J Orthop Sci (2013) 18:932–939
DOI 10.1007/s00776-013-0443-9
ORIGINAL ARTICLE
Analysis of effectiveness of therapeutic exercise for knee
osteoarthritis and possible factors affecting outcome
Miki Kudo • Kota Watanabe • Hidenori Otsubo
Tomoaki Kamiya • Fuminari Kaneko •
Masaki Katayose • Toshihiko Yamashita
Received: 4 April 2012 / Accepted: 9 July 2013 / Published online: 2 October 2013
Ó The Japanese Orthopaedic Association 2013
Abstract
Background There are numerous reports and evidences to
suggest that exercise therapy is effective for knee osteo-
arthritis (knee OA). However, there is a lack of sufficient
research concerning the factors influencing its application
and effectiveness. The purposes of this study were to
evaluate effects of the mode of treatment delivery on the
improvement of symptoms in knee OA, and to analyze
potential risk factors affecting improvement after exercise
therapies.
Methods The 209 women applicants diagnosed with knee
OA were randomly allocated into either a group perform-
ing group exercise in a class or a group performing home
exercise. The 90 min exercise program was performed
under the guidance of physiotherapists as a group exercise
therapy. The Western Ontario and McMaster Universities
Osteoarthritis Index (WOMAC) of the subjects of both
groups before and after intervention was compared to
examine the effect of exercise therapy. In addition, body
mass index, knee range of motion (ROM), the femorotibial
angle from radiographs, OA severity from Kellgren–
M. Kudo (&)  K. Watanabe  H. Otsubo  T. Kamiya 
T. Yamashita
Department of Orthopaedic Surgery, Sapporo Medical
University School of Medicine, South-1, West-16, Chuo-ku,
Sapporo, Hokkaido 060-8543, Japan
e-mail: miky_ku@yahoo.co.jp
F. Kaneko  M. Katayose
Department of Physical Therapy, Sapporo Medical University
School of Health Science, Sapporo, Japan
F. Kaneko
National Institute of Advanced Industrial Science and
Technology, 1-3-1 Kasumigaseki, Chiyoda-ku,
Tokyo 100-8921, Japan
123
•
Lawrence grade, and meniscus abnormality and subchon-
dral bone marrow lesions from MRI findings were statis-
tically analyzed as factors that may affect exercise therapy.
Results A significantly greater improvement in WOMAC
was observed in the subjects of group exercise (81 sub-
jects) as compared with the subjects of home exercise (122
subjects). There was a significantly high proportion of
subjects with knee flexion contracture among the subjects
participating in group exercise that showed only minor
symptom improvement (p \ 0.05). In addition, exercise
therapy proved to be highly effective for subjects with
limited quadriceps muscle strength (p \ 0.05).
Conclusions When prescribing exercise therapy for knee
OA, evaluation of a subject’s ROM and muscle strength is
important in deciding whether to commence exercise
therapy and what type of exercise therapy to apply; it is
also important in predicting the effect of exercise therapy.
Introduction
Knee osteoarthritis (knee OA) is a condition often seen in
the elderly; thus, the number of affected individuals has
consequently been increasing along with the aging of
society. It is imperative to systematize the treatment of this
condition, and it is clear that conservative treatment is
particularly important from a medical-economic perspec-
tive. Four items are listed as intervention strategies for
knee OA in the ‘‘National Strategy for OA 2010’’ published
by the US Centers for Disease Control and Prevention
(CDC) and Arthritis Foundation [1]: (1) self-management
education, (2) exercise therapy, (3) injury prevention, and
(4) weight management. In particular, exercise therapy is
viewed as being important in the treatment of knee OA. In
addition, randomized prospective studies and the
Therapeutic exercise for knee OA 933

Osteoarthritis Research Society International (OARSI)
guidelines regard exercise therapy as a non-drug therapy
backed by a significant amount of evidence [2–4].
