Objective. To determine whether Tai Chi or hydrotherapy classes for individuals with chronic symptomatic hip or knee osteoarthritis (OA) result in measurable clinical benefits.
Methods. A randomized controlled trial was conducted among 152 older persons with chronic symptomatic hip or knee OA. Participants were randomly allocated for 12 weeks to hydrotherapy classes (n _ 55), Tai Chi classes (n _ 56), or awaiting list control group (n _ 41). Outcomes were assessed 12 and 24 weeks after randomization and included pain and physical function (Western Ontario and McMaster Universities Osteoarthritis Index), general health status (Medical Outcomes Study Short Form 12 Health Survey [SF-12], version 2), psychological well-being, and physical performance(Up and Go test, 50-foot walk time, timed stair climb).
Results. At 12 weeks, compared with controls, participants allocated to hydrotherapy classes demonstrated mean improvements (95% confidence interval) of 6.5 (0.4, 12.7) and 10.5 (3.6, 14.5) for pain and physical function scores (range0–100), respectively, whereas participants allocated to Tai Chi classes demonstrated improvements of 5.2 (_0.8, 11.1) and9.7 (2.8, 16.7), respectively. Both class allocations achieved significant improvements in the SF-12 physical component summary score, but only allocation to hydrotherapy achieved significant improvements in the physical performance measures. All significant improvements were sustained at 24 weeks. In this almost exclusively white sample, class attendance was higher for hydrotherapy, with 81% attending at least half of the available 24 classes, compared with 61%for Tai Chi.
Conclusion. Access to either hydrotherapy or Tai Chi classes can provide large and sustained improvements in physical function for many older, sedentary individuals with chronic hip or knee OA.
Physical Activity for Osteoarthritis Management:
A Randomized Controlled Clinical Trial
Evaluating Hydrotherapy or Tai Chi Classes
MARLENE FRANSEN,1LILLIAS NAIRN,1JULIE WINSTANLEY,2PAUL LAM,3ANDJOHN EDMONDS4
Objective.To determine whether Tai Chi or hydrotherapy classes for individuals with chronic symptomatic hip or knee
osteoarthritis (OA) result in measurable clinical benets.
Methods.A randomized controlled trial was conducted among 152 older persons with chronic symptomatic hip or knee
OA. Participants were randomly allocated for 12 weeks to hydrotherapy classes (n55), Tai Chi classes (n56), or a
waiting list control group (n41). Outcomes were assessed 12 and 24 weeks after randomization and included pain and
physical function (Western Ontario and McMaster Universities Osteoarthritis Index), general health status (Medical
Outcomes Study Short Form 12 Health Survey SF-12, version 2), psychological well-being, and physical performance
(Up and Go test, 50-foot walk time, timed stair climb).
Results.At 12 weeks, compared with controls, participants allocated to hydrotherapy classes demonstrated mean
improvements (95% condence interval) of 6.5 (0.4, 12.7) and 10.5 (3.6, 14.5) for pain and physical function scores (range
0 100), respectively, whereas participants allocated to Tai Chi classes demonstrated improvements of 5.2 (0.8, 11.1) and
9.7 (2.8, 16.7), respectively. Both class allocations achieved signicant improvements in the SF-12 physical component
summary score, but only allocation to hydrotherapy achieved signicant improvements in the physical performance
measures. All signicant improvements were sustained at 24 weeks. In this almost exclusively white sample, class
attendance was higher for hydrotherapy, with 81% attending at least half of the available 24 classes, compared with 61%
for Tai Chi.
Conclusion.Access to either hydrotherapy or Tai Chi classes can provide large and sustained improvements in physical
function for many older, sedentary individuals with chronic hip or knee OA.
KEY WORDS.Osteoarthritis; Hydrotherapy; Tai Chi; Exercise.
Regular moderate physical activity provides a wide range
of health benets (1). Unfortunately, a large proportion of
individuals with osteoarthritis (OA) involving the hips or
knees are sedentary (2). Graded exercise programs are ef-
fective interventions for patients with knee OA (3), with
both strength training (4) and aerobic exercise (5) demon-
strating signicant improvements in pain, physical func-
tion, and general health status. However, ongoing adher-
ence to exercise programs is poor (6).Hydrotherapy is frequently recommended because the
weight-relieving properties of water allow for easier joint
movement. However, a systematic review of randomized
studies conducted up to 2000 identied only 4 studies
involving patients with OA and concluded that the evi-
dence for effectiveness was weak due to the poor quality of
these studies (7). A recent randomized study of 105 pa-
tients with OA found no signicant improvements in joint
pain or physical function (8). However, almost 50% of
ClinicalTrials.gov identier: NCT00123994
Supported by a National Arthritis and Musculoskeletal
Conditions Improvements grant, the University of New
South Wales, the St. George Division of General Practice,
and the Central Sydney Division of General Practice.
