Table 1

Characteristics of excluded studies

Study	Reason for exclusion
Graff-Lonnevig et al⁶	Study was not truly randomised. Allocation was based on who lived closer to the gymnasium and this group was included in the exercise training arm.
Cambach et al⁷	Study included a composite intervention and included both subjects with asthma and chronic obstructive pulmonary disease. A physiotherapist run programme included breathing retraining, mucus evacuation, and exercise.
Svenonius et al⁸	Not randomised as the subjects could choose which group they would like to belong to for the study.
Bundgaard et al⁹	Both the groups were trained and the only difference was the intensity of training with no difference in duration or frequency of training.
Dean et al¹⁰	The study was too short, being only for 5 days.
Edenbrandt et al¹¹	Frequency of physical training was low, subjects only exercised once a week.
Orenstein et al¹²	Not truly randomised, subjects were assigned to groups according to the availability of transport.
Hirt et al¹³	Mentioned as randomised, but all patients who were in hospital were assigned to the group. Subjects who had severe asthma were assigned to the control group.
Henriksen et al¹⁴	Subjects were said to be randomly chosen but the intervention group of 28 were chosen from a total of 42 because they were inactive in sports and physical games and had poor physical fitness. Control groups were more physically active than the subjects in the intervention group.
Neder et al¹⁵	Not truly randomised, subjects were assigned to groups consecutively. First 26 subjects entered the training group and the next 16 subjects had no training.