Ref (type)	Population	Outcome, Interventions	Results and statistical analysis	Effect size	Favours
Adverse effects
[52] RCT 4-armed trial	854 people	Discontinuation of treatment 12 months with theophylline (220 mg or 300 mg slow-release formulation) with placebo Absolute results not reported	P <0.002 (theophylline v placebo) The theophylline arm was open label	Effect size not calculated	placebo
[53] RCT	110 people	Proportion of people reporting an adverse effect 12 months 10/57 (18%) with theophylline (100 mg twice daily) 3/53 (6%) with placebo	P = 0.076		Not significant