Women with PCOS are often resistant to the action of insulin and may produce high levels of insulin to compensate for this. Insulin resistance can lead to increased production and activity of androgens (male hormones), and associated symptoms such as hirsutism (excess hair growth) and acne. Weight loss can help in PCOS by reducing high insulin and androgen levels. Metformin is prescribed in PCOS with the expectation that it will also reduce the effects of high insulin and androgen levels.
Guidelines and recommendations
The Endocrine Society 2018 guideline on evaluation and treatment of hirsutism in premenopausal women advises against the use of metformin. Instead, the combined oral contraceptive pill (COC) is recommended first-line for most women who need medicinal treatment, with antiandrogens being added in after six months, if necessary.
Due to the higher risk of deep vein thrombosis (DVT), the World Health Organization (WHO) recommends that COCs with 35 micrograms of ethinylestradiol and cyproterone acetate should only be used to treat moderate to severe hirsutism or acne. Lower risk preparations, i.e. those containing levonorgestrel, norethisterone or norgestimate, are recommended first line for mild to moderate hirsutism.
An international evidence-based guideline for the assessment and management of PCOS, developed by the Centre for Research Excellence in Polycystic Ovary Syndrome in partnership with multiple organisations, recommends:
- the COC first-line for managing high levels of androgens in adult women with PCOS
- the COC as an option for hirsutism and acne in adolescent women with PCOS
- metformin combined with lifestyle changes (e.g. diet, exercise) for adult women with PCOS for the management of weight, hormonal and metabolic outcomes
- metformin combined with lifestyle changes in adolescent women with a diagnosis of PCOS or clear symptoms of PCOS before diagnosis
- healthy eating and regular physical activity in all women with PCOS.
The licence for metformin only covers the treatment of type 2 diabetes so the prescriber takes full responsibility when prescribing metformin for PCOS.
If metformin is prescribed for PCOS, patients should be made aware that its use is ‘off label’. Potential adverse effects should be discussed.
As a general principle, individuals’ characteristics, preferences and values should be taken into account when recommending a medicine for PCOS.
Generally, studies using metformin for hirsutism have been small and reduction of hirsutism may not have been the main focus of the studies. Also, studies may not have been long enough; it is unlikely that a response to treatment will be seen within six months because of the nature of the hair-growth cycle. In many cases, the Ferriman-Gallwey (FG) score is used, but this has its own limitations. For instance, a low score may be recorded even in the presence of copious hair growth in one or two areas.
In adolescents with PCOS, metformin has been found to improve hirsutism more than placebo, as has metformin plus the COC. In adult women, a Cochrane review on interventions for hirsutism found that metformin may be no more effective than placebo at reducing FG scores. In another Cochrane review comparing metformin with the COC, metformin was as or less effective than the COC at improving hirsutism. However, adding metformin to COC treatment improved adult women’s hirsutism more than the COC alone.
A systematic review and meta-analysis on metformin therapy for acne in patients with PCOS analysed 13 studies which suggested that acne scores (an objective measure of acne improvement) are significantly decreased with metformin use. However, studies using metformin for acne are generally of poor quality and vary greatly, with different acne assessment scales, metformin doses and treatment durations being used. Based on three randomised, controlled studies (RCTs) analysed in a Cochrane review comparing metformin with the COC, metformin is less effective than the COC for subjective improvement of acne. There are too few studies to determine whether metformin improves acne scores more effectively than placebo or the COC alone/with metformin.
- Re-worked format of Q&A and updated evidence