1Marlene Fransen, MPH, PhD, Lillias Nairn, DipPhysio:
The George Institute for International Health, University of
Sydney, Sydney, Australia;
2Julie Winstanley, PhD: Univer-
sity of the Sunshine Coast, Queensland, Australia;3Paul
Lam, MBBS: University of New South Wales, Sydney, Aus-tralia;
4John Edmonds, MBBS: The St. George Hospital,
Kogarah, New South Wales, Australia.
Dr. Lam received royalties from the sale of Tai Chi for
Arthritis videoDVD and a book titledOvercoming Arthritis.
Address correspondence to Marlene Fransen, MPH, PhD,
PO Box M201 Missenden Road, Camperdown, New South
Wales, Australia 2050. E-mail: mfransen@thegeorgeinstitute.
Submitted for publication June 4, 2006; accepted in re-
vised form August 25, 2006.
Arthritis & Rheumatism (Arthritis Care & Research)
Vol. 57, No. 3, April 15, 2007, pp 407 414
2007, American College of Rheumatology
these study participants were on the orthopedic surgery
waiting list. Previous research has clearly demonstrated
limited responsiveness to graded exercise among patients
with severe structural disease (9).
In recent years, Tai Chi has been gaining popularity. Tai
Chi consists of slow, continuous movements that incorpo-
rate elements of strengthening, balance, postural align-
ment, relaxation, and concentration. Although many
claims have been made about the health benets afforded
by Tai Chi, there are few randomized studies specically
directed at individuals with chronic OA. Two small stud-
ies reported some improvements in quality of life indica-
tors and functioning; however, both studies had method-
ologic problems likely to inate estimates of treatment
The goal of the present study was to determine whether
hydrotherapy or Tai Chi classes for individuals with
chronic symptomatic OA of the hips or knees are accept-
able physical activity options that can provide measurable
improvements in joint pain and physical function. If hy-
drotherapy or Tai Chi classes were shown to be as effective
as traditional land-based exercise in reducing pain and
improving physical function, such classes could prove to
be effective, inexpensive, attractive, and accessible treat-
ment and regular physical activity options for older per-
sons with OA.
PARTICIPANTS AND METHODS
A randomized controlled clinical trial with blinded out-
comes assessment was conducted among community-
dwelling older persons with symptomatic OA of the hips
or knees. This study was conducted in compliance with
the Helsinki Declaration and was approved by the South
Eastern Sydney Area Health Service Human Research Eth-
ics Committee. Written informed consent was obtained
from all participants prior to randomization. The study
was registered with the National Institutes of Health.
Participants.Participants were recruited via advertise-
ment in local newspapers, through presentations at local
social clubs for older persons, and through referral from
local general practitioners and rheumatologists. The inclu-
sion criteria were age 59 85 years, a diagnosis of OA
involving the hip or knee as per American College of
Rheumatology criteria (12,13), and current and chronic
(1 year) hip or knee pain. Exclusion criteria were current
participation in recreational physical activity more than
twice per week; inability to walk indoors without a walk-
ing aide; unstable cardiac conditions or severe pulmonary
disease; incontinence, fear of water, or uncontrolled epi-
lepsy; low back pain referred to the lower limbs; joint
replacement surgery in the previous year; arthroscopic
surgery or intraarticular injections within previous 3 months;
and current participation in Tai Chi or hydrotherapy.
Randomization.A computerized randomization sched-
ule, in blocks of 30, was generated at an offsite location,
from which participants were informed of their allocation
by telephone after completing their baseline assessment.Participants were randomized to 1 of 3 groups: hydrother-
apy classes, Tai Chi classes, or a waiting list control group
(12 weeks) prior to randomization to hydrotherapy or Tai
Chi. Class size was restricted to a maximum of 15, and
classes were conducted at St George Public Hospital be-
tween January 2004 and October 2005. Participants were
required to make a single donation of $35 to assist with
Interventions.Participants were required to attend
classes for 1 hour, twice a week for 12 weeks. While class
attendance was recorded, home practice was not moni-
Four different registered physiotherapists conducted hy-
drotherapy classes over the study period. The hydrother-
apy program (Appendix A) was designed by the senior
Four different Tai Chi instructors trained in a specially
designed Tai Chi program (Tai Chi for Arthritisvideo, Paul
Lam) conducted classes over the study period. This pro-
gram is a modication of 24 forms from the Sun style of Tai
Chi and includes a preliminary 10-minute warm-up ses-
sion. Participants were able to purchase, if they desired, a
Tai Chi video to assist with home practice.
Outcomes.The study project manager, who remained
blind to participants group allocation, carried out all out-
comes assessments. Assessments were conducted at base-
line (pretreatment), 12 weeks (posttreatment), and 24
weeks (followup). Controls were assessed following the
same schedule after completion of the 12-week waiting list
period. Baseline participant characteristics included a val-
idated evaluation of comorbidity (14). Participants were
asked to identify a signal (most painful) hip or knee joint
for all further assessments.
The primary outcome measures were pain and physical
function measured by the Western Ontario and McMaster
Universities Osteoarthritis Index (WOMAC; Likert ver-
sion) (15). Scores were standardized to a 0 100 range, with
higher scores indicating greater pain or physical disability.
