Rheumatoid Arthritis



Rheumatoid arthritis (RA) is an autoimmune disease that affects the whole body but is most obvious in the joints. Most people are adults when they become aware they have rheumatoid arthritis although young people can have a juvenile form of rheumatoid arthritis.

Inflammation and degeneration of the joints results in pain, stiffness, and swelling of joints with serious fatigue that can vary from hour to hour and day to day. The joints are stiff and painful so that movement is restricted. This means that a person with arthritis can lose muscle strength and has difficulties with daily activities like grooming and dressing, cooking a meal, cleaning, shopping, and applying themselves to their work, family, social and leisure activities. Rheumatoid arthritis is often progressive with an unpredictable course of exacerbations and remissions. The joints can become deformed.

Medications used range from non-steroidal anti-inflammatory drugs such as ibuprofen; corticosteroid injections into a joint (intra-articular injection); disease modifying immunosuppressants such as methotrexate; to the newer TNF alpha (tumour necrosis factor) antagonists and rituximab. The effectiveness and severity of adverse side effects of these interventions varies for different individuals.

We need to be able to bend (flex) the knees to carry out a number of daily tasks; a minimum of 65 degrees to walk normally, 90 degrees to go down stairs one leg after the other and at least 105 degrees to get up from the toilet or a low chair. Seating in close rows or sitting on public transport can present problems without the ability to bend the knees some 90 degrees.  


Acupuncture can decrease pain, modulate stress and support internal equilibrium (homeostasis) or balance and is used by some rehabilitation specialists as part of therapy for the symptomatic treatment of rheumatoid arthritis.

Acupuncture is a traditional technique used in Chinese medicine for many medical conditions. Thin acupuncture needles are inserted into the skin at set points (meridians) to tap into the streams of body energies. Once the needles are in position they are either manipulated with the fingers or activated with a small electrical current, which is known as electroacupuncture.

What the synthesised research says

In one small study, electroacupuncture to the knee reduced symptomatic knee pain for people with RA. Pain relief was measured 24 hours after treatment and a small reduction in pain was still evident four months after treatment

In another study, traditional acupuncture to one point for general health (the liver point on each foot) did not improve disease state, pain, number of affected joints, general health or analgesic intake compared with sham treatment.

How it was tested

The researchers made a thorough search of the medical literature and found two controlled trials that randomised a total of 84 people to active acupuncture or sham treatment. One study used traditional acupuncture and the other used electroacupuncture.

Acupuncture with manipulation of the needles

The one trial using traditional acupuncture involved 64 people aged between 46 and 66 years who had RA; 8 of whom withdrew before treatment began. The trial did not find any benefit of acupuncture over sham treatment on disease state, pain, number of affected joints, general health or analgesic intake. The therapist applied acupuncture at a single specific documented point believed to affect overall health; one needle (0.25 x 30 mm) was inserted at the liver 3 (Li3) point found on each foot for four minutes. The needles were manipulated for five seconds two minutes after being inserted. Sham treatment involved using the guides but not inserting the needles. The participants received five sessions at one week intervals.


The one trial involving 20 patients that showed positive results for the use of electroacupuncture on symptomatic RA knees applied electroacupuncture to three specifically documented points around the knee or to three incorrect points. The people who received active electroacupuncture reported a significant reduction in pain 24 hours after the treatment and persisting for most of the ten people for three months.

Electroacupuncture at 6.26 5mA was applied once for 15 minutes using three 1.5 cm long needles. The needles were inserted in one of the knees at the GB 34, SP 9 and S43 acupuncture points. The other knee was injected with 50 mg of hydrocortisone. In the placebo group, electroacupuncture was applied to three incorrect points around the knee using the same instrumentation as the experimental group. The other knee was also injected with 50 mg of hydrocortisone.

Side effects and general cautions

This review is limited by few adequate studies providing evidence for the use of acupuncture. The conclusions are limited by methodological considerations such as the type of acupuncture (acupuncture versus electroacupuncture), the site of intervention, the low number of clinical trials and the small number of participants and reporting of outcome of therapy in the included studies.

In the study using traditional acupuncture the acupuncture technique was questionable. In the other study, of electroacpuncture, giving an injection of hydrocortisone to the other knee could have affected the results.


L Casimiro, L Barnsley, L Brosseau, S Milne, VA Robinson, P Tugwell, G Wells. Acupuncture and electroacupuncture for the treatment of rheumatoid arthritis. Cochrane Database of Systematic Reviews 2005, Issue 4. Art. No.: CD003788.pub2. DOI: 10.1002/14651858.CD003788.pub2.


Conventional therapies do not always give full relief of symptoms and the drugs used can cause troublesome side effects.

Herbal therapy, or phytotherapy, is a traditional approach to controlling symptoms. These are measured as level of pain on a scale of 1 to 10 or 100, number of tender or swollen body joints, morning stiffness, grip strength and the time to walk 15 metres. The Chinese herb Thunder God vine (Radix Tripterygium wilfordii hook F) has been used for rheumatoid arthritis for centuries. In Europe herbs used include devil's claw (Harpagophytum species), Salix species and plant seed oils (such as evening primrose, borage). Evening primrose oil, black currant seed oil and borage seed oil all contain the fatty acid nutrient gamma-linolenic acid (GLA). Herbs can cause adverse effects, making whether or not they are effective important.

Individualised treatment is fundamental to many forms of complementary therapy, which makes it difficult to have adequately control clinical trials, although many herbs are now available as standardised over-the-counter products.

What the synthesised research says

Evening primrose oil, black currant seed oil and borage seed oil, sources of gamma-linolenic acid, may be of benefit for people with rheumatoid arthritis. This conclusion was made after bringing together the results of seven controlled studies. In the better quality studies gamma-linolenic acid reduced pain, joint tenderness and morning stiffness compared with non-active placebo.

