Complementary Therapies in Clinical Practice 22 (2016) 87e92

Contents lists available at ScienceDirect

Complementary Therapies in Clinical Practice

journal homepage: www.elsevier.com/locate/ctcp

Knee osteoarthritis pain in the elderly can be reduced by massage

therapy, yoga and tai chi: A review
Tiffany Field
University of Miami School of Medicine, Fielding Graduate University, USA

a r t i c l e i n f o a b s t r a c t

Article history: Background and methods: This is a review of recently published research, both empirical studies and
Received 3 November 2015 meta-analyses, on the effects of complementary therapies including massage therapy, yoga and tai chi on
Received in revised form pain associated with knee osteoarthritis in the elderly.
24 December 2015
Results: The massage therapy protocols have been effective in not only reducing pain but also in
Accepted 12 January 2016
increasing range of motion, specifically when moderate pressure massage was used and when both the
quadriceps and hamstrings were massaged. The yoga studies typically measured pain by the WOMAC.
Keywords:
Most of those studies showed a clinically significant reduction in pain, especially the research that
Knee osteoarthritis
Massage therapy
focused on poses (e.g. the Iyengar studies) as opposed to those that had integrated protocols (poses,
Yoga breathing and meditation exercises). The tai chi studies also assessed pain by self-report on the WOMAC
Tai chi and showed significant reductions in pain. The tai chi studies were difficult to compare because of their
highly variable protocols in terms of the frequency and duration of treatment.
Discussion: Larger, randomized control trials are needed on each of these therapies using more stan-
dardized protocols and more objective variables in addition to the self-reported WOMAC pain scale, for
example, range-of-motion and observed range-of-motion pain. In addition, treatment comparison
studies should be conducted so, for example, if the lower-cost yoga and tai chi were as effective as
massage therapy, they might be used in combination with or as supplemental to massage therapy.
Nonetheless, these therapies are at least reducing pain in knee osteoarthritis and they do not seem to
have side effects.
© 2016 Published by Elsevier Ltd.

Knee osteoarthritis pain in the elderly can be reduced by mas- [1].The pain experience of knee osteoarthritis is apparently due to
sage therapy, yoga and tai chi. activation of sensory pain fibers in the arthritic joint and to
Knee osteoarthritis affects some 80% of elderly people. Phar- weakening of the surrounding muscles.
maceuticals have been relatively effective for the treatment of knee Knee osteoarthritis is reputedly the most common joint disease
osteoarthritis but often have undesirable side effects. Comple- in the elderly and the largest cause of functional disability with
mentary therapies have also been effective in reducing knee pain some 80% of people over 65 years of age showing radiological
but without side effects. This paper is a review of recent (this past symptoms of osteoarthritis [2]. In the U.S. alone, reportedly 27
decade) empirical studies and meta-analyses (that appeared on million people are affected by knee osteoarthritis with associated
PUBMED) on complementary therapies that have reduced knee treatment costs of $185.5 billion per year [3]. The incidence has
pain including massage therapy, yoga and tai chi. supposedly doubled in women and tripled in men over the last 20
years [4]. Leading risk factors for osteoarthritis aside from age and
1. Knee osteoarthritis genetics include female gender and obesity as well as excessive
sports or occupational stress [5].
Knee osteoarthritis involves degeneration of the cartilage in the The main focus of treatment has been to relieve pain, to restore
joint with pain in and around the joint as well as joint stiffness and function and to slow the progression of the disease. The treatments
restricted movements that ultimately lead to muscle weakness have been classified as pharmacological, non-pharmacological and
surgical or combinations of these [2]. Anti-inflammatory drugs as
well as non-opioid analgesics have been prescribed for the
E-mail address: tfield@med.miami.edu.

http://dx.doi.org/10.1016/j.ctcp.2016.01.001
1744-3881/© 2016 Published by Elsevier Ltd.
