, opening angle of the hip, and maximum muscle strength were noted at 3 month follow-up (P<0.0001-P<0.001). Among patients with HHS below 60 points at baseline, 38 had unilateral OA and 49 had bilateral OA. This study’s main limitation was that it was retrospective and uncontrolled.


Introduction
Most studies that have examined exercise as non-surgical treatment for Oteoarthritis (OA) of the hip have involved strength training, stretching, functional training, and aerobic fitness programs [1][2][3][4][5]. The reported inclusion criteria for exercise therapy include a Harris Hip Score (HHS) [6] between 60 and 95 points, whereas Fernandes et al. and Svege et al. [7,8] recommend Total Hip Arthroplasty (THA) for patients with an HHS below 60 points. Numerous studies have included participants with mild to moderate OA of the hip [1,[7][8][9][10][11][12]. However, patients with an HHS below 60 points may wish to postpone surgery more than patients with an HHS above 60 points. Most patients with an HHS above 60 points do not have much disturbance of Activities of Daily Living (ADL). The main reasons for avoiding or postponing surgery are work, child-rearing, or caring for one's parents. Longterm treatments to postpone or prevent surgery are needed that allow patients to exercise at home while continuing to work. Manual therapies that require hospital visits are not suitable as long-term treatment to postpone or prevent surgery.
Several studies have reported that neuromuscular exercise was effective at treating severe OA of the hip [13][14][15]. In those studies, exercise was performed preoperatively to improve postoperative outcomes of THA, not to postpone or prevent hip surgery. In a randomized clinical trial, Bennell et al. [16] implemented a multimodal physical program for patients with OA of the hip. The program consisted of manual therapy techniques (hip thrust manipulation and hip-lumbar spine mobilization), strengthening of the hip abductors and quadriceps, stretching, Range-of-Motion (ROM) exercises, functional balance and gait drills, and provision of a walking stick if appropriate. Bennell et al. [16] found that the physical therapy program conferred no additional clinical benefit over a realistic sham for 51% of patients with moderate to severe OA of the hip; they also found that the program was associated with relatively frequent but mild adverse effects [17]. No reported programs appear to have improved hip function in patients with an HHS below 60 points sufficiently to allow postpone or prevention of surgery.
Worsening OA of the hip results in anterior pelvic tilt, elevation Dramatic Effects of New Home Exercises for Patients with Osteoarthritis of the Hip: Pericapsular Soft Tissue and Realignment Exercises of the greater trochanter, and adduction contracture of the affected hip [18][19][20][21]. When the pelvis is tilted anteriorly, muscles may follow paths that differ from their anatomical paths. The load distribution may also differ from physiological load distribution. Anterior pelvic tilt and elevation of the greater trochanter cause apparent shortening of leg length on the affected side. This apparent leg-length difference [20] may result in instability while walking. Failure to correct pelvic malalignment before initiating strengthening exercises may cause imbalance between the affected and nonaffected legs and exacerbate OA of the hip. Therefore, we developed an exercise to realign the pelvis.
In our past experience, we have found that many patients with OA of the hip complained of hip pain on standing and at first steps when starting to walk, with decreased hip pain during walking. Many patients have motion pain with absent or mild walking pain. Even patients with an HHS below 60 points have this motion pain. We hypothesized that motion pain could be caused by contracture of the hip joint, whereas walking pain could be caused by contact with the subchondral bone resulting from loss of cartilage from the joint surface. Motion pain seems to improve with exercise therapy, whereas walking pain may require surgery. Walking instability caused by anterior pelvic tilt and elevation of the greater trochanter causes repetitive Pericapsular Soft Tissue (PST) injury to the joint capsule, ligaments, and inner muscles, resulting in contracture of the affected hip. Therefore, correction of pelvic malalignment before initiating strengthening exercise is necessary to decrease motion pain, in addition to correction of apparent leg-length differences. We hypothesized that decontracure exercise for PST and realignment of the pelvis (PSTR exercise) may improve hip function for patients with OA of the hip and an HHS below 60 points. Decontracture may provide a useful method to ease joint stiffness.
The current study investigated the effect of PSTR exercises on function in patients with OA of the hip; we also compared the results of these exercises in patients with unilateral versus bilateral OA of the hip. It is recognized that the results of unilateral OA of the hip is better than that of bilateral OA of the hip for one treatment (exercise, surgery etc.). So the result for one exercise is analyzed in unilateral versus bilateral OA of the hip separately in general. MRI scans routinely performed during initial examination were evaluated to rule out patients who met these exclusion criteria. In patients with SIF of the femoral head, weight was kept off the affected side and medication was administered to treat osteoporosis. If pain resulting from OA of the hip remained after the SIF healed, patients performed PSTR exercises but were excluded from this study.