Secondary outcomes included general health status,
psychological well-being, patients global assessment of
treatment effectiveness for the signal joint, patients global
assessment of current status of the signal joint, and phys-
ical performance measures. General health status was as-
sessed with the Medical Outcomes Study Short Form 12
Health Survey (SF-12), version 2 (16). From the 8 domains
of general health status evaluated by the SF-12, 2 scores
were computed: physical component summary (PCS) and
mental component summary (MCS) (16). The PCS and the
MCS are transformed to have a meanSD score of 5010
in the general US population. Psychological well-being
was measured with the Depression, Anxiety and Stress
Scale (DASS21) (17). Each of the 3 subscales has a score
range of 0 42, with higher scores indicating poorer psy-
chological well-being. Patients global assessment of treat-
ment effectiveness for the signal joint is a 5-level categor-
ical scale ranging from much better to much worse.
Patients global assessment of current status of the signal
joint is a 5-level scale ranging from excellent to poor
408Fransen et al
(18). Physical performance measures included the timed
50-foot walk test (19), stair climb test, and Up and Go
Statistical analysis.The primary statistical analysis
was per intent-to-treat with a priori planned comparisons
of 1) hydrotherapy classes against controls and 2) Tai Chi
classes against controls. Participants lost to followup were
assigned scores as per their last assessment but adjusted
for any systematic changes as ascertained in the control
group. An overall signicance level of 2-sidedP0.05
was set and adjustments were made for multiple compar-
isons relating to secondary outcomes using the Bonferroni
correction. For all change scores, positive scores were
improvements. Unadjusted and adjusted analyses were
performed, with unadjusted analyses consisting primarily
of pairedt-tests and chi-square analysis. The standardized
response mean (SRM; mean change divided by the stan-
dard deviation of change) was calculated for those out-
comes demonstrating signicant improvement. The SRM
95% condence intervals (95% CIs) were estimated as-
suming a normal distribution of the SRM (21).
For improved clinical relevance, analyses were con-
ducted according to the Outcome Measures in Rheumatol-
ogy Clinical Trials (OMERACT)Osteoarthritis Research
Society International (OARSI) responder criteria (scenario
D) for OA (22). As such, responders were dened as par-
ticipants achieving50% improvement in pain or physi-
cal function scores (and an absolute change of20) or20
improvement in both pain and physical function scores
(and an absolute change of10%) on the WOMAC sub-
Sample size.Sample size estimates were based on the
primary outcome measure, the WOMAC pain and physical
function scores. Evidence from the literature suggests that
12 mm (0 100-mm scale) represents a clinically signi-
cant difference, with an SD of 22 mm (23). To provide the
study 80% power to detect such an effect, and allowing for
up to a 25% dropout rate, a total sample size of 150
participants was required (24). The study would also have
80% power to demonstrate a 2.5 increased risk of being an
OMERACTOARSI responder (22).
Of the 637 individuals initially screened by telephone,
more than one-third (n254 40%) did not meet the
inclusion criteria and more than one-third (n231 36%)
were not interested or not available (Figure 1). The most
common reasons for not meeting the inclusion criteria
were health or surgical reasons (n68 27%), asymptom-
atic for the past year (n49 19%), pain referred from the
lumbar spine (n43 17%), too young or too old (n33
13%), too active (n33 13%), having other physical
treatments or corticosteroid injection in the past 3 months
(n21 8%), and unable to participate in hydrotherapy
(n7 3%). The reasons provided by the other 231
potential participants were unavailable to attend classes at
the provided times (n80 35%), not interested in exer-cise classes (n55 24%), unwilling to accept random-
ization (n50 22%), problems with transportation and
mobility (n33 14%), and unknown reason (n13
6%). A total of 152 participants were randomized (Fig-
Posttreatment assessments were completed for 141 par-
ticipants (93%) and followup assessments were completed
12 weeks later for 133 participants (88%). Of the 11 par-
ticipants who withdrew from the study, 3 had been allo-
cated to hydrotherapy (transportation difculties, classes
held too late in the afternoon, not willing to forego phys-
ical treatment during the control period) and 8 had been
allocated to Tai Chi (disliked Tai Chi classes n2,
exacerbation of knee pain n2, wanted hydrotherapy,
lost interest, prohibitive family commitments). Apart from
differences in study signal joint and the DASS21 stress
subscale scores, there were no other signicant differences
between the 3 allocation groups at baseline (Table 1).
Primary outcome (WOMAC).Posttreatment, signicant
improvements were evident in pain and function for both
the hydrotherapy group and the Tai Chi group (Table 2).
The magnitude of the treatment effect for physical func-
tion was moderate for both hydrotherapy and Tai Chi
classes (SRM 0.62; 95% CI 0.49, 0.75 and SRM 0.63; 95%
CI 0.50, 0.76, respectively) compared with the control
group (Table 3). Only the hydrotherapy classes resulted in
signicant improvement in pain scores, with a small treat-
ment effect (SRM 0.43; 95% CI 0.30, 0.56) compared with
the control group (25).