88 T. Field / Complementary Therapies in Clinical Practice 22 (2016) 87e92

reduction of inflammation and pain [1]. Although these have with arthritis in their upper limbs, the moderate pressure massage
effectively reduced the inflammation and pain, they have led to group versus the light pressure massage group had less pain and
undesirable side effects in long-term follow-up studies including, greater grip strength following the first and last sessions.
for example, heart failure and hypertension [6]. Because of these By the end of the one-month treatment period the moderate
potential adverse side effects, the American College of Rheuma- pressure group was reporting and showing less ROM-related pain
tology has advised the use of non-pharmacological therapies behavior (e.g., grimacing), and greater range of motion. Further, the
including physical therapy and exercise [7].With joint pain and massage in our study was focused on the quadriceps muscle inas-
limited mobility, however, most individuals with knee osteoar- much as researchers have reported a relationship between quad-
thritis do not participate in regular physical activity [7]. These riceps weakness, increased pain and altered walking patterns [9].
epidemiological data highlight the need for complementary/inte- But, the hamstrings were also massaged given that the previous
grative therapies such as massage therapy and lower-impact studies [10,11] failed to find ROM increases when focusing only on
physical exercise such as yoga and tai chi. The following sections the quadriceps. The results of this study on increased ROM and
of this paper are reviews of research on these therapies. decreased self-reported pain as well as decreased ROM-related pain
are consistent with those we have previously reported, i.e. changes
2. Massage therapy for knee osteoarthritis in ROM and pain following moderate pressure massage in adults
with arthritis of the upper limbs [14] and in the neck [15]. Other
Research on the effects of massage therapy on knee osteoar- researchers have noted a reduction in knee osteoarthritis pain
thritis pain has been limited, although the results have consistently following massage, but only by self-report (WOMAC), not by direct
suggested that the pain from that condition can be reduced by observation of ROM-related pain [10,11]. In one of the few studies
massage therapy. The pain, however, has typically been self- on massage for adults with knee osteoarthritis, for example, pain
reported on visual analogue pain scales (e.g. happy to sad faces or and stiffness were reduced and functionality was increased [10].
0e100 thermometer scales) and on the Western Ontario and However, no ROM changes resulted from this self-massage study,
McMaster Universities Arthritis Index (WOMAC) [8e11] with all the possibly because it was a self-massage study and because it is not
limitations of self-report studies. In this research, more objective clear that moderate pressure massage was applied. As we have
observation measures such as range of motion (ROM)-related pain noted in our earlier studies, moderate pressure is necessary for
were not taken. ROM was not measured in two of the four studies positive changes to occur [13e15]. The positive effects in our self-
[8,9]. In the first of these, pain was measured on a 10-point Likert massage studies may have derived from moderate pressure being
scale, and pain decreased even though, according to the authors of applied and/or the combination of therapist massage (once a week)
that study, only the patient's healthy foot, hands and upper parts of and the participants' self-massage (once a day).
the shoulders were massaged “shallowly” for 20 min each day of Another potential interpretation for the inconsistent findings
their hospitalization [8]. between the increased ROM we noted and the lack of change in
In the other two WOMAC studies, ROM was measured but did ROM following massage reported by the other group is that their
not change [10,11]. In one of these studies a self-massage protocol knee self-massage was focused solely on the quadriceps muscle
was used [10]. This raises the possibility that the pressure being group {10}. The authors of that self-massage study suggested that
applied was not sufficient (moderate pressure massage being key to despite the earlier research on joint cartilage degeneration as the
positive effects). This would be especially true if the participants key factor in knee arthritis, the more recent research had noted that
were not instructed to use moderate pressure and given that they quadriceps muscle weakness that affects joint loading and propri-
would not be inclined themselves to apply pressure to the area oceptive deficits contributed to knee arthritis [10]. Others have also
around the painful joint. In the second study, pain and stiffness found relationships between weak quadriceps muscles and
were reduced, and increased function was noted on the self-report increased pain and limited walking [9]. That was the rationale for
WOMAC scale, but the ROM results were negative [11]. This could their self-massage protocol focusing on the quadriceps muscle [10].
be related not only to the use of low pressure massage but also to However, the findings from our study suggest that it may be
the massage not being focused on the affected leg [11]. Only 50% of necessary to massage both the hamstrings and the quadriceps
the hour-long massages were applied to the affected leg, and the muscles to achieve increased ROM, especially since the hamstring
massage protocol, again, may have lacked sufficient pressure to muscles are noted to work together to flex the knee (12).
