Chinese herbal medicine for oligomenorrhoea and amenorrhoea in polycystic ovary syndrome: A randomised feasibility study in the United Kingdom

摘要

Purpose: Polycystic ovary syndrome (PCOS) affects 6-18% ofwomen of reproductive-age and oligomenorrhoea and amenorrhoeare cardinal symptoms. Conventional management isassociated with side-effects and anecdotal evidence suggestsChinese herbal medicine (CHM) can help. Individualised CHMis regarded as more effective than standardised, but requiresin vestigation in randomised controlled trials (RCTs). Thisstudy explores the feasibility of conducting an RCT comparingstandardised and individualisedCHMfor regulating menses inPCOS.Methods: This pragmatic, practitioner-blinded feasibilitystudy randomised 40 PCOS participants with oligo- or amenorrhoeainto 2 parallel groups - standardised or individualisedCHM - prescribed at 16 g granules/day as a tea for 6 months.Our primary aim was to evaluate feasibility of offering standardisedand individualised CHM within an RCT and collectmenstrual data for sample size calculation. Secondary dataincluded body mass index, weight, hirsutism and safety dataon liver/kidney function and adverse events.Results: 40 women were recruited within our planned 7-month recruitment-period. 29 participants (72.5%) completedthe study, 3 were lost-to-follow-up (7.5%) and 8 withdrew(20%). Analysis of covariance (ANCOVA) of menstrual rateper month showed statistically significant improvements instandardised CHM (MD 0.18±SE0.06, 95%CI 0.06 to 0.29) andin individualised CHM (MD 0.27±SE0.07, 95%CI 0.15 to 0.39).This did not reach between-group statistical significance (MD0.10±SE0.08, 95%CI -0.07 to 0.26, p=0.26). ANCOVA of secondarymeasures suggest no important changes in body massindex or weight. Liver/kidney function at Week 4 was normal(n=35), abnormal ALT (n=1); at final visit was normal (n=30),abnormal (n=0). The case of abnormal ALTwas later confirmedan acute response to alcohol.Conclusion: We have demonstrated that a CHM RCT forPCOS is feasible and preliminary data suggests promisingmenstrual response in both groups. This data will be used toinform sample-size calculation and design of a main studythat will incorporate an active or placebo-control.