Background We assessed the impact of menopausal hormone therapies (MHTs) on the risk of colorectal cancer (CRC) in Finnish women. In addition, we analysed various MHT characteristics in relation to risk and anatomic site of CRC. Methods Users of MHT and age-matched non-users (both groups n = 364,304) in 1994–2019 were identified from the Finnish Medicine Reimbursement Register and the National Population Register. Cohorts were followed for the occurrence of primary CRC from the Finnish Cancer Registry. CRC risks were calculated as hazard ratios (HRs) and 95% confidence intervals (CIs) with a Cox proportional hazards model. Results Over a mean 18-year follow-up (13 million person-years), 4602 MHT users (1.3%) and 5103 non-users (1.4%) were diagnosed with CRC (p < 0.001). Ever use of MHT for 5 years or more was associated with a decreased CRC risk (HR 0.88, 95% CI 0.85–0.92). Estrogen-only therapy (ET) (0.81, 0.76–0.87) showed a larger risk reduction than combined estrogen-progestogen therapy (EPT) (0.92, 0.88–0.96). Tibolone was accompanied by the greatest risk decrease (0.54, 0.41–0.73), but the number of users was modest. Dydrogesterone EPT tended to confer 31–36% lower CRC risk than EPTs containing norethisterone or medroxyprogesterone acetate. Risk reductions were of similar magnitude for right- and left-sided colon cancer, but rectal cancer risk was decreased only with 10 or more years of MHT use. The risk falls persisted for over 5 years after the cessation of MHT use. Conclusions Use of ET or EPT for 5 years or more is associated with reduced CRC risk, likely mediated by estrogen, while some progestogens – but not dydrogesterone – may attenuate this effect.
Background Perimenopause is a dynamic neuroendocrine transition with increased risk of cognitive, psychological and physical symptoms. How population variability in symptom burden relates to different aspects of cognitive functioning remains unclear. Methods We measured menopausal symptom burden, task-specific cognitive performance, task confidence and metacognitive bias using the online platform “Cognitron” with 1614 participants aged 45–60 from the Great British Intelligence Test. Symptom burden was standardised (mean = 0, SD = 1), allowing coefficients to be interpreted in standard deviation (SD) units. Results Exploratory analyses showed early and late perimenopause to be a significant predictor of higher menopausal symptom burden (respectively 0.259 SD, 95% CI 0.116, 0.401, and 0.284 SD, CI 95% 0.115, 0.452, both p < 0.001). Global cognitive performance was significantly lower in individuals with higher symptom burden (−0.092 SD, 95% CI -0.133, −0.050, pFDR < 0.001, ηₚ2 = 0.0117). Higher symptom burden also predicted lower task confidence (−0.225 SD, 95% CI -0.272, −0.178, pFDR < 0.001, ηₚ2 = 0.0527) and negative bias (−0.135 SD, 95% CI -0.181, −0.089, pFDR < 0.001, ηₚ2 = 0.0206). Exploratory analyses showed that symptom burden significantly predicted poorer performance on tasks measuring executive function (Stroop interference), working memory (digit span), spatial reasoning (block design), and verbal analogical reasoning. Conclusions Our findings highlight menopausal symptom burden as a significant predictor of cognitive and metacognitive differences and longitudinal research is needed to identify those at greatest risk of cognitive decline.