However, few reports have examined the factors influ-
encing the application and effectiveness of exercise therapy
in detail. Age, gender, obesity, leg alignment, leg strength,
joint distension, and a history of injury are cited as risk
factors for knee OA and are believed to be involved in its
onset and progression. Moreover, these factors may influ-
ence the effect of exercise therapy on knee OA.
Exercise therapy for knee OA can be divided into
muscle isolation strength training, range of motion (ROM)
exercises, and weight training. In addition, although muscle
isolation strength training is given as a single category,
many variations of muscle isolation strength training exist.
Its content can range from straight leg raises and other
bodyweight exercises to training using specialized equip-
ment. Furthermore, the muscles trained can either be lim-
ited to the quadriceps only or expanded to include muscles
around the hip as well as the trunk. Moreover, the form of
exercise intervention can be divided into exercise therapy
performed under the direct guidance of a physiotherapist
(either in a class or individually) and exercise performed by
the patient at home. Questions regarding the differences in
the effects of exercise therapy as a consequence of the
intervention format still persist.
The main focus of this study was to analyze potential
factors affecting improvement after exercise therapy for
patients with knee OA. This study tested the following
hypotheses: (1) exercise therapy is an effective treatment
for knee OA, and group-guided therapy has a greater effect
than home therapy, and (2) factors influencing the effec-
tiveness of exercise therapy exist, and they vary according
to the method of exercise therapy employed.
with motion-limiting conditions other than knee OA, and
applicants adjudged as Kellgren–Lawrence grade (KL
grade) 0 [7] from simple knee X-ray imaging of an ante-
rior–posterior view with full extension in weight-bearing.
Ultimately, 209 applicants were included in this study. The
participants were informed that data from the case would
be submitted for publication and gave their consent, and
this study was given approval by the institutional review
board (IRB) of our facility.
Subjects were randomly allocated into either a group
performing group exercise therapy conducted in a class
(group exercise) or a group performing home exercise ther-
apy (home exercise) by drawing lots. The number of par-
ticipants in the group exercise was limited to 81, because
there was a limit of a room and equipment for exercise. The
group that waited for 3 months before commencing home
exercise was designated the ‘‘control group’’ (control). There
were 81 subjects in the group exercise group, with an average
age of 63.8 years, average height of 152.6 cm, average
weight of 55.1 kg, and average body mass index (BMI) of
23.8; there were 128 subjects in the control group, with an
average age of 65.6 years, average height of 152.4 cm,
average weight of 55.3 kg, and average BMI of 23.8. This
control group subsequently performed home exercise ther-
apy, with 122 subjects ultimately completing the program
(the home exercise group). There were no statistically sig-
nificant differences in age, height, weight, or BMI between
the group exercise and control/home exercise groups
(Table 1).
The test items included simple front upright radiographs
and MRIs of the subjects’ knees in addition to their height,
weight, ROM, and muscle strength. In addition, the WO-
MAC score were used as an index of clinical symptoms.
Table 1 Characteristics of participants at baseline

Materials and methods Group exercise (n = 81) Control (n = 128)

Age, years 63.8 ± 5.9 65.6 ± 5.8

Trial participants were invited with local newspapers to Height, cm 152.6 ± 4.9 152.4 ± 5.1
participate in classes to test the effectiveness of exercise Body weight, kg 55.1 ± 7.3 55.3 ± 7.3
therapy on knee OA. The conditions required for the par- BMI, kg/m2 23.8 ± 2.9 23.8 ± 3.0
ticipants were females with knee pain aged between 55 and
KL grade, no. (%)
75 years. In total, 495 females responded during the
Grade 1 24 (29.6) 31 (24.2)
2006–2008 invitation period. These 495 applicants were
Grade 2 26 (32.1) 49 (38.3)
examined by orthopedic specialists and underwent imaging
Grade 3 27 (33.3) 42 (32.8)
before a decision was made to include them in the trial. The
Grade 4 4 (4.9) 6 (4.7)
exclusion criteria for participation in the trial included the
Mink grade, no. (%)
following: applicants with a Japanese Orthopaedic Asso-
Grade 0 8 (9.9) 29 (22.7)
ciation score [5] of\65 points or a score of 100 points, or a
Grade 1 22 (27.2) 33 (25.8)
the Western Ontario and McMaster Universities Osteoar-
Grade 2 24 (29.6) 34 (26.6)
thritis Index (WOMAC) score [6] of more than or equal to
Grade 3 27 (33.3) 32 (25.0)
96 points, applicants who had difficulty in walking, cases
where the applicant’s pain worsened acutely, applicants Values are the mean ± standard deviation unless otherwise indicated

123
934
This index (raw score out of 96) was normalized by mul-
tiplying it by 100/96, and the improvement in the nor-
malized index—pre-intervention normalized WOMAC
subtracted by post-intervention normalized WOMAC was
used as an indicator of clinical improvement. This calcu-
lated value was defined as W0 in this study.