increase ROM. Our results may be inconsistent with those of Perlman et al. [11]
A comparison between Thai massage and Swedish massage for a for different reasons. They again found changes on the WOMAC
sample of older people with knee osteoarthritis further supported self-report scale on pain, but no changes in ROM even though their
the need for moderate pressure [9]. In that study, the group who massages were longer (30e60 min), more frequent (two to three
received Thai massage (which typically involves more pressure times weekly) and for a longer study period (8 weeks) than ours. As
than Swedish massage) reported a greater reduction in pain on the already mentioned, their Swedish massages may have lacked suf-
WOMAC than the group who received Swedish massage. ficient pressure, and, as already mentioned, the lower limbs were
Based on these mixed findings, we recently conducted a knee only massaged 50% of the sessions. Their results were inconsistent
osteoarthritis massage study in which moderate pressure massage with ours in that they only observed reduced pain after 5 weeks of
was applied to the affected leg by massage therapists [12]. Because 60- minute massages two or three times weekly (as opposed to
earlier research was only focused on the quadriceps muscles, we their lower dose group receiving only 30- minute massages two or
designed a massage therapy protocol that was focused on the three times weekly) [11]. One possible explanation for the positive
hamstrings as well as the quadriceps muscles, thinking that both effects following the shorter and less frequent massages in our
sets of muscles were involved in ROM (flexion and extension of the study (20 min weekly for 4 weeks) is our use of moderate pressure
knee). And the assessments not only included self-reported pain, massage [12], although it is not clear what pressure was used in
but also ROM and ROM-related pain. Moderate pressure massage their study [11]. Cross-study comparisons are difficult because of
therapy (moving the skin) was used inasmuch as it has been noted the different massage protocols and the various outcome measures
to be more effective than light pressure massage (light stroking) used, i.e. self-report pain scales in their study [11] and the more
with adults with hand pain [13], upper arm and shoulder pain [14] directly observed ROM-related pain measures in our study.
and neck arthritis pain [15]. For example, in the study on adults Combining therapist-delivered massage with daily self-
 T. Field / Complementary Therapies in Clinical Practice 22 (2016) 87e92 89

massages has been effective now in at least three studies and weeks as opposed to 2 weeks, although the yoga sessions in this
suggests that this may be a more cost-effective therapy for in- case were not preceded by a standard treatment like physiotherapy.
dividuals with arthritis pain and one that may have more sustain- In another Iyengar yoga study EMG biofeedback was used prior to
ability [13e15]. The data from our knee osteoarthritis study also the yoga sessions [20]. The group who received EMG biofeedback
highlight the importance of designing massage therapy protocols followed by yoga versus the EMG biofeedback alone group reported
that target muscle groups affected by the joint movements and a greater decrease in pain (57 vs. 38% on a visual analogue pain scale
then assessing those specific range of motion measures and the and 59 vs. 34% on the WOMAC scale).
ROM-associated pain. In a meta-analysis, 18 studies on older adults (greater than age
60) met the criteria for a meta-analysis [21]. The meta-analysis
3. Potential underlying mechanisms for massage therapy revealed that yoga benefits exceeded those of conventional exer-
alleviating knee osteoarthritis pain cise interventions for strength, sleep and depression, although the
effect sizes were modest and the methodological quality of the
The underlying mechanism for the relief of knee osteoarthritis studies was mixed. Their conclusion that yoga was more effective
pain and the increase in ROM is not clear. We have reported else- than conventional exercise for osteoarthritis may have related to
where that moderate pressure massage is accompanied by greater compliance by the participants in the more gentle exercise-
decreased heartrate, suggesting a relaxed state [16]. We have also yoga studies versus the more rigorous conventional exercise pro-
noted that the stimulation of pressure receptors results in increased tocols. In addition, this meta-analysis was not limited to patients
serotonin (the body's natural pain suppressor) which may be the with knee osteoarthritis and the studies that were included in the
primary underlying mechanism for pain relief [16]. A related pos- meta-analysis were of mixed quality and yielded only modest
sibility is the substance P decreases (substance P causing pain) that effects.
we have documented following massage therapy in fibromyalgia The underlying mechanisms for yoga effects are not clear,
patients when they are experiencing more deep/restorative sleep although some have suggested hormonal and neurotransmitter
following massage [16]. Further research is needed to explore the changes, better posture and improved muscle tonus as well as a
potential underlying mechanisms for the reduction of knee osteo- lessening of depression and anxiety [22]. The significant variety of
arthritis pain and other pain syndromes following massage therapy. yoga styles and the different outcome measures have made it
In addition, the moderate pressure massage protocol needs to be difficult to interpret results across studies. Nonetheless, the meta-
replicated in a more representative sample than the medical school analysis studies, the reviews of the literature and the empirical
employees we have assessed. studies just reviewed converge to suggest that yoga can be an
The clinical implications of these data include the effective use effective complementary therapy for reducing knee osteoarthritis
of moderate pressure massage [16] and massaging both the ham- pain.
