An epidemiologic study on hand osteoarthritis in peri-to-postmenopausal women

Abstract

Osteoarthritis is a slowly developing chronic joint disease mainly characterized by joint pain and nodes with no curative treatment available except for joint replacement. The etiology of osteoarthritis is not exactly known, but the hypothesis that osteoarthritis not only evolves from wear-and-tear but is also inherent to systemic components is widely accepted. Systemic inflammation as in obesity or dyslipidemia was shown to negatively influence osteoarthritis of non-weight bearing joints such as joints in the hands. Hand osteoarthritis develops frequently in postmenopausal women. Menopausal transition in women mainly occurs from age 45 to 54, involves changes in sex hormones, and is associated with vasomotor and genitourinary symptoms mainly treated with systemic and vaginal hormone replacement therapy, respectively. Further associated symptoms such as joint pain or osteoarthritis (symptomatically treated with painkillers) or increased lipid levels (mainly treated with statins to reduce the risk of a cardiovascular event) are less known to the general public but carry a high disease burden. By means of epidemiologic studies using women’s primary care health records in the United Kingdom, this thesis aimed to help find drugs potentially delaying hand osteoarthritis onset by describing and assessing drugs treating symptoms evolving in menopausal transition in association with hand osteoarthritis. Potential negative associations may result in a decreased burden of this incurable disease. In a first descriptive study, we described incidence rates of hand osteoarthritis and of hormone replacement therapy use in women aged 40 to 69 years. We observed that rates of hormone replacement therapy initiation and of new diagnoses of hand osteoarthritis behaved inversely over time and uniformly in 5-year age groups between 40 to 54 years but not in older age groups. Hormone replacement therapy initiation rates shaped in a skewed Gaussian curve with a tail in older age groups while onset of hand osteoarthritis plateaued from age 55. In a second nested case-control study, observing women from age 45 longitudinally, we assessed the association between systemic hormone replacement therapy initiation and hand osteoarthritis overall and in women with recorded menopause only as recorded menopause was a major confounder. Most hand osteoarthritis cases occurred shortly after menopause, therefore, we assessed the timing of hormone replacement therapy initiation relative to menopause in current users as well as of hormone replacement therapy cessation relative to hand osteoarthritis diagnoses in past users, compared to non-users. The association between hormone replacement therapy use and hand osteoarthritis yielded an increased risk of hand osteoarthritis of 32%. However, in women with recorded menopause, the risk of hand osteoarthritis in hormone replacement therapy users disappeared compared to non users. Furthermore, we observed a 28% decreased risk of hand osteoarthritis if hormone replacement therapy was initiated around menopause and used continuously, compared to non-users. This potential beneficial effect diminished the later hormone replacement therapy was initiated. However, we also observed a statistically non-significant 25% increased risk of hand osteoarthritis shortly after therapy cessation. In a third cohort study, in women aged 45 to 64 years, we assessed the association between statin initiation and hand osteoarthritis and between statin initiation and generalized osteoarthritis (i.e. multiple joints affected, hand osteoarthritis is usually part of generalized osteoarthritis), overall, stratified by age, and by pre-existing dyslipidemia. Furthermore, we used psoriasis and tinnitus as negative control outcomes to control for confounding by differential menopause onset (psoriasis) and healthcare seeking behavior (psoriasis, tinnitus). We observed that statin use was neither associated with hand osteoarthritis nor with generalized osteoarthritis irrespective of age or pre existing dyslipidemia. The use of negative control outcomes corroborated this finding. Our results support the existing hypothesis that menopause is a risk factor of hand osteoarthritis. However, it is likely not the only risk factor for hand osteoarthritis because otherwise we would have expected hand osteoarthritis incidence rates to decline similarly to those of hormone replacement therapy use among older age groups. Furthermore, our results suggest that timely initiation of hormone replacement therapy relative to menopause may be crucial for a potential delay of hand osteoarthritis onset to at least after hormone replacement therapy cessation. Finally, our results suggest that the lipid lowering effect of statins does not seem to translate into a reduced risk of hand osteoarthritis in peri to postmenopausal women.