A single controlled study is considered to be insufficient evidence to base therapy on yet this was all that was available for feverfew, Tripterygium wilfordii hook F (T2), capsaicin and Reumalex, which is an over-the-counter preparation that contains willow bark. Single studies of feverfew, Tripterygium wilfordii hook F (T2), capsaicin and Reumalex, an over-the-counter preparation that contains willow bark, are inconclusive.

How it was tested

The review authors found a total of 11 studies that met set criteria, reported on between 1988 and 1996. Seven of the studies, with a total of 286 participants, compared supplements of gamma-linolenic acid (GLA) to olive oil, sunflower oil, liquid paraffin, cottonseed oil or soybean oil. These oils were used as non-active placebo, although olive oil has been reported to ease some symptoms by some authors. Daily doses of GLA ranged from 525 mg to 2.8 g, taken for 6 weeks to 12 months.

People on GLA showed a strong trend for a greater reduction in pain compared to taking placebo (WMD -33, range -56 to -9, 3 trials on a 100 mm visual analogue score). Joint tenderness was reduced with GLA (WMD in count -37%, range -56 to -19%, 3 trials), joint swelling less so. Duration of morning stiffness decreased more in the participants taking GLA. Self evaluation of overall disease activity tended to favour GLA treatment.

A study that used olive oil as placebo and evening primrose oil as the source of GLA, reported a reduction in morning stiffness with evening primrose oil but no change in pain or wellbeing. This compared to clear reductions in articular index and pain, a trend to reduction in morning stiffness and no changes in wellbeing in the olive oil group. Possible therapeutic effects of olive oil have been observed by other authors.

For evening primrose oil, over 90% of people in one study reported an overall subjective improvement compared to a little over 40% in the placebo group. This was without any clear changes of morning stiffness, grip strength or pain.

One study comparing black currant seed oil to placebo reported a clear reduction in morning stiffness and improvement in wellbeing with a trend to improvements in other clinical outcomes.

In separate single studies, Tripterygium Wilfordii Hook F (60 mg daily for 12 weeks) and topical capsaicin (0,025% 4 times a day for 4 weeks) were found to produce clear improvements in arthritic symptoms.  In a trial where 31 people were randomly assigned to capsaicin or placebo people on capsaicin had greater pain relief and physicians also tended to assess them as doing better.

Tripterygium Wilfordii Hook F in a study which had 70 participants reported benefits on joint tenderness and swelling, morning stiffness, grip strength and the time to walk 15 metres.

Reumalex (2 tablets at a time for 2 months) was tested in one trial involving 20 people with rheumatoid arthritis and 52 with osteoarthritis and was effective in improving arthritis pain score compared with placebo. Joint tenderness was also reduced but people did not reduce their intake of painkillers (analgesics).

Feverfew (70 to 86 mg daily for 6 weeks) was not found to have any significant benefits for people with rheumatoid arthritis when compared with taking placebo in a single trial involving 41 participants.

Side effects and general cautions

The review was limited by few adequate studies to provide evidence of benefit. Methodology and study quality was variable and trials suffered from poor reporting of methods and data.

The review authors included any whole plant extract but excluded homeopathy, aromatherapy or any synthetic preparation or a herb combined with a non-herbal substance.

Optimum dose and duration of treatment with GLA is uncertain. It is suggested that effects on joint synovitis are not expected to occur for some weeks. Three studies showing fewer clinical responses provided a relatively low daily dosage of GLA (between 525 mg and 540 mg) for a duration of only 6 to 12 weeks. By comparison, three studies that used doses between 1.4 g and 2.8 g GLA for a duration of 24 weeks reported greater improvements in symptoms of arthritis. In one trial people who went on with GLA for a second six months continued to show improvement.

One problem seen in studies where larger doses of GLA were given was the large size and quantity of capsules that had to be taken.

Asking people with rheumatoid arthritis to stop taking NSAIDs both before and during a study may influence the findings.

Five of the GLA studies reported adverse reactions and many of these were related to digestive upset.

Adverse reactions were seen mainly in patients receiving Tripterygium Wilfordii Hook F and resulted in four withdrawals from the study. Severe reaction (fever and aplastic anaemia) occurred in one patient following an overdose of this herbal therapy.

The only adverse reaction attributable to capsaicin was a burning sensation at the site of application, which occurred in 44% of participants receiving capsaicin.

With Reumalex, four participants in each group withdrew due to side effects, which were relatively minor. Five additional patients (two in Reumalex group and three in placebo group) reported exacerbation of arthritis.

For feverfew, one participant in each treatment group reported mild side effects resulting in one withdrawal from the study (from the placebo group).


CV Little, T Parsons. Herbal therapy for treating rheumatoid arthritis. Cochrane Database of Systematic Reviews 2000, Issue 4. Art. No.: CD002948. DOI: 10.1002/14651858.CD002948.


Spa therapy traditionally involves bathing in thermal or mineral waters including sea water baths. It is also known as balneotherapy (from the Latin) as the Romans used water for the therapeutic treatment of bone and joint conditions.

Spa therapy is soothing and may help people living with constant pain to cope.

Hydrotherapy is where physical therapists provide exercise programs in heated pools to improve the range of joint motion, improve muscle strength, relieve muscle spasm and maintain or improve functional mobility. The water is warmed to a temperature around 34 degrees Celsius and supports the weight of the body. The painful and rheumatic joints can then be moved more freely. 

What the synthesised research says

In most of the trials the participants reported improvements after therapy in the short term. Information on pain was limited. 
No clear benefits were reported after three months.