At the 12-week assessment, 27 participants (49%), 19
participants (34%), and 6 participants (15%) allocated to
Figure 1.The Physical Activity for Osteoarthritis Management
study ow chart. TKRtotal knee replacement; MImyocardial
Hydrotherapy and Tai Chi in OA Management409
hydrotherapy, Tai Chi, and the control group, respectively,
were treatment responders according to OMERACT
OARSI responder criteria D (22). The difference in treat-
ment responder rate between the 3 allocation groups was
212.4, 2 df,P0.002).
Secondary outcome measures.The hydrotherapy group
improved signicantly in the SF-12 PCS, 2 DASS21 sub-
scales, and all 3 physical performance measures (Up and
Go, 50-foot walk time, and stair climb), whereas the Tai
Chi group only improved signicantly in the timed stair
climb (Table 2). Only the hydrotherapy group demon-
strated signicant improvements, above the control group,
for the SF-12 PCS and all 3 measures of physical perfor-
mance (Table 3) (25).
Globally, 37 participants (67%), 26 participants (46%),
and 6 participants (15%) in the hydrotherapy, Tai Chi, and
control groups, respectively, reported that their signal hip
or knee joint was much better or better compared with 3
months earlier. Similarly, 27 participants (49%), 19 par-
ticipants (34%), and 4 participants (10%), respectively,
reported that their signal hip or knee was excellent, very
good, or good. The percentage of patients with at leastdaily analgesia use remained relatively unchanged at week
12 compared with baseline (Table 1): 49%, 36%, and 51%
for the hydrotherapy, Tai Chi, and control groups, respec-
Followup assessment.The followup assessment was
conducted for all participants on the basis of their active
treatment allocation, i.e., the results included those
achieved by the control participants after completion of
the hydrotherapy or Tai Chi classes (Figure 1). The signif-
icant improvements achieved at 12 weeks were generally
sustained for 3 months after cessation of classes (Table 4).
In this active treatment cohort, 32 (42%) and 23 (31%)
participants in the hydrotherapy and Tai Chi groups, re-
spectively, were OMERACTOARSI treatment responders
at 12 weeks. Among these 12-week responders, 21 (66%)
and 18 (58%) participants in the hydrotherapy and Tai Chi
groups, respectively, were still treatment responders at 24
Treatment adherence.Of participants allocated to hy-
drotherapy and Tai Chi, 62 (81%) and 46 (61%) partici-
Table 1. Baseline personal characteristics and disease severity measures*
Female sex, no. (%) 40 (73) 38 (68) 34 (83)
Age, years 70.06.3 70.86.3 69.66.1
230.05.0 29.65.9 30.75.0
Comorbidity score (016) 4.53.0 4.42.7 5.22.3
EuroQol 5D (01) 0.620.20 0.630.20 0.560.26
Symptom duration, no. (%)
6 years 17 (31) 26 (46) 9 (22)
610 years 19 (35) 15 (27) 19 (46)
10 years 19 (35) 14 (25) 12 (29)
Both knees involved, no. (%) 41 (74) 34 (61) 33 (80)
Both hips involved, no. (%) 11 (20) 16 (29) 12 (29)
Signal joint knee, no. (%) 51 (93) 41 (73) 36 (88)
Signal joint hip, no. (%) 4 (7) 15 (27) 5 (12)
Joint surgery, no. (%) 17 (31) 14 (33) 14 (34)
History of joint trauma, no. (%) 11 (20) 14 (33) 11 (27)
At least daily oral analgesia, no. (%) 27 (49) 23 (41) 19 (46)
Global assessment joint: poorfair, no. (%) 52 (95) 48 (86) 35 (85)
Pain 38.217.4 40.319.0 44.417.0
Function 46.320.4 47.220.6 50.819.3
SF-12 (meanSD 5010)
PCS 31.98.5 35.69.6 33.210.1
MCS 53.411.1 50.911.4 47.712.4
Depression 6.86.8 7.48.5 9.510.3
Anxiety 4.96.3 5.55.7 6.97.7
Stress 9.58.2 9.38.4 13.79.7
Up and Go 8.92.0 9.12.4 8.92.0
50-foot walk 11.22.3 11.32.3 11.32.1
Stair climb 15.65.3 15.36.3 16.04.7
* Values are the meanSD unless otherwise indicated. BMIbody mass index; WOMACWestern Ontario and McMaster Universities
Osteoarthritis Index; SF-12Medical Outcomes Study Short Form 12 Health Survey; PCSphysical component summary; MCSmental component
summary; DASS21Depression, Anxiety and Stress Scale.
410Fransen et al
pants, respectively, attended12 of the available 24
Adverse events.During the study period, 11 partici-
pants reported a serious adverse event requiring hospital-
ization. None of these events took place during class at-
tendance or could be related to the interventions. One
participant withdrew from hydrotherapy and one with-
drew from Tai Chi due to exacerbation of low back pain.
