Identifying memory processes in decline during menopausal transition and the neuroprotective effects of extraneous oestrogen using machine learning
Approved Research ID: 76586
Approval date: October 26th 2021
Women of menopausal age are the fastest growing demographic in the workforce. It is estimated that about 13 million women are currently transitioning to the menopause in the UK. This project will study one of up the 34 menopause symptoms: brain fog including memory decline. It will:
* explore in detail the changes in grey and white matter that occur;
* make inferences from the areas of brain changes about the type of memory processes that decline (e.g. working memory or episodic memory); and
* improve understanding of how taking hormone replacement therapy (HRT) in the 'early window' or 'late window' affects memory processes.
In the past, researchers hypothesised that brain atrophy in areas of memory function was associated with a decrease in oestrogen as women transitioned to the menopause. Their studies used lower quality scans from Magnetic Resonance Imaging (MRI) than are now available from UK Biobank, and they controlled for a relatively small number of confounders (e.g. Body Mass Index, education level, etc.). Testing proved that additional memory impairment had taken place during the menopause transition when controlling for ageing effects.
However, structural imaging-based results were inconclusive about the effects on brain atrophy of oestrogen delivered in HRT. Use of scans employing other MRI protocols and machine learning methods have been limited.
The aim of this study is to follow on from this original research using UK Biobank's high-quality MRI brain imaging volumes and metrics, and HRT data, while controlling for confounders. It will use machine learning methods: a novel deep learning method that has been developed at King's College London which has been validated against UK Biobank brain imaging data and stacked generalisable ensembles. Both methods produce interpretable results.
The study will compare the results between post-menopause women who have taken HRT in the 'early window', post-menopause women who have taken HRT in the 'late window', post-menopause women who have not taken HRT, pre-menopausal women who have not taken HRT, and age-matched men.
The project aims to improve clinical support for women and has two experienced external clinical advisers. The results will be used for the development of an innovative app for women transitioning to the menopause to coach them at home and at work through brain fog generally and memory decline specifically.