Knee OA severity was classified using KL grade. WO-
MAC was assessed both prior to commencement of the
exercise program and 3 months after commencement. The
group exercise therapy was performed under the direct
guidance of physiotherapists. The 90 min program included
stretching (warm up), followed by riding an exercise bike,
and muscle strength and stabilization exercises (focusing on
the knee joint and incorporating pelvic and trunk areas),
followed by more stretching (cool down). Open kinetic chain
exercises and a combination of isotonic contraction and
isometric contraction of the muscle were firstly performed to
the muscles around the knee, the trunk, the hip, and the ankle
as the muscle strength exercises. The time and the repetition
number of the exercises were gradually increased, and then
closed kinetic chain exercises were added to the program
including a squat and a calf raise. The physical therapists
were paying attention to pain and symptoms of the subjects to
prevent those deterioration during the exercises. Riding an
exercise bike was started from the exercise intensity for
20 min to be around 55–65 % of the predicted maximum
heart rate, and carried out for 40 min in the end. The balance
ball and balance cushion were used in stabilization exercise
to improve the balance in the standing posture and trunk
muscle strength. Subjects participated in this program twice
a week for 3 months. A workshop was held for the home
exercise group, both prior to intervention and during inter-
vention, in which guidance was given concerning the content
of the program. The home exercise group was provided with
self-check sheets having attached photographs explaining
each of the stretching, muscle strengthening, and stabiliza-
tion exercises (similar to those performed by the group
exercise group); this group was recommended to exercise at
least twice a week for 3 months. The subjects did not have
any other conservative treatments such as medication during
the participating period of the program.
Age, BMI, ROM, quadriceps muscle strength, leg
alignment, severity of knee OA (from KL grade), and
findings of meniscus abnormality and subchondral bone
marrow lesions from MRI were analyzed as factors that
may affect exercise therapy. Subjects were divided into two
groups according to the median age (65 years), and based
on their degree of obesity: a group with a BMI of\25 and a
group with a BMI of more than or equal to 25. In terms of
ROM, subjects whose knee extension was limited by 5° or
more were defined as having flexion contracture. Quadri-
ceps muscle strength was measured using a muscle strength
measurement instrument (Kenkonice, Hulia Co. Ltd.,
123
M. Kudo et al.
Tokyo). Maximum voluntary isometric force of the quad-
riceps muscle was measured while holding maximum
output for 3 s at 60° knee flexion in sitting position. This
measurement was repeated twice and the higher value was
chosen as a representative value. Then, the weight ratio
calculated from the values obtained, and the subjects
divided into two groups based on the median value.