This randomized trial demonstrates that access to 12
weeks of hydrotherapy or Tai Chi classes for older com-
munity-dwelling individuals with OA of the hip or knee
joints will produce measurable improvements. At the end
of the available classes, joint pain and physical function
clearly improved. These improvements were sustained for
most participants for a further 12 weeks.
Due to an absolute paucity of well-conducted clinical
trials with comparable samples of patients with OA, it is
not useful to compare the results of this study with thoseof previous studies evaluating hydrotherapy or Tai Chi
classes. However, the treatment effect sizes for joint pain
achieved by the hydrotherapy classes (Table 3) were
within the 0.30 to 0.47 95% CI demonstrated in a recent
meta-analysis of randomized trials evaluating land-based
formal graded exercise programs among patients with knee
OA (3). Interestingly, the treatment effect sizes for physical
function demonstrated by the hydrotherapy or Tai Chi
classes (Table 3) were signicantly higher than the 0.23 to
0.39 95% CI demonstrated in the same meta-analysis. This
larger benecial effect on physical function may be related
to the focus on whole body movement in both hydrother-
apy and Tai Chi programs compared with traditional
graded therapeutic exercise (3). Although the signicant
self-reported improvements in physical function were ac-
companied by signicant moderate to large improvements
in all 3 physical performance measures for hydrotherapy,
Tai Chi classes did not result in any signicant improve-
ments in these objective measures (Table 3).
In contrast to the SF-12 PCS, there were no signicant
improvements in the SF-12 MCS for either treatment allo-
cation. This nding may be related to the inuence of a
Table 2. MeanSD score and mean change (95% CI) at 12 weeks*
MeanSD 27.318.7 30.718.9 40.016.2
Change (95% CI) 10.9 (6.5, 15.3) 9.6 (5.4, 13.7) 4.4 (0.2, 8.6)
MeanSD 34.823.7 36.620.9 49.919.0
Change (95% CI) 11.4 (6.3, 16.6) 10.6 (5.6, 15.7) 0.9 (3.6, 5.4)
SF-12 (mean 50)
MeanSD 35.79.8 37.611.2 33.110.6
Change (95% CI) 3.8 (0.6, 7.0) 1.9 (0.4, 4.2)0.2 (2.4, 2.0)
MeanSD 54.68.9 50.910.7 48.011.4
Change (95% CI) 1.2 (0.9, 3.3)0.0 (2.7, 2.6) 0.2 (3.1, 3.6)
MeanSD 4.76.1 7.08.3 9.011.0
Change (95% CI) 2.2 (0.9, 3.4) 0.4 (1.9, 2.7) 0.5 (1.6, 2.6)
MeanSD 4.65.2 5.16.0 7.37.8
Change (95% CI) 0.3 (1.0, 1.5) 0.3 (1.1, 1.8)0.4 (2.1, 1.3)
MeanSD 7.18.0 8.18.6 12.610.9
Change (95% CI) 2.4 (0.9, 3.8) 1.1 (0.6, 2.8) 1.1 (1.6, 3.6)
Up and Go
MeanSD 8.21.7 8.83.0 9.22.2
Change (95% CI) 0.7 (0.3, 1.1) 0.2 (0.2, 0.7)0.3 (0.6, 0.1)
50-foot walk time
MeanSD 10.32.2 11.03.3 11.12.1
Change (95% CI) 0.9 (0.6, 1.3) 0.3 (0.2, 0.8) 0.2 (0.3, 0.6)
MeanSD 13.85.1 14.26.5 15.85.4
Change (95% CI) 1.8 (0.9, 2.7) 1.1 (0.4, 1.8) 0.2 (0.5, 1.0)
* 95% CI95% condence interval; see Table 1 for additional denitions.
Hydrotherapy and Tai Chi in OA Management411
ceiling effect, with baseline MCS scores in this patient
sample being comparable with population norms. How-
ever, the DASS21 provided a more detailed questionnaire
on psychological well-being, with 21 questions evaluating
depression, anxiety, and stress. Interestingly, there was a
trend towards signicant improvements in the depression
and stress subscales for patients allocated to hydrotherapy,
but not for patients allocated to Tai Chi (Tables 3 and 4).
This study aimed to evaluate both the clinical effective-
ness and the acceptance of hydrotherapy and Tai Chi
classes for individuals with chronic symptomatic OA.
Class attendance rates were lower for Tai Chi. In addition,
4 Tai Chi class participants withdrew from the study be-
cause they disliked the classes or believed the classes
exacerbated their knee pain. The lower adherence rates
may simply reect the almost exclusively white ethnicity
of the study sample (only 1 Asian participant) or the lessthan optimal gym environment. However, while the slow
movements involved in Tai Chi result in a low-impact
form of exercise, Tai Chi is mostly performed with bent
knees in a semisquat position requiring sustained lower-
limb muscle control. Of the 152 study participants, 84%
indicated a knee joint as the signal (most painful) joint.