How it was tested

This conclusion is based on six controlled trials that involved a total of 355 people with rheumatoid arthritis. Participants were aged from 49 to 62 years. Four trials were carried out in Israel, one in Germany and one in the UK. People were randomly assigned to spa therapy sessions, about 20 minutes long, in water containing Dead Sea salts, sulphur, tap water or radon-carbon dioxide or the comparative treatment of baths containing salt (sodium chloride), tap water or carbon dioxide; no treatment; relaxation or exercise therapy (hydrotherapy and land exercises). Sessions continued over two to four weeks. The participants were followed up at three months (and six months in one trial).

Bathing in Dead Sea salts with sulphur, Dead Sea salts or sulphur alone improved morning stiffness, walking time over 15 metres, hand grip strength, number of joints with inflammation (active joints), activities of daily living, patient self assessment after therapy.

For other types of baths (three studies) the benefits were less clear but the one high quality study found that radon-carbon dioxide baths reduced pain more than a carbon dioxide bath (RR 2.3; range 1.1 to 4.7).

Other studies also reported short term improvements after treatment with: sulphur baths, mudpacks, sulphur baths and mudpacks; seated immersion in tap water at 36 degrees Celsius, hydrotherapy (exercises in water), land exercises or relaxation therapy and compared with control groups that did not receive any treatment (two studies).

Side effects and general limitations

Two trials reported on participants who had mild heat reactions during therapy.

Information on pain relief was limited. Four studies reported a patient self-assessment of the severity of disease with a clear improvement, before and after therapy. Pooling of the data was not performed and multiple outcomes, including pain and function, were measured. Information about study design was lacking or unclear in four studies. A number of the trials had methodological limitations and only two had a control group that did not receive additional treatment.

It is possible that the way that the effects of spa therapy were measured was not sensitive or responsive enough to be able to measure any treatment effect. Also the therapy period was short. The spa ‘environment' is an important part of the therapy, and the time taken for oneself can provide benefits in itself.

Several studies were reported on in Hebrew, Japanese or an Eastern European language and are not included in the review as yet.


AP Verhagen, SMA Bierma-Zeinstra, JR Cardoso, RA de Bie, M Boers, HCW de Vet. Balneotherapy for rheumatoid arthritis. Cochrane Database of Systematic Reviews 2004, Issue 1. Art. No.: CD000518. DOI: 10.1002/14651858.CD000518.


Thermotherapy uses heat (hot packs, paraffin wax baths for the hands) or cold (ice or cold gel packs) to relieve arthritis symptoms including pain, muscle spasms, puffy swelling and poor circulation. They can be applied at home and are also used in physical rehabilitation centres together with exercises.

What the synthesised research says

The limited evidence from randomised controlled trials was unable to show any overall benefit of heat or cold therapy on arthritic hands - joint swelling, movement, pain and hand function for adults with rheumatoid arthritis.

Paraffin wax baths may help exercises to improve hand range of movement and pain during non-resisted movement (one trial only, 4 weeks of exercise with baths did show greater improvements in hand function but paraffin baths alone were not effective).

Cold therapy tended to reduce swelling after hand surgery (one trial only, 2 and 3 days after surgery).

No harms were reported in the included trials.

How it was tested

These conclusions are based on seven controlled trials that involved 328 adults with rheumatoid arthritis. The trials looked at ice therapy (2 trials), paraffin wax or faradic baths, different temperatures (1 trial) and heat versus cold.

Thermotherapy given in combination with another treatment: strengthening exercises (two trials), ultrasound (one trial) or medication (1 trial) had no clear benefits on objective measures of disease activity including joint swelling and tenderness, pain, medication intake, range of movement, grip strength or hand function compared to no active treatment or an alternative treatment.

A faradic bath uses a weak electrical current that produces a sensory effect as muscles start to contract (one trial).

Side effects and general cautions

No harmful effects were reported in these trials.

Thermotherapy is used as a routine part of physiotherapy practice, combined with exercises, therapeutic ultrasound or medication, which makes it difficult to assess the benefits of thermotherapy alone. The setting, length and frequency of use of thermotherapy may also be important.

The review is limited by few adequate studies that provide evidence. The trials were reported on between 1968 and 1994 Conclusions are limited by the poor methodological quality of the trials available, the small number of participants involved and the large number of unclear results.


VA Robinson, L Brosseau, L Casimiro, MG Judd, BJ Shea, P Tugwell, G Wells. Thermotherapy for treating rheumatoid arthritis. Cochrane Database of Systematic Reviews 2002, Issue 2. Art. No.: CD002826. DOI: 10.1002/14651858.CD002826.


In rheumatoid arthritis, the small joints of the hands and feet can become tender, swollen, stiff and disfigured.

People sometimes use TENS (transcutaneous electrical nerve stimulation), in addition to their medication, to give pain relief and to help them to carry out their daily activities. This involves wearing a portable unit with electrodes placed on a painful area, for some 30 minutes at a time. Pain relief is almost immediate but generally stops as soon as the TENS is turned off, although some people report more prolonged pain relief.

The electrodes can either be placed on painful areas, over nerve pathways or at acupuncture and trigger points. The level of stimulation can also be varied (with high and low frequency electrical stimulation rates and different intensities.

Conventional TENS (C-TENS) uses a high stimulation frequency (40 to 150 Hz), low intensity

Acupuncture-like TENS (AL-TENS) is given at a low frequency (1 to 10 Hz), high intensity, close to the patient's limit of tolerance.

Burst TENS uses high frequency (100 Hz) burst impulses at low intensity (1 to 2 Hz) and causes rhythmic muscle contractions.

What the synthesised research says

Applying acupuncture-like TENS to the hands reduces pain and may also improve muscle power compared with sham treatment (placebo control).

