Skip navigation

Exercise training in JIA

Clinical bottom line

Six months of aquatic exercise training in children with JIA made no difference to quality of life, physical fitness, or joint status. The number of swollen or tender joints fell in children who exercised.


Children with JIA are likely to be less physically fit than children without the condition. The joint pain, stiffness, and sometimes deformity that come with JIA are likely to lead to low levels of physical activity, and therefore of fitness. It is not known whether supervised training regimens can improve health status or quality of life.


T Takken et al.Aquatic fitness training for children with juvenile idiopathic arthritis. Rheumatology 2003 42: 1408-1414.

Randomised trial

The study randomised 54 children with JIA to a six month aquatic training regimen or control group. For inclusion children had to have a diagnosis of JIA according to set criteria, a phase of remission without medication of no longer than six months in the absence of joint pain, tenderness or morning stiffness, and an ESR within normal limits. All had received NSAIDs and DMARDs in the six months before the trial. Exclusions were systemic disease with fever, low haemoglobin, exercise contraindicated by specialist, bone marrow transplant, or not feeling confident in water.

Randomisation was stratified by oligoarticular or polyarticular disease, by an independent means. The study was open as investigators and subjects knew whether they were participating, but measurements when taken were done in ignorance of earlier evaluations. Measurements were made after randomisation and three and six months later. Quality of life, physical functioning and joint status were measured, as well as maximum exercise capacity on a cycle.

All children received usual care and medical treatment. Training children received an aquatic group exercise programme (groups of 2-4), lasting one hour a week for six months (practically about 20 sessions over six months because of holidays and pool closures).


There was only one drop out at 15 weeks, but information was included in the analysis. The mean age was about 9 years, and groups were well matched at baseline (27 children in each group). There were no differences between the groups at three or six months for functional ability, quality of life, or physical fitness.

The number of swollen and tender joints fell by 55% in the exercise group, and rose by 21% in the control group (Figure 1). This did not achieve statistical significance (p=0.07).

Figure 1: Average number of swollen and tender joints


This study was well done. It showed tantalising results with regard to swollen and tender joints, but no other significant differences. There is a terrific discussion about the difficulties in this type of study in children in general, and children with JIA in particular. Longer, and larger, studies of exercise in children with JIA might well be informative. Aquatic training, because it is not load bearing on joints, has particular meaning.