Skip navigation
Acute Pain | Chronic Pain | General

Physiotherapy exercises for back pain

Clinical bottom line: Trials are of insufficient quality to draw a clear conclusion. It is not clear whether physiotherapy is better than other conservative treatments, or whether it is better than no treatment. It is not possible to establish which types of exercises are the most effective.

About 80% of the population will suffer from back pain during their active lives. This is usually self-limiting, and will disappear within a few months in 90% of patients. However, there is a lack of consensus on which of the many treatments available is the best for chronic back pain. Physiotherapy is widely used for back complaints, usually exercise therapy given alone or in combination with other treatments (e.g. massage, heat, traction, ultrasound or short wave diathermy).

Systematic review

Koes BW, Bouter LM, Beckerman H, van der Hiejden GJMG, Knipschild PG. Physiotherapy exercises and back pain: a blinded review. 1991; BMJ 302: 1572-6.

Date review completed: 1990

Number of trials included: 16

Control group: active/no intervention/placebo

Main outcomes: pain, mobility

Inclusion criteria were randomised controlled trials of physiotherapy for back pain; physiotherapy given individually, not in groups; physiotherapy given alone or with additional treatment; back pain present at baseline; full journal publication.

Reviewers provided a descriptive summary of trials, methodological assessment together with a 'positive' or 'negative' conclusion for each trial as concluded in the original report.


Of the 16 trials included, most were of poor methodological quality, and trials were generally not adequately designed to assess the intervention - small group sizes, high rates of loss-to-follow-up, internal sensitivity, poor statistical analysis. Type of exercise, control interventions and length of intervention varied between trials. Ten trials had negative conclusions and six had positive conclusions. Reviewers note that there is a general trend for higher quality trials to have positive results. Blinding status was unclear, but ten of sixteen appeared to be evaluator-blind.

Inactive/placebo comparisons

Four trials compared exercise therapy with no therapy or placebo therapy. One of four trials showed a benefit of exercise therapy for pain and activity at four and 12 weeks, but this did not persist to three months. The remaining trials showed no benefit for acute pain, sciatic symptoms and chronic low back pain.

Active comparisons

Seven trials compared exercise therapy with other conservative treatments. Two of seven trials showed significant benefit compared with hot packs and rest and with mini back school (i.e. one session). The second study showed benefit was still present at one year. Five of seven trials showed no benefit when compared with manual therapy, home care instructions, non-steroidal anti-inflammatory drugs, manipulation, manipulation and mobilisation or short wave diathermy.

Different exercise therapies compared

Eight trials compared different types of exercise therapy - mainly isometric flexion exercises compared with extension exercises. Four of eight trials showed no difference (although all were flawed). Four showed some benefit - one favouring three months of intensive dynamic back extensor exercises over a less intensive treatment plus massage and heat, one favoured extension over flexion exercises, and two favoured flexion exercises over mobilisation plus other exercises and, in one trial, massage and heat.

Adverse effects

Reviewers did not report on adverse effects

Further reading

Related topics