Conventional TENS had no clear benefit on pain compared with placebo. 

How it was tested

This conclusion is based on three controlled trials that randomised a total of 78 people, more women than men, with rheumatoid arthritis affecting their hands to receive TENs or sham treatment. The people’s ages ranged from 18 to 72 years.

Acupuncture-like TENS and conventional TENS were compared to placebo and to each other. These trials were reported on between 1978 and 1984.

In one study, applying acupuncture-like TENS for 15 minutes a week over three weeks, resulted in a significant decrease in rest pain (45 points absolute benefit on a 100 mm VAS scale) compared with sham treatment but without improvements in pain when gripping.

In the second study, a single 20-minute session of conventional TENS or acupuncture-like TENS did not reduce pain intensity (resting or gripping the hand) compared with each other or conventional TENS with sham treatment. Joint tenderness scores  did improve with conventional TENS compared with sham treatment.

The third study compared three different TENS applications given for 5 minutes, once a day, for 15 days: acupuncture-like TENS applied at the wrist under study, conventional TENS applied at the wrist under study or between the shoulder-blades, on either side of the spinal processes of the back. Conventional TENS and acupuncture-like TENS were not clearly different in relieving joint pain.

Participants did report a greater change in disease with conventional TENS over acupuncture-like TENS (Number needed to treat 5).

Side effects and general cautions

No adverse events were reported with TENS, which is considered to rarely cause adverse effects The different ways in which TENS was used (15 minutes once per week for 3 weeks; 20 minutes once only; daily application for 5 minutes for 15 consecutive days using either acupuncture-like or conventional TENS makes it hard to compare studies. The study participants had rheumatoid arthritis for from 1 to 44 years.

The methodological quality of the trials was poor, were relatively small all carries out before 1985.



L Brosseau, KA Yonge, V Robinson, S Marchand, M Judd, G Wells, P Tugwell. Transcutaneous electrical nerve stimulation (TENS) for the treatment of rheumatoid arthritis in the hand. Cochrane Database of Systematic Reviews 2003, Issue 2. Art. No.: CD004377. DOI: 10.1002/14651858.CD004377.


People with rheumatoid arthritis are looking to improve muscle performance and reduce the loss of joint mobility and function.

Yet muscles have to contract strongly to improve muscle strength and resistance to fatigue, which is difficult for people for people with painful and inflamed joints. Therapists in rehabilitation centres can apply added electrical stimulation to increase the number of motor units contributing to muscle contraction, and therefore the strength of its activity.

What the synthesised research says

Electrical stimulation applied to muscles of the hand improved grip strength and fatigue resistance for people with wasted muscles of the hand because of rheumatoid arthritis. More favourable results were obtained using a patterned stimulation mimicking that from a fatigued motor unit of a normal hand rather than with a fixed 10 Hz stimulation frequency of stimulation.

How it was tested

The researchers made a thorough search of the medical literature and found only one controlled trial that involved 15 adults aged 30 to 75 years with rheumatoid arthritis affecting their hands.

This trial looked at the first joint of the index finger of the dominant (right) hand and measured sprain, deformity (partial dislocation and deviation) and muscle wastage (disuse atrophy). The trial participants were randomly assigned to an untreated control group (3 participants) or to receive electrical stimulation over the first joint of the index finger for 70 sessions over 10 weeks (168 hours in total) (6 participants with each type of stimulation).

Patterned electrical stimulation strengthened hand strength, grip, pushing a button (button test) and deformity, and pinch strength and the isometric force exerted by the index finger of the dominant hand and fatigue resistance during continuing maximum voluntary contraction).

A 10 Hz stimulation also increased pinch strength and fatigue resistance of the first joint of the index finger during maximum contraction but less so.

Side effects and general cautions

Side effects of electrical stimulation were not reported, nor how well the participants kept to their treatment schedules. Conclusions are limited by the low methodological quality of the single trial found and the small number of participants in that trial. It is difficult to base practice on the results of only one trial designed as a randomised controlled trial to minimise biases.


L Pelland, L Brosseau, L Casimiro, VA Robinson, P Tugwell, G Wells. Electrical stimulation for the treatment of rheumatoid arthritis.Cochrane Database of Systematic Reviews 2002, Issue 2. Art. No.: CD003687. DOI: 10.1002/14651858.CD003687.


Low-level laser therapy used as one of a range of physical therapies in the treatment of rheumatoid arthritis. Light of a single wavelength is applied to the skin. The effect is not thermal, but rather triggers responses in the cells. It has been reported to stimulate cartilage, reduce inflammation and relieve pain.

There are different types of the therapy (Classes I, II, and III) with wavelengths varying from 632nm to 1064 nm. Characteristics of treatment (e.g. low versus high dose, wavelength, nerve versus joint application, and treatment duration) and the actual device used can all differ and may influence the effectiveness of treatment.

Low level laser therapy may positively modify the sensory input to the central nervous system and provide an improvement in the perception of pain localized to the area of treatment, which would mean that it is most effective when nerves as well as joints are treated.  

What the synthesised research says

Low level laser therapy applied locally to the hands or thumb or the knee for at least four weeks helped reduce morning stiffness and relieve pain. 

How it was tested

This conclusion is based on six controlled trials involving a total of 222 adult people with rheumatoid arthritis. Five were randomised controlled trials of laser therapy versus sham treatment in a separate control group (204 participants). Treatment was generally given in 2 to 3 sessions per week for 3 to 4 weeks (10 weeks in one study). Relative to sham therapy, low level laser reduced pain by 1.1 points on a visual analogue scale (1 to 10 scale) (range 1.8 to 0.4, 3 trials), reduced morning stiffness duration by 27.5 minutes (range 3 to 52 minutes, 3 trials) and increased finger or thumb tip to palm flexibility by 1.3 cm (range 0.8 to 1.7, 2 trials). Other outcomes such as functional assessment, grip strength, range of motion (hand and fingers, ankle) and local swelling did not clearly differ between treatment groups.