strings and quadriceps to achieve better knee ROM and the
reduction of knee ROM-related pain. The addition of self-massage
on the days between massage therapy sessions might lead to
even greater therapy effects, as has been noted in our earlier studies 5. Potential underlying mechanisms for yoga reducing pain
on arthritis of other joints (13e15).
The mechanism that has been most frequently used to explain
4. Yoga therapy for knee osteoarthritis massage therapy effects on pain syndromes, the Gate Control the-
ory [16] might also pertain to yoga inasmuch as yoga is a form of
Osteoarthritis of the knee has also been treated by yoga. Yoga self-massage, as in limbs rubbing against limbs and against the
has mostly been used as a loosening and strengthening form of floor and stimulating pressure receptors. According to the Gate
exercise with young adults. However, interest in yoga is growing Control theory, pain stimulates shorter and less myelinated (or less
among older adults [17]. Yoga, like massage therapy, may be insulated) nerve fibers so that the pain signal takes longer to reach
effective in pain management because of the stimulation of pres- the brain than the pressure signal which is carried by nerve fibers
sure receptors, in turn, increasing vagal activity and serotonin and that are more insulated and longer and therefore able to transmit
slowing the production of cortisol and substance P [16]. the stimulus faster. The message from the pressure stimulation
Yoga studies on knee osteoarthritis have varied on many di- reaches the brain prior to the pain message and “closes the gate” to
mensions including the different types of yoga practiced as well as the pain stimulus. This metaphor for the electrical and biochemical
the differing length of sessions, and the frequency and duration of changes that likely occur has been commonly used in explaining
the treatment. Several of the studies have used an integrative the effect of grabbing your crazy bone when it has been bumped.
program of yoga consisting of not only physical poses (asanas) but Another theory that is commonly referenced is the deep sleep
also breathing (pranayama) and meditation exercises which further theory. In deep sleep, less substance P is emitted and therefore less
confounds the question of which aspect/protocol of yoga is effec- pain occurs because substance P causes pain. As already mentioned,
tive. In one of these studies a 40% reduction in pain followed one we directly tested the “enhanced deep sleep leading to less sub-
week of yoga in a camp setting [18]. And, an assessment on the 15th stance P” theory in our study on fibromyalgia [16]. Following a
day revealed a 34% reduction in resting pain and a 69% decrease in period of massage therapy, more time was spent in deep sleep, and
morning stiffness. These decreases were significantly greater than lower levels of substance P were noted in the saliva samples taken.
the decreases noted for the randomized control group. Still another theory is that less pain results from increased se-
These studies were not only confounded by the different types rotonin levels [16], serotonin being the body's natural anti-pain
of yoga (the poses, breathing and meditation exercises) being chemical. Serotonin, in turn, decreases cortisol and depression
combined but also by their being preceded by a standard treatment which are also important effects of yoga. And, serotonin decreases
for yoga e.g. transcutaneous electrical stimulation and ultrasound substance P and other pain-causing chemicals, highlighting the
or physiotherapy. In a study that used physical poses alone (Iyengar complex interaction between yoga effects on biochemistry. Future
standing yoga poses that are each held for a minute or so), a 47% yoga studies might use multiple physiological and biochemical
decrease in pain occurred [19].That study may have yielded a measures to enhance our understanding of the mechanisms un-
greater reduction in pain because the treatment period was 8 derlying the pain reduction effects of yoga.