Participants
The study was carried out in accordance with the Helsinki Declaration and was approved by the Fukuoka Wajiro Hospital Medical Research Ethics Committee. Authors obtained informed consent from participants when required for protection of human subjects.
The patients were observed by one hospital and any special recruitment such as advertisement in a local newspaper was not performed for this study. The inclusion was performed at Arthris Center of Fukuoka Wajiro Hospital in Fukuoka, Japan. One orthopedic surgeon examined all radiographs and three physical therapists of PSTR exercise specialist rated the patient's symptoms (HHS).

Precautions during everyday activities
Patients were instructed to avoid strain or impact on the hip joint and overloading. In principle, lifting less than 5 kg was permitted; lifting up to 10 kg was permitted for up to 15 min, if absolutely necessary. Based on this principle, patients were also instructed regarding sports, recreation, and other social activities.

Precautions during exercise
Patients were instructed to stop exercising if they experienced increased pain. Patients then attempted to exercise fewer times per day. If the pain recurred, patients were instructed to stop the program.  Exercise to correct apparent leg-length differences: Many patients with OA of the hip have apparent leg-length differences resulting from the posture they have adopted for a prolonged period because of pain. Apparent leg-length differences cause pain in the lower back and in other joints, such as the knee. This exercise increases flexibility of the erector spinae, the transversus abdominis, and internal oblique muscles.

Back-and-forth exercise:
This exercise is performed before the CKC exercise involving the hip abductors to obtain ROM that more closely approximates the physiological ROM. can be done at any time of day after the pelvic realignment exercise. However, the back-and-forth exercise is done before the CKC exercise involving the hip abductors.
Patients were instructed by a physical therapist and were supervised while exercising once every 2 weeks during the first 12 weeks. After 12 weeks, patients were instructed to perform the exercises at home as part of their daily routine.
Outcome measures: 792 patients out of 1,077ones who performed PSTR exercise were excluded and 285 participants were analyzed. Patients were divided into two groups: the Unilateral OA group (no pain [HHS pain score of 44 points] in opposite hip) and the Bilateral OA group (pain in both hips, HHS pain score below 40 points bilaterally). In the Bilateral OA group, the more painful hip joint was analyzed.
Measures of characteristics at baseline: age, sex, body mass index, duration of pain, work status, and Kellgren-Lawrence (KL) arthritis grade [23].

Measures of symptoms at baseline, 3-month, and 1-year follow-up
Month follow-up: HHS, HHS pain score, pain on a Numerical Rating Scale (NRS) [24], ROM, opening angle of the hip (opening angle of the hip according to a modified Patrick's test, (Figure 1) [25], maximum strength of hip abductors (a hand-held dynamometer was used to evaluate muscle strength), and SF-36 scores [26][27][28].
Year follow-up

HHS, HHS pain score, and SF-36 scores
No change in the KL grade from baseline to 1 year follow-up was noted. None of the participants received analgesics, including nonsteroidal anti-inflammatories, paracetamol, opioid analgesics, or glucosamine/chondroitin products.

Statistical analyses
Only treatment teams entered and collected data from electronic medical charts. Data scrubbing and statistical analyses were performed by the Clinical Research Support Center Kyushu. 1 A paired t-test was used for intra-personal change in endpoints between baseline and 3-months later or between baseline and 1 year later. Mean difference and its 95% confidence interval of endpoints were estimated by using t-distribution. All analyses were performed with Stata version 13 (Stata Corp., College Station, Texas). A value of p<0.05 was statistically significant.

