Skip navigation

Ankylosing spondylitis: Knowing what to do for the best

Ankylosing spondylitis is an inflammation of the spine involving the lower back and sometimes peripheral joints. It predominantly affects young men, usually before age 30. It has a prevalence of about 0.1% (1 in 1,000) and a strong association with HLA B27 histocompatability group (95% of patients).

Physical therapy

Treatment often includes exercise and physiotherapy, regarded as being important in slowing the deterioration in spinal mobility, and maintaining quality of life. How these treatments are best delivered is not known with any certainty. A randomised trial of three different physiotherapy regimens in ankylosing spondylitis demonstrates the real problems in research to determine what to do for the best.

Randomised trial

The study [1] at Leeds randomised 44 patients to receive either:-
  1. intensive in-patient physiotherapy
    - a three-week admission which involved daily intensive exercise, stretching, aerobics, hydrotherapy and other treatments as needed.
  2. out-patient hydrotherapy and home exercises
    - a six-week period of twice-weekly out-patient hydrotherapy
  3. home exercise alone
    - patients were instructed in the home exercise regime and given diary cards to complete, with review at six weeks. They could have further diaries if wanted.

All patients were advised to continue exercising after completing the study.


A number of measurements were made immediately before starting treatment, after completing treatment, and two, four and six months afterwards. Cervical rotation, chest expansion, lumbar movement and a visual analogue scale for both pain and stiffness were measured.


Despite some initial differences, at six months there were no differences in outcomes between the groups. The bottom line was that there was no bottom line.


The authors were frank about the problems they encountered while doing this research, and it is instructive to look at some of the issues which makes a negative result suspect.
  1. Poor recruitment rate. Many patients refused to enter; if they were employed they were not willing to commit to in-patient treatment, and if unemployed they preferred to have in-patient treatment.
  2. High drop out rate. At six months 87%, 60% and 57% in each group attended for assessment. Some patients who were randomised to groups other than those they preferred, and when thwarted, defaulted.
  3. Measurements were not blinded. There was insufficient funding for measurements to be made by a physiotherapist who did not know what treatments had been given.
  4. Measurements were imprecise given the magnitude of changes expected. In a pilot study the repeatability of measurements was studied over three days. For cervical rotation, for example, the mean difference between measurements was 1°, with a standard deviation of 8.5° (95% confidence range is four times a standard deviation, in this case 34°). The mean change seen with treatment was 20° at most.


Doctors and physiotherapists have to take a pragmatic approach to treatment, and making changes is difficult where patients have chronic disorders and when they think they know what works for them. Knowing what to do for the best where good evidence is lacking will never be easy.

Nor is it easy for researchers, as this paper shows. Studies like this take an enormous amount of enthusiasm, time and dedication. The failure to generate a definite answer may be a disappointment, but highlighting these important practical problems in getting good quality answers is very important.


1 PS Helliwell, CA Abbott, MA Chamberlain. A randomised trial of three different physiotherapy regimes in ankylosing spondylitis. Physiotherapy 1996 82: 85-90.

previous or next story in this issue