90 T. Field / Complementary Therapies in Clinical Practice 22 (2016) 87e92

6. Tai chi for knee osteoarthritis week for 20 weeks suggests that this was not a cost-effective
intervention. Based on the group differences at 9 weeks, it would
Tai chi is a Chinese martial art/exercise that combines many appear that this program could be provided for only 8 weeks for it
poses that are made very slowly and smoothly in a continuous, to be successful.
circular movement. Tai chi has most notably been used for The authors noted that this 12-form SUN style tai chi program
enhancing balance and muscle strength in the elderly, although it has been endorsed by the American Arthritis Foundation and that it
has rarely been used for older adults with knee osteoarthritis (see would be expected to be good for arthritic knee pain because the
Ref. [23] for a review). Typically, the control group has been a weight-bearing involved in tai chi could strengthen the leg muscles
stretching, an awareness education, or a muscle relaxation group, including the quadriceps and the hamstrings. It could re-establish
and the Tai Chi groups have usually shown better performance than the normal mechanics of the joint. When the joint is stabilized
the control groups on all of the measures assessed. pain should be reduced because the stress and strain on the joint
In one study, kinematic analysis suggested that the Tai Chi where the pain receptors are located are reduced.
practitioners used a more cautious walking strategy, including Recent reviews of the literature further support the effective-
slower gait and shorter and slower steps than the control group ness of tai chi for patients with knee osteoarthritis. One of these
[24]. The more complex gait involved in Tai Chi would also research reviews yielded 74 studies, but 57 of them did not meet
contribute to greater balance. Compared with normal gait, elders the inclusion criteria and 11 of them did not meet the exclusion
during Tai Chi had significantly larger knee and hip flexions and criteria, leaving only 6 studies for analyses [34]. As in the previous
longer co-activation of most leg muscle pairs and greater activation studies, tai chi was noted to reduce pain and improve physical
of the leg muscles [25]. In another study, faster reflex reaction time function in patients with knee osteoarthritis.
was noted in the hamstrings and gastrocnemius muscles and a In a systematic review and meta-analysis, 5 randomized control
longer balance time on a tilt board [26].Among swimmers/runners, trials met inclusion criteria [35]. The meta-analysis showed mod-
the elderly people who regularly practiced Tai Chi not only showed erate evidence for short-term effectiveness of tai chi for reducing
better proprioception at the ankle and knee joints than sedentary pain and stiffness and increasing physical function. These authors
controls, but also better ankle kinesthesis than swimmers/runners concluded, as have most of the authors of the research already
who did not practice tai chi [27]. In a review of 24 studies, Tai Chi summarized, that more quality randomized controlled trials are
had beneficial effects on balance and postural impairments, espe- needed to replicate these results.
cially those associated with aging including improved balance and Another more rigorous meta-analysis identified 7 randomized
dynamic stability, increased musculoskeletal strength and flexi- controlled trials that met criteria [36]. This time the Jadad scale was
bility, improved performance on activities of daily living, reduced used. This scale is the most rigorous criterion for meta-analysis and
fear of falling, and general improvements in psychological well- includes items that rate randomization, blinding and drop-out/
being. In still another study, increased muscle strength followed withdrawals to determine the quality of the studies. Standard
Tai Chi including the strength of the knee extensors and flexors mean differences (changes) were 45% for pain, 31% for stiffness and
[28]. The control group had less strength than the jogging group 61% for physical function. The authors noted that an average change
and less strength than the Tai Chi group. of 32.2e36.4% in the outcomes was greater than the minimum
More recently Tai Chi has also been studied for pain relief in clinically important difference. As in most of the studies already
older adults with knee osteoarthritis [29,30]. In one elaborate study reviewed, these authors only analyzed WOMAC scores. They found
involving gait kinematics, tai chi training was provided for 1 h twice that most studies lacked objective outcome measures such as ex-
a week for 6 weeks [31]. After 6 weeks of the tai chi exercise the ercise performance and muscle strength which would be more
stride length, the stride frequency and the gait speed increased reliable and robust evidence for the effects of tai chi on knee
significantly and knee pain was decreased. Unfortunately, this trial osteoarthritis. Nonetheless, they found no adverse effects of tai chi
did not have a control or treatment comparison group. and relatively high adherence in most studies. They also pointed
In another knee osteoarthritis study, the 12-form Sun tai chi for out the cost-effectiveness features such as the low cost of sessions
arthritis [32] was adapted for elders with knee osteoarthritis [33]. given that no special setting was required and multiple benefits for
In this randomized trial on a 20-week tai chi program for the the body, suggesting that tai chi might be considered a good
elderly, several assessments were made including the WOMAC, the alternative to other forms of exercise for knee osteoarthritis.