Characteristics of the patients
PSTR exercises were performed by 1,077 patients with OA of the hip who visited the Arthritis Center at Fukuoka Wajiro Hospital from April 2011 to January 2014. Of these, 792 patients were excluded due to 1 The differences in the background parameters between unilateral and bilateral groups were tested by t-test for continuous variables and Chi-square test for discrete variables.    exclusion criteria and 285 fulfilled the inclusion criteria were divided into Unilateral Group (n=154) and Bilateral Group (N=131) ( Figure  2). Baseline characteristics of the patients were presented in Table 1. Exclusion criteria from beseline to 1 year follow-up were as follows. Patients fulfilled the following criteria were excluded as far as we confirmed in the electronic medical charts. In Unilateral Group 38 patients had HHS below 59 points (25.0%) and 114 patients had HHS above 60 points (75.0%). Two patients were eliminated because some of the data on their HHS scores were missing. In Bilateral Group 49 patients had HHS below 59 points (37.4%) and 82 patients had HHS above 60 points (62.6%) ( Table 1). 115 patients were excluded prior to 1-year follow-up in Unilateral Group (taking analgesics: n=12, received chiropractic treatment or other hip therapy: n=48, had not visited the clinic 2 weeks before or after 1-year followup: n=46, underwent surgery: THA; n=8, Ostoetomy; n=1). 98 patients were excluded prior to 1-year follow-up in Bilateral Group (taking analgesics; n=14, received chiropractic treatment or other hip therapy: n=37, had not visited the clinic 2 weeks before or after 1-year follow-up: n=33, underwent surgery: THA; n=11, Ostoetomy; n=3).

Changes in outcome measures except SF-36
Changes from baseline to 3-month follow-up: The results of all included patients are shown in Tables      differences in HHS were noted in patients in both the Unilateral and Bilateral OA groups (p<0.0001). Significant differences in HHS were noted between baseline and 3 month follow-up for patients with a KL grade of 1-3 but not for patients with a KL grade of 4 (Table 2c). Significant differences in HHS were noted between the two time points in patients in the Bilateral OA group who had a hip opening angle smaller than 30° according to Patrick's test, but not in the Unilateral OA group. Significant differences in HHS were noted between the two time points for patients in both the Unilateral and Bilateral OA groups who had an opening angle of the hip greater than 30° (Table  2d). Significant differences in HHS were noted in patients in both the Unilateral and Bilateral OA groups who had an HHS below 60 points at baseline (p<0.0001) (Table 2e). Table 2e shows the results of patients with an HHS above 60 points.
Changes from baseline to 1-year follow-up: Table 2a shows the results of HHS and HHS pain scores. Significant differences in HHS were noted in patients in both the Unilateral (p=0.003) and Bilateral OA groups (p<0.0001). Significant improvements in HHS were noted in both the Unilateral (p<0.0001) and Bilateral OA groups (p<0.001) among patients with an HHS below 60 points at baseline (Table 2e). Table 2e shows the results of patients with a baseline HHS above 60 points.

Changes in SF-36 scores Changes from baseline to 3 months follow-up:
The results of all investigated patients are shown in Tables 2f and 2g. In the Unilateral OA group, patients with an HHS below 60 points at baseline had significant improvements in their Mental Component summary score. In the Bilateral OA group, patients with an HHS below 60 points at baseline had significant improvements in their Mental Component and Role/Social Component summary scores (Tables 2h and 2i). Tables 2h  and 2i show the results of patients with a baseline HHS above 60 points. Tables 2f and 2g. Patients in the Unilateral OA and Bilateral OA groups with an HHS below 60 points at baseline had no significant changes in any of the component summary scores between baseline and 1-year follow-up (Tables 2h and 2i). Tables 2h and 2i show the results of patients with a baseline HHS above 60 points.

Discussion
Pelvic malalignment must be corrected and decontracture of the   affected hip must be performed to decrease motion pain, thus improving function in patients with OA of the hip and an HHS below 60 points. In this study, PSTR exercises were not indicated for patients with a KL grade of 4 and an opening angle of the hip smaller than 30° according to a modified Patrick's test. Patient education including ADL instruction is essential for PSTR exercises to be effective and to maintain adequate function and QOL over the long term [29][30][31][32].

Continuation of exercises
Once symptoms improved, patients had difficulty continuing all of the PSTR exercises daily as preventive treatment. Even if pelvic balance was restored, once patients with prolonged symptoms stopped performing the pelvic realignment exercise, the anterior pelvic tilt and elevation of the greater trochanter on the affected side tended to recur within a few days. Therefore, the following treatment program was implemented after the first 3 months of PSTR exercises.