One trial found a positive effect of laser on knee range of motion after treatment (WMD 32, range 4 to 32), but not for the ankle.

Two trials followed up patients three months after stopping therapy and were unable to detect any long-term benefit.

In the single study of 18 people aged 53 to 67 years that applied treatment to one hand and used the other hand to compare (as control), laser therapy did not improve pain or the duration of stiffness in the mornings (baseline morning stiffness ranged from 60 to 90 minutes).

Side effects and general cautions

No side effects were reported. Only one study applied low level laser to the nerves as well, with a trend towards greater pain reduction when nerves as well as joints were irradiated.

The major limitation of this review is the variety of ways laser therapy was applied, including different dosages and wavelengths. Treatment was to the hands in all but one of the trials.


L Brosseau, V Robinson, G Wells, R deBie, A Gam, K Harman, M Morin, B Shea, P Tugwell. Low level laser therapy (Classes I, II and III) for treating rheumatoid arthritis. Cochrane Database of Systematic Reviews 2005, Issue 4. Art. No.: CD002049.pub2. DOI: 10.1002/14651858.CD002049.pub2.


Therapeutic ultrasound involves the application of high frequency mechanical vibrations, continuous or pulsed, to the affected area. Pulsed ultrasound is used to reduce inflammation during a flare of inflammation. Continuous ultrasound generates heating (thermal) effects and is used to relieve pain, reduce muscle spasms and increase blood flow to help flush out toxins. Therapeutic ultrasound that penetrates deeply enough to increase connective tissue (collagen) elasticity may also be useful in a flexibility program early in the disease process.

What the synthesised research says

Therapeutic ultrasound to the hands improved grip strength in one trial; with borderline effects for improving wrist dorsal flexion, morning stiffness and number of swollen or painful joints.

Ultrasound in combination with exercises, faradic current and wax bath treatment modalities was not supported by findings from one controlled trial.

How it was tested

The researchers made a thorough search of the medical literature and found two controlled trials involving a total of 80 participants randomly assigned to ultrasound or sham therapy.

In the one study that gave positive results, the ultrasound was applied in water to the dorsal and palmar aspects of the hand, at 0.05 W/cm2 continuously using a circular round head. The treatment regimen lasted ten minutes, given on alternate days for three weeks, a total of 10 sessions.

The second trial found no clear difference using ultrasound on pain score, grip strength, circumference of finger joints, articular index, range of motion, or level of activity when comparing three groups of treatment interventions: a) exercises and wax bath, b) exercises with ultrasound, c) exercises with ultrasound and faradic (electric current) hand baths. The ultrasound was applied in water to the palm of the hand, at 0.250 W/cm2 with a constant beam of 3 MHz for three minutes, five times a week for three weeks. Combined therapy reflects common clinical practice in physiotherapy.

Side effects and general cautions

Continuous ultrasound waves were used in both included trials. No harmful side effects were reported. These trials looked at treatment of the hand, with no trials available for other joints.

Conclusions are limited by the poor quality of the two included trials with small numbers of participants.


L Casimiro, L Brosseau, V Robinson, S Milne, M Judd, G Wells, P Tugwell, B Shea. Therapeutic ultrasound for the treatment of rheumatoid arthritis. Cochrane Database of Systematic Reviews 2002, Issue 3. Art. No.: CD003787. DOI: 10.1002/14651858.CD003787.


Painfully inflamed and stiff joints interfere with balance, joint stability, physical activities and work. Exercise therapy is aimed at either preserving or restoring overall functional ability by improving joint mobility, muscle strength, endurance and physical (aerobic) capacity. While range of movement and non weight-bearing isometric exercises put little stress on the joints, dynamic exercises are needed to improve muscle strength and physical capacity. Such exercise programs aim to increase heart rate for 20 minutes or more at least twice a week over a minimum of six weeks.

Note. Isometric exercises work specific muscles while avoiding joint movement. Isotonic exercises contract muscles in a way that does cause joint movement.

What the synthesised research says

Dynamic exercise therapy improved physical capacity in terms of aerobic capacity (five trials), muscle strength (from two trials) and joint mobility.

Exercise therapy over up to three months resulted in a small improvement in daily functional ability, from four trials. The consistency about the safety of vigorous exercises for people with rheumatoid arthritis was striking.

How it was tested

These conclusions are based on six controlled trials each with 18 to 100 people aged round 50 years, where stated, randomised to an exercise program or no exercise (four trials), range of motion and isometric exercises (two trials) as control. Generally people had non-active to moderate active disease and mildly restricted daily functioning. One trial specifically evaluated the effect of exercise therapy in elderly patients with a history of corticosteroid treatment. Muscle strength deteriorated less in the dynamic exercise group than in the control group (3% compared with 28%).

Four studies included dynamic, full weight bearing exercises in the exercise program. Bicycle exercises were most frequently used for fitness (conditioning); walking or pool exercises were also used. Exercise programs lasted for up to three months except in one study where the program was for two years, without clear results on functional ability.

The methodological quality of the trials was moderate to good.

None of the trials reported negative effects of dynamic exercise therapy on the inflammation of the joints or on the level of the acute phase reactants and pain was no different between dynamic exercise and control groups. In two studies the number of clinically active joints was statistically significant decreased within the dynamic exercise group.

Side effects and general cautions

There is no evidence of damaging effects of dynamic exercise therapy for 12 weeks or less - on pain, joint inflammation or disease activity.