Get-up-to-go test, the Sit-to-stand test and the Geriatric Depression In all of these studies there may have been an overestimation of
scale. The WOMAC was used to measure subjective pain, physical the treatment effects since the samples have been relatively small.
functioning and stiffness. The Get-up-to-go (GTG) measure Another limitation is that the duration of the tai chi sessions has
assessed the speed in getting up to standing from an arm chair and been highly variable as have the targeted samples. For examples,
walking as fast as possible for 50-feet and returning to the chair to the samples have had different durations of osteoarthritis as well as
sit down. The Sit-to-stand (STS) test has the participants cross their different ethnic and cultural backgrounds and wide variability in
arms across their chest and rise to standing 5 times. The attention the treatment protocols on several dimensions including the type
control group participated in health education, culture-related and of tai chi form, the length of the form and the duration of the
other social activities for the same 20-week period. treatment period.
On the primary variable, the WOMAC pain scale, the tai chi One of the real advantages of Tai Chi is that it is a simple,
group score was significantly lower than the control group score convenient workplace intervention that may promote musculo-
starting at 9 weeks, suggesting that an 8-week trial was required to skeletal health without special equipment [37]. For example, in a
show a significantly greater reduction of pain. The scores on the workplace study, female computer users participated in two 50 min
WOMAC physical function and stiffness subscales also decreased Tai Chi classes per week for 12 consecutive weeks. Positive results
more for the tai chi than the attention control group. Although the included reduced heartrate and waist circumference and increased
tai chi group also showed greater improvement on the GTG and the grip strength. The Tai Chi program improved musculoskeletal
STS measures, the differences only tended to be significant. While fitness as well as psychological well-being.
an attention control group was included in this study, that condi- Although Tai Chi has improved balance, gait and strength, some
tion did not include any physical activity, suggesting that it was not studies have the limitations of a small sample size or a limited
a good comparison group. In addition, having two sessions per number of variables. The reviews have also shown that Tai Chi
 T. Field / Complementary Therapies in Clinical Practice 22 (2016) 87e92 91

participants have improved performance on activities of daily living [6] J. Forman, M. Stampfer, G. Curban, Non-narcotic analgesic dose and risk of
herpertension in US women, Hypertension 46 (2005) 500e507.
as well as general psychological well-being [38,39]. Unfortunately,
[7] J. Richmond, D. Hunter, J. Irrgang, M. Jones, L. Snyder-Mackler, et al., Surgeons
Tai Chi has not yet been compared with similar forms of movement clinical practice guideline on the treatment of osteoarthritis (OA) of the knee,
therapy such as yoga. J. Bone Jt. Surg. Am. 92 (2007) 990e993.
[8] M. Eghball, H. Lellahgani, N. Alimohammadi, R. Daryabeigi, Z. Ghasempour,
Study on effect of massage therapy on pain severity in orthopedic patients,
7. Limitations of these complementary studies Iran. J. Nurs. Midwifery Res. 15 (2010) 1e7.
[9] P. Peungsuwan, P. Sermcheep, P. Hammontree, W. Eungpinichpong,
Although many limitations of the literature on massage therapy, R. Puntumetakul, U. Chatchawan, J. Yamauchi, The effectiveness of Thai ex-
ercise with traditional massage on the pain, walking ability and QOL of older
yoga and tai chi with knee osteoarthritis have already been people with knee osteoarthritis: a randomized controlled trial in the com-
mentioned, they are briefly summarized here. Many of the massage munity, J. Phys. Ther. Sci. 26 (2014) 139e144.
therapy studies that showed lesser effects did not use moderate [10] D. Atkins, D. Eichler, The effects of self-massage on osteoarthritis of the knee:
a randomized, controlled trial, Int. J. Ther. Massage Bodyw. 6 (2013) 4e14.
pressure and/or did not include the hamstrings in their massage [11] A. Perlman, A. Ali, V. Njike, D. Hom, A. Davidi, S. Gould-Fogerite, C. Milak,
protocols. The yoga studies were frequently confounded by the D. Katz, Massage therapy for osteoarthritis of the knee: a randomized dose-
several types of yoga that were integrated in the protocols (poses, fnding trial, PLos ONE 7 (2012) 1e9.
