Effect of menopausal status on cognitive function and sleep in breast cancer survivors.

Abstract

e12709 Background: The majority (70-80%) of invasive breast cancers (BC) express estrogen receptors, making endocrine therapy a cornerstone of most BC treatments. However, this therapy can induce menopause or hypoestrogenism in premenopausal women, and is associated with numerous negative symptoms (e.g., sleep problems). Many BC survivors also report cognitive dysfunction both during and after treatment, which is also associated with poor sleep quality, underscoring the need to examine cognitive function and sleep quality in relation to menopausal status. This study examines differences in cognitive function and sleep disturbance among BC survivors by menopausal status (1: pre/peri-menopausal, 2: naturally post-menopausal, and 3: cancer- or treatment-induced post-menopausal). Methods: This secondary, cross-sectional analysis included baseline data from 180 BC survivors from two cohorts (n = 128 stages 0-III; n = 52 stage IV). Sociodemographic information, clinical characteristics, and PROMIS Sleep Disturbance 4a assessments were collected via REDcap surveys. Administration of remote cognitive test batteries (via BrainCheck) included the Trail Making Test – Parts A and B (TMT-A, TMT-B), Digit Symbol Substitution Test (DSST), Stroop Color-Word Test, and the Immediate and Delayed Memory Tests. Analysis of covariance was conducted to assess the effect of menopausal status on the standardized cognitive test scores and sleep disturbance (t scores), covarying for years of education and age. Post hoc comparisons with Bonferroni corrections and effect sizes were also calculated. Results: Significant main effects of menopausal status were found for sleep disturbance (p = 0.012; ω² = 0.049) and TMT-A scores (p = 0.045; ω² = 0.036). Post hoc comparisons indicated that pre/peri-menopausal survivors had significantly lower sleep disturbance than naturally post-menopausal survivors (p = 0.009, Cohen’s d = 0.84). Naturally post-menopausal survivors performed worse than cancer-induced post-menopausal survivors on the TMT-A (p = 0.043, Cohen’s d = 0.51). Since Homogeneity of Variance was violated for Delayed Memory scores, nonparametric models (Kruskal-Wallis) and post hoc tests (Dunn’s) were conducted. A significant main effect for menopausal status was found (p = 0.02; rank ε² = 0.045), with pre/peri-menopausal survivors having significantly higher scores than naturally post-menopausal survivors (p = 0.022) and cancer-induced post-menopausal survivors (p = 0.022). Conclusions: After controlling for age, BC survivors who have gone through natural menopause before treatment may be more vulnerable to sleep disruption, slower processing speed, and worsened memory performance compared to pre/peri-menopausal BC survivors. Further, BC survivors who have gone through cancer- or treatment-induced menopause may be more vulnerable to worsened memory performance compared to pre/peri-menopausal BC survivors.