The femorotibial angle (FTA) was measured from simple
front upright knee radiographs as an index of leg alignment,
and subjects were subsequently divided into a group with an
FTA of\178° and a group with an FTA of more than or equal
to 178° which was the median in the subjects. Subjects were
divided into a KL grade 1 and 2 group and KL grade 3 and 4
group based on the severity of knee OA; they were divided
into a Mink grade 0–2 and grade 3 group on the basis of
meniscal MRI findings [8]. Mink grade 3 is considered as
meniscal tear [8]. The degree of extension of an area of
abnormal low intensity on the femoral condyle was classified
into four types by sagittal proton-density MRI [9]: normal,
no abnormal low intensity area; spot, an area of abnormal
low intensity is shown as a single spot on the subchondral
bone; moderate, the abnormal low intensity area is extended
from the subchondral bone to the proximal bone marrow, and
extension is less than one-third of the femoral condyle; and
large, the extension of the abnormal low intensity area is
more than one-third of the femoral condyle. MRI imaging
was performed using a 1.5-T system (General Electric
Medical Systems, USA). Sagittal and coronal images were
acquired with proton density-weighted images (TR/
TE = 1900/31.7; flip angle 90°) and T2-weighted images
(TR/TE = 3800/102; flip angle 90°). Each image was
obtained using a 160 mm, 320 9 256 matrix with 3 mm
slices. All radiographs and MR images were analyzed by an
experienced orthopaedic surgeon.
Firstly, W0 was compared between pre- and post-inter-
vention in the groups of control, group exercise, and home
exercise. Statistical analysis was performed using paired
t test with statistical significance set at p \ 0.05 level.
Then, the groups of home exercise and group exercise were
divided into more effective group and less effective group
respectively according to the value of W0 , and relationship
was analyzed between the effectiveness of therapeutic
exercise and the various factors mentioned previously.
Unpaired t test was used for statistical analysis of W0 and
the factors of BMI, quadriceps strength, and FTA, and
Mann–Whitney U test was used for flexion contracture, KL
grade, Mink grade, and subchondral bone marrow lesion
analysis with statistical significance set at p \ 0.05 level.
Finally, subjects of group exercise were divided into two
groups according to the factors described above, and then
W0 was compared between the two groups. Statistical
analysis was performed using Mann–Whitney U test with
statistical significance set at p \ 0.05 level.
Therapeutic exercise for knee OA
Results
Eighty-one subjects in the group exercise group and 122 in
the home exercise group completed their exercise programs
ultimately. No significant change in the WOMAC score of
the control group was observed before and after the
3 month waiting period according to paired t test. How-
ever, W0 improved by an average of 10.2 points in the
group exercise (SD, 10.3)—a statistically significant
improvement from the beginning of exercise and compared
with the control (p \ 0.05). Although W0 improved by an
average of 3.2 points in the home exercise (SD, 8.7), it was
not recognized as a statistically significant improvement
from the beginning of exercise according to paired t test.
The group exercise showed significantly higher W0
improvement than the home exercise according to unpaired
t test (p \ 0.05). Figure 1 shows changes of normalized
WOMAC before and after intervention in the home exer-
cise and the group exercise.
Next, the factors obtained from physical, radiographical,
and MRI findings were compared between the group whose
symptoms improved with exercise (more effective group)
and the group where there was minimal improvement of
symptoms (less effective group) (Table 2). Subjects whose
W0 improved by 6 points or more were allocated to the
more effective group, and the subjects whose W0 improved
by\6 points were allocated to the less effective group. The
results of this comparison showed that there was a signif-
icantly high proportion of subjects with flexion contracture
of the knee in the less effective group (p \ 0.05) and
Fig. 1 Improvements of normalized WOMAC pre- and post-exercise
in the home exercise and the group exercise. Normalized WO-
MAC = (raw score out of 96) 9 100/96 Statistical analysis was
performed using paired t test (* p \ 0.05)
935
quadriceps strength was significantly lower in the more
effective group (p \ 0.05).
Figure 2 shows whether each factor affected improve-
ment in WOMAC scores in group exercise. The WOMAC
scores of the group of subjects with quadriceps strength
lower than the median value improved significantly
(p \ 0.05). Besides this, no other factor was observed to
significantly affect improvement in WOMAC scores in
group exercise.
