Women with the e4/e4 genotype who have used HRT had lower hippocampal, parahippocampal, and thalamus volumes than those with the e3/e3 genotype who had never used HRT, with differences equivalent to approximately 1-2 years of hippocampal atrophy observed in typical healthy ageing trajectories in midlife.
Key Findings
Results
Women with the e4/e4 genotype who used HRT had significantly lower hippocampal volume compared to women with the e3/e3 genotype who never used HRT.
Hippocampal volume was 1.82% lower in e4/e4 HRT users versus e3/e3 non-users
The difference is equivalent to approximately 1-2 years of hippocampal atrophy observed in typical healthy ageing trajectories in midlife (0.98%-1.41% per year)
The sample included N=207,595 women with mean age 56.25 years (SD=8.01)
Women with the e4/e4 genotype who used HRT had lower parahippocampal and thalamus volumes compared to women with the e3/e3 genotype who never used HRT.
Parahippocampal volume was 2.4% lower in e4/e4 HRT users versus e3/e3 non-users
Thalamus volume was 1.24% lower in e4/e4 HRT users versus e3/e3 non-users
All brain volumes were normalised for head size
Analyses of structural brain regions further controlled for scanner site
Results
The two-way interaction between HRT use and APOE status was not detected for measures of cognition.
Despite significant brain volume differences, no corresponding cognitive differences were found in the HRT × APOE interaction
Generalised linear regression models were used to quantify cross-sectional associations between HRT and brain health
The study included women from the UK Biobank with no self-reported neurological conditions
Results
No clinically meaningful three-way interaction between APOE, HRT, and age was detected for either brain volumes or cognitive measures.
The three-way interaction was interpreted relative to the scales of the cognitive measures used and normative models of ageing for brain volumes
The absence of a clinically meaningful three-way interaction was consistent across both cognitive and structural brain health outcomes
Age was included as a continuous variable in three-way interaction models
Results
Post hoc sensitivity analyses within the APOE e4/e4 group showed consistent effect sizes, suggesting observed brain volume differences may be partly attributed to HRT use rather than solely to APOE status.
Effect sizes were consistent within the APOE e4/e4 group in post hoc sensitivity analyses
The authors note that the cross-sectional design means they cannot exclude the possibility that women who used HRT may have a predisposition for poorer brain health (i.e., the healthy user bias or indication bias)
The study examined HRT use, age of initiation, and duration of use in relation to brain health outcomes
Methods
The study examined women across a broad range of APOE genotypes including e2/e2, e2/e3, e3/e3, e3/e4, and e4/e4 in relation to HRT use and brain health outcomes.
Five APOE genotype groups were included: e2/e2, e2/e3, e3/e3, e3/e4, and e4/e4
Two-way interactions between HRT and APOE status were modelled, in addition to three-way interactions including age
The study population was N=207,595 women from the UK Biobank with a mean age of 56.25 years (SD=8.01 years)
Covariates included time since attending centre for completing brain health measure, surgical menopause status, smoking history, BMI, education, physical activity, alcohol use, ethnicity, socioeconomic status, vascular/heart problems, and diabetes
Ambikairajah A, Khondoker M, Morris E, de Lange A, Saleh R, Minihane A, et al.. (2024). Investigating the synergistic effects of hormone replacement therapy, apolipoprotein E and age on brain health in the UK Biobank.. Human brain mapping. https://doi.org/10.1002/hbm.26612