Three months from baseline
If anterior pelvic tilt and elevation of the greater trochanter were alleviated and symptoms improved 1. Patients temporarily stopped performing the pelvic alignment exercise. If the anterior pelvic tilt and elevation of the greater trochanter were absent for 2 weeks, patients discontinued the pelvic realignment exercise. Patients continued to perform the exercise to correct apparent leg-length differences and the back-and-forth exercise. Instructors followed patients for up to 1 year from baseline.
2. If anterior pelvic tilt and elevation of the greater trochanter recurred within 2 weeks, patients continued to perform the pelvic alignment exercise for 3 more months.
If anterior pelvic tilt and elevation of the greater trochanter were alleviated but symptoms did not improve: Instructors re-evaluated the program and re-educated the patient regarding ADL. If the patient had a KL grade of 4 and an opening angle of the hip smaller than 30° according to a modified Patrick's test, instructors considered THA in consultation with an orthopedic surgeon. A SIF of the femoral head or massive tearing of the capsule, labrum, and pericapsular ligaments may be present in these cases.
If the anterior pelvic tilt and elevation of the greater trochanter remained and symptoms did not improve: Patients continued to perform the pelvic realignment exercise. If the patient had a KL grade of 4 and an opening angle of the hip smaller than 30° according to a modified Patrick's test, instructors considered THA in consultation with an orthopedic surgeon.
If the anterior pelvic tilt and elevation of the greater trochanter remained but symptoms improved: Patients continued to perform the pelvic realignment exercise. Symptoms could readily recur; surgery (osteotomy, THA, etc.) was considered when symptoms recurred if the anterior pelvic tilt and elevation of the greater trochanter remained. THA was considered for patients with a KL grade of 4 and an opening angle of the hip smaller than 30° according to a modified Patrick's test.

Six months from baseline
If anterior pelvic tilt and elevation of the greater trochanter were alleviated and symptoms improved: After performing the pelvic realignment exercise for 6 months, almost all patients had improvement of anterior pelvic tilt and greater trochanter elevation, with alleviation of symptoms. Patients stopped performing the pelvic realignment exercise but continued to perform the exercise to correct apparent leg-length differences and the back-and-forth exercises. Instructors followed patients for up to 1 year from baseline.
Patients performed PSTR exercises only as home exercises; instruction during clinic visits stopped. Patients who determined that elevation of the greater trochanter had recurred visited the clinic for reevaluation by instructors. Therefore, instruction in selfevaluation of the greater trochanter on the affected side was important (Supplementary Information). When the greater trochanter was difficult to locate, the position of the medial malleolus of the ankles was compared [20]. The medial malleolus of the ankle on the affected side is often elevated proximally in patients with hip pain. Patients cannot compare the position of the medial malleolus of their own ankles, so family members were taught this technique.
If anterior pelvic tilt and elevation of the greater trochanter were alleviated but symptoms did not improve: Instructors re-evaluated the program and reeducated the patient regarding ADL. THA was considered in patients with a KL grade of 4 and an opening angle of the hip smaller than 30° according to a modified Patrick's test.
If anterior pelvic tilt and elevation of the greater trochanter remained and symptoms did not improve: Instructors re-evaluated the program and patients continued to perform the pelvic realignment exercise for 6 more months. THA was considered in patients with a KL grade of 4 and an opening angle of the hip smaller than 30° according to a modified Patrick's test.
If the anterior pelvic tilt and elevation of the greater trochanter remained and symptoms did improve: Patients continued to perform the pelvic realignment exercise. Symptoms could readily recur; surgery (osteotomy, THA, etc.) was considered when symptoms recurred and the anterior pelvic tilt and elevation of the greater trochanter remained. THA was considered for patients with a KL grade of 4 and an opening angle of the hip smaller than 30° according to a modified Patrick's test.

One year from baseline
If symptoms improved: Patients continued to perform the backand-forth exercise.
If symptoms did not improve: Surgery (osteotomy, THA, etc.) was considered. If patients wanted to continue performing the exercises, instructors reevaluated the program and re-educated the patient regarding ADL for up to 2 years from baseline.

If symptoms improved
Patients continued to perform the back-and-forth exercise.
If symptoms did not improve: Surgery (osteotomy, THA, etc.) was considered.
The Mental Component and Role/Social Component summary scores on the SF-36 improved in both the Unilateral OA and Bilateral OA groups to near 50 points according to norm-based scoring. However, the Physical Component summary score improved only slightly, remaining below 50 points according to norm-based scoring (Tables 2f and 2g).
Low back pain may have contributed to the low Physical Component summary scores. A lumbosacral disturbance or a sacroiliac disturbance may have been involved [18,21]. The current authors plan to develop new exercises to address these issues and to establish guidelines for the timing of hip surgery based on screening involving PSTR exercises.