Discussion
The first item examined in the study was a comparison of
the effects of group exercise and home exercise. The result
indicated that group exercise was superior to home exercise
for the treatment of knee OA. There were several factors
which might influence effects of exercise therapies on
patients with knee OA. In the group exercise, the subjects
performed their exercises with other subjects under the
guidance of therapists in a class. They could use exercise
machines such as an exercise bike. These factors might
improve not only quality of the exercise but also motiva-
tion and compliance of the subjects. We thought that these
differences between two groups might lead to superior
results of the group exercise therapy in this study. Roddy
et al. [10] performed a systematic literature review and
provided 10 evidence-based recommendations for the role
of exercise in the management of OA of the hip or knee,
known as ‘‘the MOVE consensus’’. One of the 10 recom-
mendations was ‘‘group exercise and home exercise are
equally effective, and patient preference should be con-
sidered’’. Another meta-analysis [11] showed that the mode
of treatment delivery (individual treatments, exercise
classes, and home programs) was not significantly associ-
ated with the magnitude of treatment benefit. They ana-
lyzed supervision occasions and found that programs
providing fewer than 12 direct supervision occasions
demonstrated only small mean benefits for pain reduction
and physical function. In this study, direct supervision was
provided on 24 occasions over 3 months in the group
exercise program. However, direct supervision was only
provided once during the 3 month home exercise therapy
period, and there may have been limited opportunities to
check and re-teach the exercise method. In any event, it
was clear that the group exercise program adopted in this
study was effective.
Of the factors possibly influencing the effectiveness of
exercise therapy on knee OA, this study suggested that
flexion contracture and quadriceps strength affected group
exercise. Because a significant improvement in the WO-
MAC score was not observed after the home exercise in
this study, we analyzed these factors in the group exercise
123
936 M. Kudo et al.

Table 2 Relationship between the effectiveness of therapeutic exercise and various factors
Group exercise (n = 81) Home exercise (n = 122)
Less effective More effective p value Less effective More effective p value
(n = 35) (n = 46) (n = 81) (n = 41)

W0 1.4 ± 2.8 16.9 ± 8.7 \0.001 -1.4 ± 6.1 12.1 ± 5.8 \0.001
BMI, kg/m2 23.5 ± 2.7 24.0 ± 2.9 0.473 23.7 ± 2.8 24.0 ± 3.5 0.554
Flexion contracture, no. (%) 19 (54.3) 20 (43.5) 0.045 25 (30.9) 17 (41.5) 0.268
Quadriceps strength, Nm/kg 1.53 ± 0.51 1.29 ± 0.49 0.044 1.38 ± 0.52 1.38 ± 0.48 0.978
FTA, degrees 177.8 ± 4.0 178.1 ± 3.6 0.775 178.2 ± 3.6 178.1 ± 3.9 0.825
KL grade, no. (%) 0.172 0.458
Grade 1,2 19 (54.3) 31 (67.4) 53 (65.4) 24 (58.5)
Grade 3,4 16 (45.7) 15 (32.6) 28 (34.6) 17 (41.5)
Mink grade, no. (%) 0.455 0.856
Grade 0–2 20 (57.1) 34 (73.9) 58 (71.6) 30 (73.2)
Grade 3 15 (42.9) 12 (26.1) 23 (28.4) 11 (26.8)
Subchondral bone marrow lesion, 0.271 0.346
no. (%)
Normal 12 (34.3) 18 (39.1) 33 (40.7) 13
Spot 12 (34.3) 17 (37.0) 34 (42.0) 19
Moderate 11 (31.4) 10 (21.7) 12 (14.8) 9
Large 0 (0) 1 (2.2) 2 (2.5) 0
0
More effective: Group where W improved by 6 points or more
Less effective: Group where W0 improved by less than 6 points
W0 : pre-intervention normalized WOMAC subtracted by post-intervention normalized WOMAC
Normalized WOMAC = (raw score out of 96) 9 100/96
Unpaired t test was used for statistical analysis of W0 and the factors of BMI, quadriceps strength, and FTA, and Mann–Whitney U test was used
for flexion contracture, KL grade, Mink grade, and subchondral bone marrow lesion analysis

group. Subjects with low quadriceps strength experienced
superior symptomatic relief as a result of group exercise. In
a study on the relationship between muscle strength prior to
the commencement of an exercise program and symp-
tomatic improvement, Irie et al. [12] evaluated the efficacy
of quadriceps setting exercise programs for relieving pain
in knee OA and investigated background factors influenc-
ing efficacy. Fifty-three knees of 43 patients were evalu-
ated until 12 months after initial examination. Pain was
relieved at a mean of 2.3 months in half of the patients. No
association was identified between pain relief and age,
BMI, radiographic OA grading, or initial muscle strength.
