Women’s brain aging: effects of sex-hormone exposure, pregnancies, and genetic risk for Alzheimer’s disease

Abstract

Estrogen exposure may influence women's risk of Alzheimer's disease, but little is known about how it affects normal brain aging. Recent findings from the UK Biobank demonstrate less evidence of brain aging in women with a history of multiple childbirths. Here, we investigated the link between brain aging, estrogen exposure, and APOE genotype beyond the effects of parity in 16,854 UK Biobank women. Machine learning was used to predict brain age based on neuroimaging-derived measures, and the difference between an individual's predicted and chronological age was used as an estimate of brain aging. Cumulative estrogen exposure was estimated using an index including age at menarche and menopause, BMI, time since menopause, and duration of hormone replacement therapy. Endogenous hormone exposure was approximated by reproductive span, while exogenous exposure was estimated by usage, onset, and duration of hormone replacement therapy and oral contraceptives. Higher cumulative, endogenous, and exogenous estrogen exposure were each linked to higher brain age relative to chronological age. Earlier onset of hormone replacement therapy, particularly before menopause, was associated with less evident brain aging in APOE e4 carriers only, while higher circulating estradiol levels during menopause were linked to more evident brain aging in carriers and less evident brain aging in non-carriers. The results indicate that estrogen exposure and parity may differentially relate to women's brain aging, and that APOE e4-specific associations between estrogen and brain aging may be of importance for optimizing hormone replacement therapy regimes in perimenopausal women. Women's health | Estrogens | Parity | Apolipoproteins E | Neuroimaging | Machine learningCorrespondence: a.m.g.d.lange@psykologi.uio.no claudia.barth@medisin.uio.no

Towards an understanding of women's brain aging: the immunology of pregnancy and menopause

“…Differences in content, dosage, and administration may contribute to inconclusive findings across observational studies and randomized trials [225,226], and neuroprotective effects of estrogen could potentially depend on 'optimal' exposure, which could vary between individuals. For instance, some evidence points to genotype-specific influence of estrogen exposure on brain aging: increased E2 levels induced by HRT have been associated with reduced risk of developing AD in apolipoprotein E type 4 (APOE e4) non-carriers, but not in carriers [227,228], and higher menopausal levels of E2 have been linked to more evident brain aging in APOE e4 carriers, and less evident brain aging in non-carriers [229].…”

Exogenous Sex Hormone Effects on Brain Microstructure in Women: A diffusion MRI Study in the UK Biobank

“…Determining effects on the brain's white matter of HT in preand post-menopausal years would require decades of longitudinal research following individuals across the lifespan. However, non-invasive in vivo imaging biomarkers that are associated with age and with cognition have been increasingly used as proxies to assess effects of menopausal HT on the brain (with large recent studies in the UK Biobank 25,26 ). In this current paper, we set out to measure and model ageassociated effects on white matter microstructure to infer possible effects of exogenous sex hormones on the brains of preand post-menopausal women.…”

“…Supplemental analyses were carried out to test the link between duration of therapy, age at HT/OC onset and white matter aging on diffusivity metrics in women. We analyzed single-and multi-shell diffusionweighted brain MRI data from the UK Biobank 26 and processed them using 3 increasingly complex diffusion models. Diffusion indices were extracted and averaged across a whole brain white matter skeleton.…”

Fast qualitY conTrol meThod foR derIved diffUsion Metrics (YTTRIUM) in big data analysis: UK Biobank 18608 example

“…Diffusion magnetic resonance imaging (dMRI) provides a range of structural brain features based on routine clinical measurements, which has contributed to its popularity across fields and applications (Kochunov et al 2015), (de Lange et al 2019), (Westlye et al 2010). Advanced dMRI is technically challenging and often involves time-consuming acquisitions placing high demands on the performance and stability of the scanner hardware.…”

“…Derived diffusion metrics from diffusion or kurtosis tensors are sensitive to a range of subjectspecific factors such as age or various brain disorders, but also to applied numerical algorithm or its programming implementation (Lebel et al 2012), (Grinberg et al 2017), (Maximov et al 2015), (David et al 2019). The effects of noisy observations on subsequent between-subjects analysis involving the derived diffusion metrics can be mitigated using simple outlier detection procedures (see, for example, (Richard et al 2018), (Tønnesen et al 2018), (de Lange et al 2019)). However, few publications have directly assessed the effects of QC filtration of final data and performing a sanity check of the derived scalar maps before the statistical analysis.…”