Activity level, fatigue, work capacity and endurance were not included in the questionnaires used to measure functional activity in the included studies.

The effects on damaged joints need to be monitored. Only one study evaluated the radiological progression of damage to the joints in the short time period of these studies.

The results of the different trials could not be pulled together (pooled) because of the different ways in which the results (outcomes) were measured and presented.


CHM Van den Ende, TPM Vliet Vlieland, M Munneke, JMW Hazes. Dynamic exercise therapy for treating rheumatoid arthritis. Cochrane Database of Systematic Reviews 1998, Issue 4. Art. No.: CD000322. DOI: 10.1002/14651858.CD000322.


People with rheumatoid arthritis lose muscle strength and range of movement, made worse by the pain, inflammation and stiffness of joints they live with. Exercise therapy sets out to help with flexibility, balance, strength, endurance, aerobic capacity.  

Tai Chi is a traditional Chinese martial art modelled on a fight between a crane and a snake that combines relaxation technique with health-promoting exercise. It works on the understanding that good health is dependent on the body's vital energy, the Chi, circulating freely. Tai Chi uses deep diaphragmatic breathing and slow gentle movements while holding good posture. It has low impact on the joints but demands a considerable amount of work by the leg muscles because the moves (forms) are completed with bent knees in a squat-like position. As it is weight bearing, it has the potential to stimulate bone formation and strengthen connective tissue.

What the synthesised research says

Tai Chi clearly improved the ability to bend the joints of the lower extremities (range of motion), in particular the ankle. Shoulder and upper extremity ranges of motion were not clearly different.

Practising Tai Chi did not worsen arthritic symptoms.

How it was tested

These conclusions are based on four randomised trials that involved 206 adults aged 16 to 80 years. Onetrial was from China and three from the USA. The Tai Chi programs lasted from 8 to 10 weeks.

Tai Chi-based exercise programs had no clear effects on activities of daily living, tender and swollen joints or patient-assessed overall rating.

One study found that a Tai Chi dance program was more enjoyable and led to higher participation than conventional range of movement and rest programs, assessed both immediately and 4 months after completion of the 8-week Tai Chi program. Overall, more people withdraw from the control groups than the Tai Chi programs; 11 out of 101 (11%) for the Tai Chi groups and 25 out of 88 (28%) for the control groups.

Side effects and general cautions

Improvements in lower extremity joint range of movement did not always lead to better functioning in daily activities.

In terms of safety, there were no detrimental effects of Tai Chi on clinical disease activity. The authors of two trials reported that approximately one-third of the participants complained of soreness in the knee, shoulder or lower back during the first three weeks of the studies. Their pain did subside and they were able to continue the program; except for one person with knee pain.

The methodological quality of the trials was low and it is not possible to blind people to the fact they are participating in an exercise program.


A Han, MG Judd, VA Robinson, W Taixiang, P Tugwell, G Wells. Tai chi for treating rheumatoid arthritis. Cochrane Database of Systematic Reviews 2004, Issue 3. Art. No.: CD004849. DOI: 10.1002/14651858.CD004849.


Splints and orthoses are often recommended to people with rheumatoid arthritis as a way of decreasing pain and swelling, to provide joint stability with fewer muscle spasms and to prevent deformity. They support and cushion, align, position, immobilise, prevent or correct deformity, assist weak muscles, or improve function.  Devices include resting hand splints, rigid or flexible wrist supports, static or dynamic finger splints, and special shoes and shoe inserts for the feet.

Splints and insoles (orthoses) are generally low cost but require skilled fabrication and follow up of their use and effectiveness. They require a commitment on the part of the individual to wear what is often a cumbersome device. 

What the synthesised research says

While resting hand and wrist splints do not seem to affect pain or range of movement, participants in the trials preferred wearing a resting splint to not wearing one. Firm conclusions about the effectiveness of working wrist splints in decreasing pain or increasing function for people with rheumatoid arthritis are not possible from the identified randomised controlled trials.

Potential adverse effects such as decreased range of movement do not seem to be an issue although some splints do decrease grip strength and dexterity.

Extra-depth shoes and moulded semi-rigid insoles (orthotics) provide pain relief during weight-bearing activities such as standing, walking and climbing stairs. Supported insoles do not have a clear impact on pain or function but they may be effective in preventing progression of deformity (hallux abductus angle).

How it was tested

The review authors identified 10 randomised controlled trials that had a total of 449 participants. These trials investigated resting hand and wrist splints (2 trials), working wrist splints (5 trials), special shoes and insoles (3 trials). None of the studies were similar enough to allow pooling of results.

Resting wrist and hand splints

In one trial, participants assigned to wearing a splint for one month preferred using a splint (low temperature thermoplastic splint or a padded medium temperature thermoplastic splint) to no splint (relative ratio 5.5, range 2.1 to 14.5).

The participants in the second trial who wore splints for six months had no overall differences in pain, grip strength, or number of swollen joints versus those who did not. After two months of wear, participants preferred use to non-use, and padded resting splints to unpadded splints.

Working wrist splints

- to limit movement (circumduction and reduce torque) during heavy wrist activities.

There is evidence that regularly wearing wrist splints during work over up to six months does not benefit pain, morning stiffness, pinch grip or quality of life and may decrease grip strength.

Special shoes and insoles (orthotics)

The one trial of participants assigned to wearing special shoes over two months or regular footwear showed clear benefits of wearing extra-depth shoes. They had less pain when walking or climbing stairs and had more minutes of pain free walking. (100 mm visual analogue pain score (VAS) when either walking (MD -19, range -28 to -9) or stair climbing (MD -27, range -38 to -16), and more minutes pain free walking time (MD 18, range 8 to 28). This was without clear changes of pain when not weight bearing, fatigue or sense of wellness.