[12] T. Field, G. Gonzalez, M. Diego, C.G. Funk, Knee pain is reduced and range of
breathing and meditation), making it difficult to determine the real motion is increased following massage therapy, Complement. Ther. Clin.
treatment effects. The yoga studies also typically focused on pain as Pract. 21 (2015) 233e237.
measured by the WOMAC, a self-report measure that might not be [13] T. Field, M. Diego, M. Hernandez-Reif, C.G. Funk, Hand arthritis pain is reduced
by massage therapy, J. Bodyw. Mov. Ther. 11 (2007) 21e24.
as reliable as more objective measures such as observed ROM-
[14] T. Field, M. Diego, J. Delgado, D. Garcia, C.G. Funk, Rheumatoid arthritis in
related pain. The tai chi studies also primarily focused on pain as upper limbs benefits from moderate pressure massage therapy, Complement.
self-reported on the WOMAC. The tai chi studies were also difficult Ther. Clin. Pract. 18 (2013) 101e103.
[15] T. Field, M. Diego, G. Gonzalez, C.G. Funk, Neck arthritis pain is reduced and
to compare because of their highly variable protocols in terms of
range of motion is increased by massage therapy, Complement. Ther. Clin.
frequency and duration of treatment. Pract. 10 (2014) 219e223.
[16] R. Melzack, P.D. Wall, Pain mechanisms: a new theory, Science 150 (1965)
971e979.
8. Future directions
[17] K. Roland, J. Jakobi, G. Joes, Does yoga engender fitness in older sdults? A
critical review, J. Aging Phys. Act. 19 (2011) 62e79.
Larger, randomized control trials are needed on each of these [18] R. Ranjitha, & R. Nagarathna, Effect of yoga on pain, mobility, gait and balance
therapies using more standardized protocols and more objective in patients with osteoarthritis of the knee.
[19] S. Kowalinski, M. Garfinkel, A. Tsai, Iyengar Yoga for treating symptoms of
variables in addition to the self-reported WOMAC pain scale, for osteoarthritis of the knees: a pilot study, J. Altern. Complement. Med. 11
example, range-of-motion and observed range-of-motion pain. (2005) 689e693.
Further, treatment comparison studies should be conducted. So, for [20] G. Nambi, A. Shah, Additional effect of Iyengar yoga and EMG biofeedback on
pain and functional disability in chronic unilateral knee osteoarthritis, Int. J.
example, if the lower-cost yoga and tai chi were as effective as Yoga 6 (2013) 123e127.
massage therapy, they might be used in combination with or as [21] N. Patel, A. Newstead, R. Ferrer, The effects of yoga on physical functioning
supplemental to massage therapy. Nonetheless, the studies thus far and health related quality of life in older adults: a systematic review and
meta-analysis, J. Altern. Complement. Med. 18 (2012) 902e917.
suggest that these therapies are at least reducing pain in knee [22] M. Ferrari, S. Thuraisingam, R. Kanel, & N. Egloff, Expectations and effects of a
osteoarthritis and they do not seem to have undesirable side ef- single yoga session on pain perception. Int. J. Yoga, 8, 154e157.
fects. In addition, for most of the studies, the reduction in pain was [23] T. Field, Tai Chi Research Review, Complement. Ther. Clin. Pract. 17 (2011)
141e146.
clinically significant and compliance/adherence rates were high,
[24] A. Ramachandran, K. Rosengren, Y. Yang, E. Hsiao-Wecksler, Effect of Tai Chi
suggesting that these therapies may be pursued by older adults on gait and obstacle crossing behaviors in middle-aged adults, Gait Posture.
with knee osteoarthritis even without being prescribed. Even 26 (2006) 248e255.
[25] G. Wu, Age-related differences in Tai Chi gait kinematics and leg muscle
though most of the studies focused exclusively on pain measures,
electromyography: a pilot study, Archives Phys. Med. Rehabil. 89 (2008)
undoubtedly other positive effects that have been reported for 351e357.
other conditions also happen for knee osteoarthritis following [26] S.M. Fong, G.Y. Ng, The effects on sensorimotor performance and balance with
massage therapy, yoga and tai chi. Tai Chi training, Arch. Phys. Med. Rehabil. 87 (2006) 82e97.
[27] D. Xu, Y. Hong, K. Chang, Effects of Tai Chi exercise on proprioception of ankle
and knee joints in old people, Br. J. Sports Med. 38 (2004) 50e54.
Acknowledgments [28] D. Xu, J. Li, Y. Hong, Effects of long term Tai Chi practice and jogging exercise
on muscle strength and endurance in older people, Br. J. Sports Med. 40
(2006) 50e54.
