Assessing the role of sleep in reducing risk and improving outcomes for breast cancer

Abstract

Conventional observational studies suggest that good sleep quality, average duration and morning-preference may reduce breast cancer (BC) risk, although residual confounding and reverse causation may limit ability to establish causality. Mendelian randomisation (MR), which uses germline genetic variants to assess causal effects, can address these limitations if certain assumptions are met. MR studies have suggested that morning-preference reduces BC risk. I have further explored sleep trait effects on BC risk and survival including potential mediators of these relationships using univariable, bidirectional, and multivariable MR analyses.
My findings suggest a protective effect of both morning-preference and ease of getting up in the morning on BC risk (OR per category increase=0.88, 95% CI=0.81, 0.96 and OR per category increase=0.67 [0.51,0.90], respectively).
Morning-preference was associated with lower total (mean SD difference per category increase=-0.07 [-0.10,-0.03]), and bioavailable testosterone (mean SD difference per category increase=-0.08 [-0.13,-0.04]), but these effects were attenuated in multivariable MR analysis. Limited evidence was found for a relationship between morning-preference and adiposity traits. Overall, there was inconsistent evidence for sex-hormone or adiposity traits as mediators of chronotype’s effect on BC.
In survival analyses, daytime sleepiness was associated with lower odds of overall BC mortality (hazard ratio (HR) per category increase=0.34 [0.14,0.80]). Insomnia symptoms were linked to lower odds of mortality in patients with ER+/PR+/HER2- tumours, receiving no chemotherapy (HR=0.18 [0.05,0.68]) and patients receiving aromatase inhibitors (HR=0.23 [0.07,0.78]), compared to no symptoms. However, results were not robust to sensitivity analyses. Therefore, advocating for shorter sleep duration in a public health setting would not be advisable.
This thesis extends previous research on the potential role of sleep traits in BC risk and outcomes. Interventions to alter sleep behaviours may help reduce BC risk and improve health outcomes in BC patients, but further longitudinal studies are needed to better understand these relationships.

Date of Award	17 Jun 2025
Original language	English
Awarding Institution	University of Bristol
Supervisor	Richard M Martin (Supervisor), Debbie A Lawlor (Supervisor), Rebecca Richmond (Supervisor) & Tim Robinson (Supervisor)