Prevalence and risk factors of sleep disturbances in breast cancersurvivors: systematic review and meta-analyses

Background Breast cancer remains the most frequently diagnosed malignancy among women worldwide, with rising incidence numbers. In Belgium, one out of eight women will be diagnosed with breast cancer. Fortunately, 80% of those breast cancer patients will still be alive 10 years after diagnosis due to improvements in screening and treatment strategies. However, an important portion of the breast cancer survivors (BCS) will face side effects, such as sleep disturbances, long after treatment ends. It has been demonstrated that untreated insomnia in BCS negatively impacts mood, physical symptoms, pain sensitivity, fatigue, and quality of life. Furthermore, insomnia is increasingly considered an independent risk factor for future depression in BCS. The importance of understanding sleep disturbances in cancer populations has been highlighted and recognized as warranting further research. Therefore, the purpose of this systematic review was to determine the prevalence and the risk factors for the development of sleep disturbances in BCS. Methods PubMed, Web of Science, and PEDro were systematically screened for studies encompassing data regarding the prevalence or risk factors of sleep disturbances in BCS. If possible, meta-analyses were performed. Subgroup analyses were undertaken based on the methodological quality, study design, type of sleep disturbance, and the use of a measurement tool with strong psychometric properties to investigate significant heterogeneity (I-2 > 50%) across studies. Results A total of 27 studies were found eligible. The pooled estimate for sleep disturbances prevalence is 0.40 (95% confidence interval (CI) = [0.29-0.52], I-2 = 100%, p < 0.00001) and ranged from 0.14 (95% CI = [0.04-0.24]) to 0.93 (95% CI = [0.91-0.95]). Subgroup analyses did not reduce the heterogeneity among studies. Meta-analyses were performed for seven risk factors. Significant differences for the odds of developing sleep disturbances were found for hot flashes (pooled OR (ORp) 2.25, 95% CI = [1.64-3.08], I-2 = 0%, p = 0.90), race (ORp 2.31, 95% CI = [1.56-3.42], I-2 = 0%, p = 0.47), and menopause (ORp 1.84, 95% CI = [1.11-3.06], I-2 = 0%, p = 0.70). After withdrawing the studies that did not rely on the use of a measurement tool with strong psychometric properties, pain (ORp 2.31, 95% CI = [1.36-3.92], I-2 = 27%, p = 0.25), depressive symptoms (ORp 3.20, 95% CI [2.32-4.42], I-2 = 0%, p = 0.63), and fatigue (ORp 2.82, 95% CI = [1.98-4.02], I-2 = 0%, p = 0.60) became significant as well, with a substantial decrease of heterogeneity. Conclusion Prevalence for sleep disturbances ranged from 0.14 to 0.93 with the vast majority of the studies investigating insomnia and sleep-wake disturbances. High heterogeneity makes it difficult to draw firm conclusions. Pain, depressive symptoms, hot flashes, fatigue, non-Caucasian race, and menopausal status were significantly associated with increased odds for developing sleep disturbances.