I would like to thank the many adults who participated in our [29] C. Wang, C. Schmid, P. Hibbard, et al., Tai Chi is effective in treating knee
knee osteoarthritis and several other arthritis studies. Further osteoarthritis: a randomized controlled trial, Arthritis Rheum. 61 (2009)
thanks go to the massage therapists, mostly Gladys Gonzalez, who 1545e1553.
[30] H. Lee, H. Park, Y. Chae, et al., Ta Chi Qigong for the quality of life of patients
designed the therapy protocols and the therapy sessions and to with knee osteoarthritis: a pilot, randomized, waiting list controlled trial, Clin.
Miguel Diego, who conducted the data analyses. I would also like to Rehabil. 23 (2009) 504e511.
thank Massage Envy for funding our arthritis studies. [31] C. Shen, C. James, M. Chyu, W. Bixby, J. Brismee, M. Zumwalt, G. Poklikuha,
Effects of Tai Chi on Gait Kinematics, Physical Function, and Pain in Elderly
with Knee Osteoarthritis-a Pilot Study, vol. 36, 2008, pp. 219e222.
References [32] Arthritis Foundation, Tai Chi Program, December 9, 2010. Available from,
http://arthritis.org/events/getinvolved/ProgramsServices/TaiChi.asp.
[1] J. Ebnezar, R. Nagarathna, R. Yogitha, H.R. Nagendra, Effects of integrated yoga [33] P. Tsai, J. Change, C. Beck, Y. Kuo, F. Keefe, A pilot cluster randomized trial of a
therapy on pain, morning stiffness and anxiety in osteoarthritis of the knee 20-week Tai Chi program in elders with cognitive impairment and osteoar-
joint: a randomized control study, Int. J. Yoga 5 (2012) 28e36. thritic knee: effects on pain and other health outcomes, J. Pain Symptom
[2] J. Ye, S. Caj, W. Zhong, S. Caj, Q. Aheng, Effects of tai chi for patients with knee Manag. 45 (2013) 660e669.
osteoarthritis: a systematic review, J. Phys. Ther. Sci. 26 (2014) 1132e1137. [34] J. Ye, S. Caj, W. Zhong, S. Caj, Q. Zheng, Effects of Tai Chi for patients with knee
[3] P. Tsai, C. Beck, J. Chang, et al., The feasibility of implementing Tai Chi for osteoarthritis: A systematic review, J. Phys. Ther. Sci. 26 (2014) 1133e1137.
nursing home residents with knee osteoarthritis and cognitive impairment, [35] R. Lauche, J. Langhorst, G. Dobos, H. Cramer, A systematic review and meta-
Activities Directors Q. Alzheimers Other Dementia Patients 10 (2009) 9e17. analysis of Tai Chi for osteoarthritis of the knee, Complement. Ther. Med. 21
[4] U. Nguyen, Y. Zhang, Zhu, et al., Increasing prevalence of knee pain and (2013) 396e406.
symptomatic knee osteoarthritis: survey and cohort data, Ann. Intern. Med. [36] J. Yan, W. Gu, J. Sun, W. Zhang, B. Li, L. Pan, Efficacy of Tai Chi on pain, stiffness
155 (2011) 725e732. and function in paients with osteoarthritis: a meta-analysis, PLos One 8
[5] T. Spector, A. MacGregor, Risk factors for osteoarthritis, J. Osteoarthr. Cartil. 12 (2013) e6i71.
(2004) 39e44. [37] H. Tamim, E.S. Castel, V. Jamnik, P.J. Keir, S.L. Grace, N. Gledhill, et al., Tai Chi
92 T. Field / Complementary Therapies in Clinical Practice 22 (2016) 87e92

workplace program for improving musculoskeletal fitness among female Arthritis Rheumatism 61 (2009) 1545e1553.
computer users, Work 34 (2009) 331e338. [39] P. Wayne, D. Krebs, S. Wolf, K. Gill-Body, D. Scarborough, C. McGibbon, et al.,
[38] C. Wang, C.H. Schmid, P.L. Hibberd, R. Kalish, R. Roubenoff, R. Rones, et al., Tai Can Tai Chi improve vestibulopathic postural control? Arch. Phys. Med.
Chi is effective in treating knee osteoarthritis: a randomized controlled trial, Rehabil. 85 (2004) 142e152.