br Methods br Results Main characteristics of study
Results Main characteristics of study population are presented in Table 1. AD cases had higher co-morbidity score and lower socioeconomic position than controls. Women with AD had undergone any surgery (unilateral/bilateral oophorectomy, hysterectomy, hysterectomy with bilateral oophorectomy) more often than controls. Use of Erlotinib Hydrochloride and combination HT was more common among AD cases than controls, regardless of the mode of HT use. Table 2 depicts the crude and adjusted associations between type (estrogen/progestogen) and duration of HT use and AD. In the model adjusted for socioeconomic status, co-morbidities, surgery, and gynecological cancer, any use of estrogen and progestogen were associated with increased risk of AD with OR (95% CI) of 1.10 (1.06–1.12) and 1.13 (1.10–1.17) respectively. Up to 10 years of estrogen, progestogen, and combination HT use was associated with increased risk of AD. Use of estrogen HT for >10years was related to lower relative risk of AD with OR (95% CI) of 0.91 (0.84–0.99), while >10years of progestogen and combination HT were not significantly related to AD with OR (95% CI) of 1.0 (0.90–1.2). Association between modes (oral/transdermal) and formulations of HT use and AD are shown in Table 3. Oral estrogen was the most commonly used type among both cases and controls. Oral use of estrogen and progestogen and combined (oral and dermal) use of estrogen were associated with higher relative risk of AD, but the absolute differences in prevalence were small. When the exposure to different treatment modalities were categorized according to duration, 0–5 years of transdermal estrogen use and 0–10 years of oral estrogen and progestogen use were associated with higher relative risk of AD, but also here the absolute differences were small.
Discussion Our study depicts a higher relative risk of AD among postmenopausal women who used HT up to 10 years (0–5 years and 6–10 years) while >10years of estrogen monotherapy and oral estrogen use were associated with lower risk of AD. However, the absolute differences were small. Oral progestogen use itself was not significantly associated with AD and similarly no association was observed between any duration of transdermal estrogen use with AD. These findings did not alter when surgery (oophorectomy, hysterectomy), and gynecological cancer were taken into account. Our findings imply that HT was not an important determinant of future AD risk. The protective association with AD among long-term estrogen users is in line with the Cache County study where reduced risk of AD was observed among women using HT for >10years . Similarly prior use of HT was protective against AD in a cohort study only when use of HT exceeded >10years . Increased risk of AD in short-term HT users in our study likely relates to initiation of HT in late postmenopausal period, as average age at start of HT in this study for cases and controls was 64.1 and 63.8 years respectively. Moreover, we could not account for HT use prior to this age, because prescription register was established in Finland in 1995. This is in accordance with the findings from the WHIMS, the biggest clinical trial to date on use of HT and all cause dementia. WHIMS explored risk of all cause dementia as a secondary outcome with use of opposed HT i.e. conjugated equine estrogens (CEE) and medroxyprogesterone acetate (CEE+MPA) and unopposed (CEE only) HT versus placebo among women >65year old. The use of opposed HT increased the dementia risk, but this did not occur with unopposed HT , . In our study, the nominal risk increase was observed among users of both opposed and unopposed HT. Findings from previous register based studies favor ours as use of estrogen alone or in combination with progestogen among≥75year old women was not beneficial for cognitive performance . Medical records based use of oral or parenteral estrogen use for ≥6 months among 264 AD cases and 264 controls did not show any significant association with AD in a case-control study which can be due to short duration of HT use and small sample size .