This weight-bearing position is likely to be difcult and
painful for persons with knee OA, particularly if arthritic
changes are present in the patellofemoral joint. Future
studies should evaluate the temporary incorporation of
patella taping to potentially allow pain-free Tai Chi prac-
tice (26). Water-based Tai Chi may also provide a less
painful avenue for physical activity; however, the clinical
benets for patients with knee OA have yet to be evaluated.
Many participants with severe structural joint disease
are likely to have been recruited because most had been
symptomatic for 6 years or more (Table 1). Inclusion of
Table 3. Effect size (0 12 weeks) compared with control group*
Hydrotherapy above control Tai Chi above control
Pain 6.5 (0.4, 12.7) 0.43 (0.30,0.56) 5.2 (0.8, 11.1) NS
Function 10.5 (3.6, 14.5) 0.62 (0.49,0.75) 9.7 (2.8, 16.7) 0.63 (0.50,0.76)
SF-12 version 2 (mean 50)
PCS 4.0 (0.8, 7.2) 0.34 (0.21,0.47) 2.1 (0.2, 4.4) 0.25 (0.12,0.38)
MCS 0.9 (1.2, 3.0) NS0.03 (2.9, 2.3) NS
Depression 1.7 (0.6, 4.0) NS0.1 (3.6, 3.1) NS
Anxiety 0.7 (1.3, 2.7) NS 0.7 (1.5, 2.9) NS
Stress 1.3 (1.5, 4.0) NS 0.01 (3.0, 3.0) NS
Up and Go 1.0 (0.4, 1.5) 0.76 (0.63,0.89) 0.5 (0.2, 1.2) 0.32 (0.19,0.45)
50-foot walk time 0.8 (0.2, 1.4) 0.49 (0.36,0.62) 0.1 (0.6, 0.8) NS
Stair climb 1.6 (0.4, 2.8) 0.55 (0.42,0.68) 0.8 (0.2, 1.9) 0.36 (0.23,0.49)
* 95% CI95% condence interval; SRMstandardized response mean; NSnonsignicant change; see Table 1 for additional denitions.
Borderline signicant change.
Table 4. Active treatment cohort: short-term effects (0 12 weeks) and sustainability (1224 weeks)*
Hydrotherapy (n77) Tai Chi (n75)
012 weeks 1224 weeks 012 weeks 1224 weeks
Pain 9.5 (5.8, 13.2)1.1 (4.4, 2.1) 7.5 (3.7, 11.2)1.4 (4.7, 1.9)
Function 11.2 (7.0, 15.4)2.2 (5.6, 1.1) 8.4 (4.1, 12.7)0.6 (3.9, 2.7)
SF-12 version 2 (mean 50)
PCS 4.6 (2.1, 7.1) 2.2 (0.2, 4.2) 2.3 (0.2, 4.3) 1.3 (0.1, 2.6)
MCS 0.7 (1.1, 2.5)2.9 (4.3,1.5)0.1 (2.3, 2.0)1.1 (2.6, 0.5)
Depression 2.5 (1.2, 3.7) 0.7 (0.5, 1.9) 0.00.6 (1.9, 0.7)
Anxiety 0.7 (0.4, 1.8) 0.2 (0.7, 1.1) 0.0 0.2 (0.9, 1.3)
Stress 3.4 (1.7, 5.2)1.1 (2.5, 0.3) 0.4 (1.1, 1.9)0.5 (1.8, 0.9)
Up and Go 0.8 (0.4, 1.1) 0.0 0.2 (0.2, 0.6) 0.0
50-foot walk time 0.7 (0.4, 1.0) 0.0 0.3 (0.1, 0.7) 0.0
Stair climb 1.5 (0.8, 2.3) 0.1 (0.4, 0.6) 0.7 (0.1, 1.5) 0.3 (0.2, 0.7)
* Values are the mean change (95% condence interval). Positive changes are improvements. See Table 1 for denitions.
412Fransen et al
many participants with severe structural disease may have
diluted treatment effects as we have shown previously that
severe structural disease, in terms of markedly reduced
tibiofemoral joint space width, is associated with reduced
responsiveness to exercise (9). Unfortunately, funding lim-
itations precluded baseline radiographs to quantify radio-
graphic disease severity in this study. The small number of
participants indicating a hip joint as the signal joint also
did not allow subgroup analysis of treatment effectiveness
according to specic joint involvement.
This clinical trial involved a physical intervention;
therefore, participants were not blinded to treatment allo-
cation, possibly inating treatment effect sizes. However,
all participants were aware that the study staff conducting
the outcomes assessments were both blinded to their treat-
ment allocation and had no interest in promoting the su-
perior benets of either of the 2 treatment options.
In addition to blinding of outcomes assessment, this
study had several other strengths. The randomization pro-
cedure was rigorous and was conducted offsite, followup
was excellent, the main outcomes were specied a priori,
several therapists were involved in the supervision of the
hydrotherapy and Tai Chi classes, and patient-relevant
outcomes were collected with well-validated questions
and supported by objective measures of physical perfor-
mance. In addition, the project manager was an experi-
enced clinical physiotherapist able to effectively screen
participants prior to randomization, reducing the number
of participants with self-reported hip and knee joint pain
referred from the lumbar spine.