Extra-depth shoes with semi-rigid insoles provided slightly better pain relief than extra-depth shoes alone after 12 weeks of wear (on a 100 mm VAS scale, MD 1.9, range -3.3 to -0.5) but had no effect on function, in one trial. When participants wore soft insoles in extra-depth shoes there was no difference on any measure compared to when they wore extra-depth shoes alone. Surprisingly therefore, after having experienced all three combinations, nearly half voiced a preference for the soft insoles, and nearly half preferred the semi-rigid insoles. Those who chose soft insoles had similar pain with both types of insoles, while participants who preferred semi-rigid insoles experienced significantly more pain with soft insoles.

The third trial examined the effects of Rohadur supporting insoles versus placebo insoles

After up to three years of wear, individuals who used posted insoles demonstrated less progression of hallux abductus angle (RR 3.6, range 2.2 to 5.9) but no difference in pain or function.

Side effects and general cautions

Resting wrist and hand splints are designed to provide pain relief through immobilisation of actively inflamed joints. Pain relief is only likely to be evident during times of active joint inflammation.

Working wrist splints are recommended to limit movement (circumduction) and reduce stress on joints during heavy wrist activities, they may also limit the dexterity required in certain activities. The splints may also become too soiled to be practical with some activities. Therefore, both wearing time and amount and type of stress on the joints may vary widely among study participants depending on the types of activities which they usually carry out. Furthermore, people wearing splints may tend to feel more secure with heavy activities so that they may expose themselves to greater joint stress.

Overall, the quality of the included studies was only fair. Most did not use a placebo or blind participants to the treatment allocation, which is difficult with these devices.Only five trials reported that evaluators were not aware of which group a participant had been allocated to (group allocation).


M Egan, L Brosseau, M Farmer, M Ouimet, S Rees, P Tugwell, G Wells. Splints and Orthosis for treating rheumatoid arthritis. Cochrane Database of Systematic Reviews 2001, Issue 4. Art. No.: CD004018. DOI: 10.1002/14651858.CD004018.


People with rheumatoid arthritis (RA) live with pain, joint stiffness and decreased muscle strength, fatigue and difficulties with daily activities of grooming and dressing, preparing meals and other housework, shopping, work and leisure activities. They are challenged with physical, personal, family, social and work aspects of daily life. 

Occupational therapy (OT) sets out to help people carry out their daily activities and to overcome barriers - to maintain, improve or compensate for decreased abilities. This is done by offering skills training for self care and productivity, education of joint protection and energy conservation skills, prescription of assistive devices and instruction on their use, the provision of splints and counseling.

Occupational and physical therapists provide treatments that aim to give additional symptom relief and to assist and improve day-to-day functioning. Occupational therapy offers skills training, counseling, education about protecting joints, provision of assistive devices and splinting. Instruction in the use of assistive devices, training of self care activities and productivity activities are the three most often chosen interventions by occupational therapists for rheumatoid arthritis patients

What the synthesised research says

Two randomised controlled trials compared comprehensive occupational therapy to no therapy over 6 to 10 weeks showing improved functional ability with occupational therapy. Neither pain nor feelings of depression were clearly reduced.

Using training of motor function, improvements in hand function in terms of pain and functional ability – grip strength and range of movement - were not definitive.

Instruction on joint protection increased level of knowledge and had some benefit on functional ability without any change in pain levels.

The incorporation of the outcomes of trials of other design led to indications that wearing a splint can reduce pain.

Provision of splints may have a decrease of dexterity as a side effect.

How it was tested

The researchers made a thorough search of the medical literature and found 38 studies (16 randomised controlled trials 6 controlled but not randomised trials and 16 trials that used other design studies, to back up findings from the controlled trials). The studies dated from 1978 to 2001.

1. Comprehensive occupational therapy (3 randomised trials, one compared to alternative treatment; 1 other design; a total of 343 participants)

Two randomised controlled trials compared comprehensive occupational therapy to no therapy over 6 to 10 weeks showing improved functional ability with occupational therapy. Neither pain nor feelings of depression were clearly reduced.

2. Training of motor function (6 randomised and controlled trials compared to either no treatment (3 trials) or alternative treatment (3 trials) and 1 other trial; involving a total of 258 people).

Improvements in hand function in terms of pain and functional ability – grip strength and range of movement - were not definitive. One trial reported problems with the upper extremities for people who performed resistance exercises. In another trial, some people found that mechanical continuous passive motion was uncomfortable, heavy work and caused fatigue.

3. Instruction on joint protection and energy conservation (5 randomised and controlled trials (3 compared to no therapy) and 4 trials comparing before and after therapy; involving a total of 370 people with rheumatoid arthritis). People were followed up for between 3 weeks and 6 months and, in one trial, at one year.

Instruction on joint protection increased level of knowledge and improved functional ability without any change in pain levels.

One randomised trial reported a decrease in handgrip strength and range of motion but questioned whether this was because people were protecting their joints.

4. Advice and instruction in the use of assistive devices (one controlled trial comparing occupational therapy to alternative treatment and 1 comparing before and after instruction; involving 212 participants). This before and after study reported a decrease in pain with the use of assistive devices in the kitchen.

5. Provision of splints (16 studies of which 7 were randomised and controlled trials, 3 using a non-treated control group; involving 606 participants). The types of splints used were a working splint, resting splint, three types of anti-deformity splints and an air-pressure splint.

The incorporation of the outcomes of trials of other design led to indications that wearing a splint can reduce pain. Three trials looked at pain immediately after providing a splint and two of these reported a clear decrease in pain while wearing a working splint. The effect on pain after splinting for one week to eighteen months was assessed in ten studies. Trials which compared splinting with no treatment (2 studies) reported reductions in pain (19 to 36%).