Motion pain as mechanism for hip pain
In many of the current patients, symptoms improved markedly as a result of PSTR exercises, even in patients with loss of joint cartilage. Partial or total loss of cartilage is not directly related to hip pain, but it can trigger that pain. Excessive strain on or overloading of the synovium, periosteum, capsule, ligaments, and labrum may cause hip pain via the mechanism shown in Figure 3 [33]. Anterior pelvic tilt and elevation of the greater trochanter may cause an apparent leg-length difference in the affected hip, resulting in instability while walking. This instability may cause repetitive PST injury, resulting in joint contracture. According to several studies, the main cause of joint contracture is skeletal muscle, followed by the joint capsule; other studies have also mentioned ligament contracture [34][35][36][37][38]. Therefore, performing PSTR exercises early to improve apparent leg-length differences and contracture of the joint capsule and ligaments is crucial to alleviate hip pain in motion and to prevent progression of OA. Ligament contracture needs to be studied in the future. In this new perspective, pericapsular soft tissue must be considered a pain generator in OA of the hip that can be addressed with new home exercises to prevent surgery [39].
This retrospective study revealed that a new home exercise program may improve hip function for patients with OA of the hip and an HHS below 60 points. We have begun a multicenter prospective single-arm study to confirm the effectiveness of this program [40]. In addition we plan studies to investigate the decontracture test to differentiate between motion pain and walking pain to determine surgical indications [39].

Strengths and Limitations
This study appears to be the first to describe an attempt to develop effective physical therapy for patients with severe OA of the hip (including those with an HHS below 60 points) who wish to postpone or avoid surgery for a prolonged period. However, there is no conclusive evidence of the effectiveness of PSTR exercises because this was a retrospective observational study with no controls.
Patients who took analgesics (even if only once) were excluded because this was a retrospective observational study. However, including patients who took analgesics with no change in type or dose would result in fewer drop-outs 1 year from baseline. The current authors plan to design a prospective comparative study that includes patients who take analgesics with no change in type or dose. Patients who received chiropractic treatment or other hip therapy at baseline were also excluded because this was a retrospective observational study. Patients had difficulty continuing to perform PSTR exercises for longer than 1 year. As mentioned in the discussion, the current authors plan to create a greatly reduced exercise program for patients whose symptoms improve (alleviation of anterior pelvic tilt and elevation of the greater trochanter) and to establish an exercise program to prevent surgery. Patients who were excluded from this study as described in the steps for patient enrollment in Figure 3 were determined based on a retrospective analysis of interviews in which a patient met two or more of the exclusion criteria. In Table 1, the Unilateral OA group at baseline consisted of 154 patients; however, two patients were eliminated because some of the data on their HHS scores were missing.

Comparison to Other Studies
To our knowledge, the current study is the first to analyze the effectiveness of exercise therapy to postpone or prevent surgery for a prolonged period in patients with severe symptoms of OA, including those with an HHS below 60 points. According to the latest Cochrane review, patients generally perform land-based exercise programs consisting of traditional muscle strength training, functional training, or an aerobic fitness program [1], except for one study in which participants were enrolled in a tai chi program [3]. The current study is the first to investigate the effectiveness of a pelvic realignment exercise before muscle exercises to improve function in patients with OA of the hip. We plan to conduct a prospective comparative study to investigate the relationship between the effect of PSTR exercises and changes in pelvic malalignment over time.

Author Contributions
K. Hayashi contributed to study design, collection, analysis, and interpretation of data, and drafting and revising of the manuscript. S. Tokunaga contributed to analysis and interpretation of data, use of statistics, and critical revising of the manuscript. K. Haruguchi, D. Nakaniwa, and Y. Tobo contributed to the explanations of PSTR exercises, the discussion, and analysis of patient characteristics at baseline (Table 1). T Shimose contributed to analysis and interpretation of data and use of statistics. All authors approved the final version of the manuscript.

Ethical Considerations and Study Approval
The study was carried out in accordance with the Helsinki Declaration and was approved by the Fukuoka Wajiro Hospital medical research ethics committee.

Earlier Publication
Some of the results of this study were prepared as an abstract and poster at the 2015 OARSI World Congress, April 30-May 3, 2015 in Seattle, WA, USA [41].