Only patients \70 years of age who had initially displayed
quadriceps muscle strength below median value demon-
strated a high incidence of pain relief. Torii [13] reported
the results of 9 weeks of intervention to increase muscle
strength around the knee and hip joints for middle- or
advanced-aged women with knee OA, and analyzed the
characteristics of the group whose pain improved. In the
groups that showed improvement, body weight was hea-
vier, and the values of muscle strength per body weight
were lower at baseline. Thus, it is thought that patients with
low muscle strength prior to the commencement of an
exercise program can more easily enhance their muscular
strength as a result of muscle strength exercises; therefore,
it is easier for them to obtain symptomatic relief.
A few reports examined knee ROM as a possible pre-
dictor of treatment responsiveness of exercise therapy [14,
15]. This factor did not reveal trends in treatment effec-
tiveness in those studies. However, flexion contracture is
one of major symptoms of patients with knee OA, espe-
cially in advanced stages. Quadriceps muscle power or
contractile function of this muscle should be deteriorated in
near extended knee position in patients with flexion con-
tracture. From the results of this study, we believe that
flexion contracture and quadriceps strength are the possible
factors affecting improvement. We also think that further
study is needed to investigate effects of a specially modi-
fied exercise program to reduce knee flexion contracture on
patients with knee OA.
Although several other factors such as those examined
in this study may influence the effect of exercise therapy on
knee OA, a consensus has not yet been reached. For
example, some reports indicate that symptomatic relief is
less in cases of severe knee OA; some dismiss any corre-
lation between symptomatic relief and OA severity.

123
Therapeutic exercise for knee OA
Fig. 2 Effects of various factors on the efficacy of therapeutic
exercise in subjects of group exercise. a Age, b BMI, c flexion
contracture, d quadriceps strength, e FTA, f KL grade g Mink grade,
h subchondral bone marrow lesion W0 : pre-intervention normalized
According to the aforementioned ‘‘the MOVE consensus’’
by Roddy et al. [10], one of the 10 propositions was ‘‘the
effectiveness of exercise is independent of the presence or
severity of radiographic findings’’. They did not find any
direct evidence to support this proposition except one
937
WOMAC subtracted by post-intervention normalized WOMAC.
Statistical analysis was performed using Mann–Whitney U test
(* p \ 0.05)
randomized controlled trial (RCT) performed by Fransen
et al. [14]. This study stratified subjects according to
medial knee joint space width and demonstrated signifi-
cantly greater improvement in pain and function in subjects
with less severe loss of medial joint space. In this study,
123
938
126 patients were randomized into one of 3 allocation
arms: individual treatments, small group format program,
and waiting list control. Assessments included both self-
reporting measures (WOMAC, SF-36) and objective mea-
sures of physical performance (gait analysis and muscle
strength). Both physical therapy treatment allocations
resulted in significant improvements in pain, physical
function, and health-related quality of life above the con-
trol group.
