All 10 RCTs (n=615) included in this review were scored as high methodological quality. Sample size ranged from 22 to 126; five studies had fewer than 50 participants. Three trials used cross-over designs and seven were parallel. Six studies reported details of randomisation and five reported on allocation concealment. Three studies used an intention to treat analysis. Most data was not suitable for meta-analysis as a result of clinical and statistical heterogeneity as well as missing data.
Overall the following outcomes improved significantly more in patients who received amitriptyline than those who received placebo: pain (three out of eight trials); fatigue (three out of six trials); sleep disturbance (five out of seven trials); patient global assessment (five out of nine trials); physician global assessment (four out of seven trials); and tender point counts (one out of eight trials).
Amitriptyline 25mg versus placebo (six studies): Significant differences in favour of the active intervention were found for the following outcomes: pain (three out of five trials); fatigue (three out of five trials); sleep disturbance (four out of five trials); patient global assessment (five out of six trials); and physician global assessment (four out of five trials).
Amitriptyline 50mg versus placebo (four studies): Significant benefits for sleep disturbance were found in one of two trials that measured this outcome. There were no significant differences for any of the other outcomes measured: pain (three trials); fatigue (one trial); patient global assessment (three trials); and physician global assessment (two trials).
Adverse events were reported rigorously by six studies. The mean adverse event rate for any dose of amitriptyline was 51.84% (2.8% to 95%) and for placebo was 36.63% (2.8% to 80%). All adverse events were reported to be mild or moderate. There were no differences in reported withdrawal due to side-effects between active and placebo groups.