In conclusion, this study demonstrated that access to 12
weeks of hydrotherapy classes or Tai Chi classes for fairly
sedentary older individuals (59 years of age) with
chronic symptomatic knee or hip OA resulted in clinical
benets that were sustained a further 12 weeks. Both types
of classes resulted in large improvements in self-reported
physical function, greater than improvements demonstrated
for traditional land-based exercise. In this almost exclusively
white sample, hydrotherapy classes appeared to be more
acceptable (higher attendance), appeared to provide greater
relief of joint pain, and resulted in larger improvements in
objective measures of physical performance.
We thank the hydrotherapy physiotherapists (Guni Hinchey,
Kim Walker, Cathy Brand, and Khim Kwan) and the Tai Chi
trainers (Joan Peters, Pat Weber, Fiona Black, and Jenny Al-
fonso). Lai-Hoong Wong kindly allowed the use of the phys-
iotherapy department facilities at St George Hospital.
Dr. Fransen had full access to all of the data in the study and
takes responsibility for the integrity of the data and the accuracy
of the data analysis.
Study design.Fransen, Nairn, Lam, Edmonds.
Acquisition of data.Fransen, Nairn, Edmonds.
Analysis and interpretation of data.Fransen, Nairn, Winstanley,
Manuscript preparation.Fransen, Nairn, Winstanley, Edmonds.
Train Tai Chi instruction.Lam.
1. Macera CA, Hootman JM, Sniezek JE. Major public health
benets of physical activity. Arthritis Rheum 2003;49:122 8.
2. Hootman JM, Macera CA, Ham SA, Helmick CG, Sniezek JE.
Physical activity levels among the general US adult popula-
tion and in adults with and without arthritis. Arthritis Rheum
3. Fransen M, McConnell J, Bell M. Exercise for osteoarthritis of
the hip or knee review. Cochrane Database Syst Rev 2003;3:
4. Pelland L, Brosseau L, Wells G, MacLeay L, Lambert J, Lamothe
C, et al. Efcacy of strengthening exercises for osteoarthritis. Part
I. A meta-analysis. Phys Ther Rev 2004;9:77108.
5. Brosseau L, Pelland L, Wells G, Macleay L, Lamothe C,
Michaud G, et al. Efcacy of aerobic exercises for osteoarthri-
tis. Part II. A meta-analysis. Phys Ther Rev 2004;9:125 45.
6. Munro JF, Nicholl JP, Brazier JE, Davey R, Cochrane T. Cost
effectiveness of a community based exercise programme in
over 65 year olds: cluster randomised trial. J Epidemiol Com-
munity Health 2004;58:1004 10.
7. Verhagen AP, Bierma-Zeinstra SM, Cardoso JR, de Bie RA,
Boers M, de Vet HC. Balneotherapy for rheumatoid arthritis.
Cochrane Database Syst Rev 2003;4:CD000518.
8. Foley A, Halbert J, Hewitt T, Crotty M. Does hydrotherapy
improve strength and physical function in patients with
osteoarthritis: a randomised controlled trial comparing a gym
based and a hydrotherapy based strengthening programme.
Ann Rheum Dis 2003;62:11627.
9. Fransen M, Crosbie J, Edmonds J. Physical therapy is effective
for patients with osteoarthritis of the knee: a randomized
controlled clinical trial. J Rheumatol 2001;28:156 64.
10. Hartman CA, Manos TM, Winter C, Hartman DM, Li B, Smith
JC. Effects of Tai Chi training on function and quality of life
indicators in older adults with osteoarthritis. J Am Geriatr Soc
11. Song R, Lee EO, Lam P, Bae SC. Effects of Tai Chi exercise on
pain, balance, muscle strength, and perceived difculties in
physical functioning in older women with osteoarthritis: a
randomized clinical trial. J Rheumatol 2003;30:2039 44.
12. Altman R, Alarcon G, Appelrouth D, Bloch D, Borenstein D,
Brandt K, et al. The American College of Rheumatology cri-
teria for the classication and reporting of osteoarthritis of the
hip. Arthritis Rheum 1991;34:50514.
13. Altman R, Asch E, Bloch D, Bole G, Borenstein D, Brandt K, et
al. Development of criteria for the classication and reporting
of osteoarthritis: classication of osteoarthritis of the knee.
Arthritis Rheum 1986;29:1039 49.
14. Sangha O, Stucki G, Liang MH, Fossel AH, Katz JN. The
Self-Administered Comorbidity Questionnaire: a new method
to assess comorbidity for clinical and health services re-
search. Arthritis Rheum 2003;49:156 63.
15. 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.