The alignment of lower extremities assessed by FTA
was not correlated with the effectiveness of exercise ther-
apy in the current study. In a report that examined the
connection between muscle exercise and knee alignment,
Lim et al. [15] performed a RCT to examine whether the
effects of 12 weeks of quadriceps strengthening on the
knee adduction moment, pain, and function in people with
medial knee OA differed in those with and without varus
malalignment. The benefits of quadriceps strengthening on
pain were more evident in those with more neutral align-
ment. They concluded that knee alignment represented a
local mechanical factor that could mediate the symptom-
atic outcome from exercise interventions in knee OA. The
knee OA severity of the participants in our study was
between KL grade 1 and 3. Few participants showed severe
varus malalignment. We thought that this factor might
influence our result of FTA analysis.
The present study also examined whether abnormal MRI
findings influenced the effects of exercise therapy. There
were few reports investigating MRI findings as a possible
factor of exercise treatment responsiveness. Several studies
investigated the relationship between abnormal MRI find-
ings and pain in knee OA. Link et al. [16] evaluated
whether knee symptoms assessed by WOMAC correlated
with abnormal MRI findings including cartilage lesions,
bone marrow edema, and ligamentous and meniscal lesions
in 50 patients. Significant differences between WOMAC
scores were found only for the grades of cartilage lesions.
While Felson et al. [17] reported that bone marrow lesions
on MRI were strongly associated with the presence of pain
in knee OA. No correlation was observed between the
presence of femoral subchondral bone or meniscal abnor-
malities in MRI findings and the improvement of symp-
toms from the results of the current study. Further study of
the relationship between knee pain from OA and abnor-
malities in MRI findings is considered necessary.
The present study had limitations in that the participants
actually applied to participate in the exercise therapy for
knee OA, and therefore, may have had strong motivation to
exercise. Therefore, these results may not necessarily be
reflected in subjects in routine medical practice. However,
similar to previous studies, this study clarified that group
exercise under the guidance of a physical therapist familiar
with exercise therapy is extremely effective. Therapeutic
123
M. Kudo et al.
exercise should be encouraged for patients with quadriceps
muscle weakness. Moreover, exercise may be less effective
in providing symptomatic relief in cases where flexion
contracture is observed. Programs to improve joint con-
tracture warrant consideration in such cases. In summary,
flexion contracture of the knee and muscle strength should
be checked in patients with knee OA, and therapeutic
exercise is recommended as a treatment option.
Acknowledgments We sincerely thank Drs.Tsuyoshi Minowa,
Kenji Tateda, and Ima Kosukegawa for their help.
Conflict of interest The authors declare that they have no conflict
of interest.
References
1. Lubar D, White PH, Callahan LF, Chang RW, Helmick CG,
Lappin DR, Melnick A, Moskowitz RW, Odom E, Sacks J, Toal
SB, Waterman MB. A national public health agenda for osteo-
arthritis 2010. Sem Arth Rheum. 2010;39:323–6.
2. Richmond J, Hunter D, Irrgang J, Jones MH, Snyder-Mackler L,
Van Durme D, Rubin C, Matzkin EG, Marx RG, Levy BA,
Watters WC 3rd, Goldberg MJ, Keith M, Haralson RH 3rd,
Turkelson CM, Wies JL, Anderson S, Boyer K, Sluka P, St Andre
J, McGowan R, et al. American academy of orthopaedic surgeons
clinical practice guideline on the treatment of osteoarthritis (OA)
of the knee. J B Jt Surg Am. 2010;92:990–3.
3. Jordan KM, Arden NK, Doherty M, Bannwarth B, Bijlsma JW,
Dieppe P, Gunther K, Hauselmann H, Herrero-Beaumont G,
Kaklamanis P, Lohmander S, Leeb B, Lequesne M, Mazieres B,
Martin-Mola E, Pavelka K, Pendleton A, Punzi L, Serni U,
Swoboda B, Verbruggen G, Zimmerman-Gorska I, Dougados M.