The effect of splinting on grip strength immediately after provision of the splint was evaluated in six studies. Two high quality studies presented an increase in grip strength while wearing a splint. The effect of splinting on grip strength after a period of time was measured in four controlled trials with the highest quality trials not finding any clear benefit with splinting.

Four studies measured range of motion. The two high quality randomised controlled trials did not find any clear changes in range of movement with splints reported no significant differences between groups; another did report significant results after wearing an anti-boutonniere splint for 6 weeks.

Side effects and general cautions

No trials were identified on training of skills and counseling or on people’s ability to participate socially. participation. All studies with a multi-disciplinary intervention were excluded from the review.

One randomised controlled trial reported a decrease of grip strength when wearing working splints and people in the study reported removing the splint when doing activities that required dexterity. This was supported by other studies whereas others reported that arm and hand functions were not significantly affected by splint wear.


EMJ Steultjens, J Dekker, LM Bouter, D van Schaardenburg, MAH van Kuyk, CHM van den Ende. Occupational therapy for rheumatoid arthritis. Cochrane Database of Systematic Reviews 2004, Issue 1. Art. No.: CD003114.pub2. DOI: 10.1002/14651858.CD003114.pub2.


Rheumatoid arthritis is a ‘for life’ condition in which the people who are affected have uncertain disease progression as well as experiencing unpredictable exacerbations and remissions. Often the effects of the disease causing pain, disability, deformity and reduced quality of life persist even with treatment. Patient education is one approach used to help people manage and live with their disease. Education programmes set out to teach them about rheumatoid arthritis and how to adjust their daily activities according to the disease symptoms they experience. They are often run with groups of people that may include partners and family.

The programmes offered differ in content and can include formal instruction, exercise, biofeedback or psychosocial support components; psycho-behavioural methods promote changes in health behaviours. Behavioural treatments include: pain management; visualisation techniques; relaxation training; cognitive behavioural therapy; occupational therapy; nurse patient contracts.

This review divides the programmes into: information only, the exchange of information; counselling, mainly to provide social support and to give people the opportunity to discuss their problems; behavioural treatment, such as behavioural instruction, skills training and biofeedback.

What the synthesised research says

Patient education programmes for adults with rheumatoid arthritis had small short-term effects (at completion of the programme) on pain (4%), disability (10%), joint counts (9%), patient global assessment (12%), psychological status (5%) and depression (12%).

The reported studies did not find any clear longer-term benefits (at 3 to 18 months) although there was a trend for improved disability scores.

How it was tested

The researchers made a thorough search of the medical literature and found thirty-one randomised controlled studies that investigated the effects of patient education programmes on symptoms of rheumatoid arthritis. 

These outcomes were compared at the end of the programmes (after 6 to 20 weeks, depending on the programmes used) and at around 3 to 18 months for longer-term benefits.

At the end of the programme

Patient education had clear but small benefits on disability (SMD -0.17, range -0.25 to -0.09; 2275 participants), joint counts (SMD -0.13, range -0.24 to -0.01; 1158 participants), patient global assessment (SMD -0.28, range -0.49 to -0.07; 358 participants), psychological status (SMD -0.15, range -0.27 to -0.04; 1138 participants) and depression (SMD -0.14, range -0.23 to -0.05; 1770 participants). These outcomes involved from 358 (global assessment by patients) to 2275 participants (disability).

Anxiety was not clearly reduced and there was only a trend for a small reduction in pain immediately following patient education. (SMD -0.08, range -0.16 to 0.00; 2229 participants). In the trials using a visual analogue scale (0 to 10 cm) to measure pain (12 studies, 1112 participants) the difference in pain score was WMD -0.38 (range -0.71 to -0.05).  

The most important benefit was for disability with an effect size of -0.17. The review authors report that this is small compared with drug treatment with methotrexate (-1.48, range -1.82 to -1.14) and glucocorticoids when given in addition to such disease modifying' drugs (-0.57, range -0.92 to -0.22).

When the different components of the patient education programmes were looked ar, behavioural treatment was the only one that showed clear though small benefits.

Behavioural treatment

Showed clear effects on scores of disability, patient global assessment and depression; a trend for pain; and no change in scores for joint counts, psychological status, anxiety and disease activity.

Information only

Had no clear effects on pain, disability, joint counts, patient global assessment, anxiety, depression and disease activity. However, scores on pain and psychological status showed a trend in favour of the information-only group.


Showed no clear effects for scores on pain, disability, joint counts, patient global assessment, anxiety, depression and disease activity; a trend was found for scores on psychological status.


At longer term follow up

No clear effects of patient education were found. A trend was seen in favour of patient education for disability scores (SMD -0.09, range -0.20 to 0.02; 1308 participants). Trends favouring behavioural treatment were found for scores on disability and depression.

Side effects and general cautions

The patient education provided in the studies reviewed had small short-term effects and no obvious long-term benefits on disability, joint counts, patient global assessment, psychological status and depression. In evaluating clinical effects of patient education, it must be taken into account that patient education was provided in addition to standard medical care so the effects of patient education. People were invited to take part in the trials, which is different to routine clinical practice where they are more likely to select themselves for education sessions. Studies looking at specific parts of the body, for example the hand, were excluded from this review.

The quality of studies was, on average, not high but did improve with the more recent studies.


RP Riemsma, JR Kirwan, E Taal, JJ Rasker. Patient education for adults with rheumatoid arthritis. Cochrane Database of Systematic Reviews 2003, Issue 2. Art. No.: CD003688. DOI: 10.1002/14651858.CD003688.