Purpose: Hand osteoarthritis (OA) is more common in women and previous studies have noted its incidence increases around the time of the menopause. More recently, there has been a report in a large primary care dataset of increased incidence of hand OA following cessation of menopausal hormonal therapy (MHT). We set out to describe the temporal relationship between menopause, use of MHT and the onset of hand symptoms in a population of women with hand OA in a secondary care (rheumatology) setting. Our objectives were: i) to explore if there was a temporal relationship between the onset of menopause and the onset of hand OA symptoms; ii) to describe if current or previous use of MHT was associated with the timing of onset of hand OA symptoms; iii) To examine if cessation of MHT was associated with onset of hand OA symptoms.

Methods: This was a retrospective review of UK NHS medical records (UK IRAS ethics #282499). Sequential females aged 18 years and older referred to specialist hand OA clinic in London, UK 2007-2015 with a diagnosis of Hand OA by accepted clinical criteria were included; exclusions were other forms of arthritis or causes of hand pain. Predefined outcome measures were reported age of onset of hand symptoms, reported age of Final Menstrual Period (FMP); predefined variables were use of systemic estrogen-containing MHT (subgroups current, previous, never users) and menopausal status (premenopausal, post-menopausal [at least one year post FMP]; peri-menopausal period defined as FMP+/- 4 years). These variables, age and other demographics were from the usual healthcare records, with all predefined variables recorded by standardised proforma used routinely in this setting. Descriptive statistics and linear regression modelling (coefficient, 95% CI are given) were carried out in STATA IC.

Results: 82/98 females were post-menopausal, with mean age at FMP of 50. In these post-menopausal women, median time from FMP to hand symptom onset was 3 years (range -25, 60 years). 48/82 (59%) of these women developed their hand symptoms within the defined peri-menopausal period. In women who had discontinued MHT, the median time from cessation of therapy to onset of hand symptoms was 6 months. Past users of MHT were older at onset of reported hand symptoms than women who had not taken MHT (coeff.4.7 (0.92, 8.39); P=0.015). This association persisted when adjusted for menopausal status.

Conclusions: This study of real-world, patient-level data in women in a hospital care setting adds to the circumstantial evidence which associates the menopause with onset of OA symptoms; in this population, a significant subgroup of women are seen to present around the time of their menopause with hand OA symptoms. The use of MHT and its cessation appeared to be associated with the age of onset of hand OA symptoms in this particular population. There of limitations of this type of study, including risk of recall or referral bias, or other uncontrolled bias; also that the type and duration of MHT could not be taken into account. It is important that these findings are re-tested in different settings. It cannot be known from this type of study whether the effects of menopause are mediated by sex hormone deficiency directly or by other factors; also whether menopause is causative, is a disease modifier or a symptom modifier in these women. OA is a female-preponderant disease: the sexual dimorphism particularly observed in hand OA and the clues that the influence of menopause may give us on the underlying pathogenesis of OA deserve more attention.