16. Gandhi SK, Salmon JW, Zhao SZ, Lambert BL, Gore PR,
Conrad K. Psychometric evaluation of the 12-item short-form
health survey (SF-12) in osteoarthritis and rheumatoid arthri-
tis clinical trials. Clin Ther 2001;23:1080 98.
17. Lovibond PF, Lovibond SH. The structure of negative emo-
tional states: comparison of the depression anxiety stress
scales (DASS) with the Beck Depression and Anxiety inven-
tories. Behav Res Ther 1995;33:335 43.
18. Bellamy N, Kirwan J, Boers M, Brooks P, Strand V, Tugwell P,
et al. Recommendations for a core set of outcome measures for
future phase III clinical trials in knee, hip, and hand
osteoarthritis: consensus development at OMERACT III.
J Rheumatol 1997;24:799 802.
19. Grace EM, Gerecz EM, Kassam YB, Buchanan HM, Buchanan
Hydrotherapy and Tai Chi in OA Management413
WW, Tugwell PS. 50-foot walking time: critical assessment of
an outcome measure in clinical therapeutic trials of antirheu-
matic drugs. Br J Rheumatol 1988;27:372 4.
20. Piva SR, Fitzgerald GK, Irrgang JJ, Bouzubar F, Starz TW. Get
Up and Go test in patients with knee osteoarthritis. Arch Phys
Med Rehabil 2004;85:284 9.
21. Beaton DE, Hogg-Johnson S, Bombardier C. Evaluating
changes in health status: reliability and responsiveness of ve
generic health status measures in workers with musculoskel-
etal disorders. J Clin Epidemiol 1997;50:79 93.
22. Pham T, van der Heijde D, Altman RD, Anderson JJ, Bellamy
N, Hochberg M, et al. OMERACT-OARSI initiative: Osteoar-
thritis Research Society International set of responder criteriafor osteoarthritis clinical trials revisited. Osteoarthritis Carti-
lage 2004;12:389 99.
23. Bellamy N, Carette S, Ford PM, Kean WF, le Riche NG, Lus-
sier A, et al. Osteoarthritis antirheumatic drug trials. III. Set-
ting the delta for clinical trials: results of a consensus devel-
opment (Delphi) exercise. J Rheumatol 1992;19:4517.
24. Kirby A, Gebski V, Keech AC. Determining the sample size in
a clinical trial. Med J Aust 2002;177:256 7.
25. Cohen J. Statistical power analysis for the behavioral sci-
ences. Revised ed. New York: Academic Press; 1977.
26. Hinman RS, Crossley KM, McConnell J, Bennell KL. Efcacy
of knee tape in the management of osteoarthritis of the knee:
blinded randomised controlled trial. BMJ 2003;327:135 40.
APPENDIX A: HYDROTHERAPY PROTOCOL (Water temperature 34C).
Patients exercise with the water at approximately waist height. Patients are required to bring a bottle of water and the classes
pause between blocks for a drink.
WALKING Each direction 6 laps minimum
Forward, backward, sideways Large steps, arms draw through the water, bent knees
Forward With board (after 4 weeks)
Forward running 2 minutes
BAR WORK: One hand hold Upright posture maintained throughout
Swinging outside leg forwardbackward Straight knees, picking up speed
Hip abduction 20 reps bilateral Straight knees, picking up speed, big toe shows down
BAR WORK: Both hands on bar
Hip abduction 20 reps; bilateral Abducted leg exed
Hip rotations 10 reps; bilateral Flexed knee, small circle, forward, side, back, down
Tipping Big toe touches to side, straight leg
Push ups Kneehip extensors activated, lift elbows out of water
Push ups As before right leg, left knee forward
Walking up the wall on toes Knees fully exed
Knees open and close Knees fully exed
Both knees together to right and left Knees fully exed
Bend and straight knees Rhythmic swinging forward and backward
Heel raise to squat Emphasize quadriceps and gluteus activation
Side to side lunges On the oor or feet under the bar
Squats 10 reps Back to wall, touching wall with buttocks
Running backward 2 minutes
Lift legs20, right then left Keep thighs on seat, increase speed with improvement
Lift legs alternatively40 Really pumping
DEEP WATER NOODLE
Scissors Legs swing forward and backward
Knees to chest Leg exion
Opencross Leg abductionadduction
STEP Activate correct posture
Step up then over, turn and repeat Picking up speed with improved skill
Step up and down sideways Holding on only if necessary
BAR WORK WITH RINGSNOODLES
Push noodlelarge ring underwater with foot 10 times each, assist patients from outside of pool
Hook small ring over foot, extend knee
FREE STANDING Advanced: closed eyes, closed arms
Swinging one leg forward and backward Standing on 1 leg, both knees straight
Circle ankle, knee, and hip and reverse Standing on 1 leg, other knee is bent to 90
Trunk rotations The Hula
Twist Knees bent
Alternate cross leg lifts slow change into Touch knee with opposite hand
On the spot Advance through waist high water, gaiters
Sideways 12 minutes
Forwardbackwardquick successions After week 4
Entryexit pool310 Practice, increase repetition and speed as able
414Fransen et al