EULAR recommendations 2003: an evidence based approach to
the management of knee osteoarthritis: report of a task force of
the standing committee for international clinical studies including
therapeutic trials (ESCISIT). Ann Rheum Dis. 2003;62:1145–55.
4. Zhang W, Moskowitz RW, Nuki G, Abramson S, Altman RD,
Arden N, Bierma-Zeinstra S, Brandt KD, Croft P, Doherty M,
Dougados M, Hochberg M, Hunter DJ, Kwoh K, Lohmander LS,
Tugwell P. OARSI recommendations for the management of hip
and knee osteoarthritis, part I: critical appraisal of existing
treatment guidelines and systematic review of current research
evidence. Osteoarth Cartil. 2007;15:981–1000.
5. The Japanese Orthopaedic Association Japanese Knee Society.
Assessment criteria for knee diseases and treatments. Tokyo:
Kanehara; 1994.
6. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt
LW. Validation study of WOMAC: a health status instrument for
measuring clinically important patient relevant outcomes to
antirheumatic drug therapy in patients with osteoarthritis of the
hip or knee. J Rheumatol. 1988;15:1833–40.
7. Kellgren JH, Lawrence JS. Radiological assessment of osteoar-
throsis. Ann Rheum Dis. 1957;16:494–502.
8. Stoller DW, Martin C, Crues JV 3rd, Kaplan L, Mink JH. Men-
iscal tears-pathologic correlation with MRI imaging. Radiology.
1987;163:731–5.
9. Nakamura E, Torisu T, Tomari K, Nakamura T, Asao T, Takagi
K, Masumi S. Clinical significance of subchondral cortical signal
intensities in osteoarthritis of knee joints. Nichiji Ishi (JJMRM).
1996;16:243–51 (in Japanese).
Therapeutic exercise for knee OA
10. Roddy E, Zhang W, Doherty M, Arden NK, Barlow J, Birrell F,
Carr A, Chakravarty K, Dickson J, Hay E, hosie G, Hurley,
Jordan KM, McCarthy C, McMurdo M, Mockett S, O’Reilly S,
Peat G, Pendleton A, Richards S, et al. Evidence-based recom-
mendations for the role of exercise in the management of
osteoarthritis of the hip or knee––the MOVE consensus. Rheu-
matology. 2005;44:67–73.
11. Fransen M, McConnell S. Land-based exercise for osteoarthritis
of the knee: a metaanalysis of randomized controlled trials.
J Rheumatol. 2009;36:1109–17.
12. Irie K, Watanabe K, Tomatsu T. Effects and influencing factors
of quadriceps setting program on patients with knee osteoarthri-
tis. Undou Butsuri Ryohou (J Physical Medicine). 2002;13:
249–54 (in Japanese).
13. Torii S. The characteristics of the improved cases of mild or
moderate knee osteoarthritis by intervention to increase muscle
strength about the knee and hip joint. Heal Sci. 2003;19:122–6
(in Japanese).
939
14. Fransen M, Crosbie J, Edmonds J. Physical therapy is effective
for patients with osteoarthritis of the knee: a randomized con-
trolled clinical trial. J Rheumatol. 2001;28:156–64.
15. Lim B-W, Hinman RS, Wrigley TV, Sharma L, Bennell KL.
Does knee malalignment mediate the effects of quadriceps
strengthening on knee adduction moment, pain, and function on
medial knee osteoarthritis? A randomized controlled trial.
Arthritis Rheum. 2008;59:943–51.
16. Link TM, Steinbach LS, Ghosh S, Ries M, Lu Y, Lane N,
Majumdar S. Osteoarthritis: MR imaging findings in different
stages of disease and correlation with clinical findings.
Radiology. 2003;226:373–81.
17. Felson DT, McLaughlin S, Goggins J, LaValley MP, Gale ME,
Totterman S, Li W, Hill C, Gale D. Bone marrow edema and its
relation to progression of knee osteoarthritis. Ann Intern Med.
2003;139:330–6.
123
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.