Sleep quality and overall survival among cancer survivors in the Netherlands: A PROFILES registry study

Background: This study aimed to determine whether: 1) sleep quality was associated with overall survival (OS) among a heterogeneous sample of cancer survivors; 2) this association differed per cancer diagnosis; 3) aspects of sleep quality (e.g., sleep latency, daytime dysfunction) were associated with OS among a subsample of colorectal cancer (CRC) survivors; and 4) adjustment for depressive symptoms changed these associations. Methods: Several cohorts from the population-based PROFILES registry, including adult cancer survivors diagnosed between 1990 and 2014 with 11 cancer diagnoses, were used. Data on sleep quality (3 categories: no, non-clinically, and clinically important sleep quality impairment) was collected through the insomnia scale of the EORTC QLQ-C30 and the PSQI for CRC survivors only (n = 1245). Clinical data were obtained through the Netherlands Cancer Registry. Cox regression analysis was used to assess adjusted hazard ratios (HRs). Results: 7195 participants were included, of which 36 % died (median follow-up since inclusion, 9 years). Clinically impaired sleep quality was associated with lower OS (HR, 1.17[1.05; 1.30]) compared to no sleep problems. Stratification by cancer diagnosis suggested a consistent pattern. After adjusting for depressive symptoms, sleep quality was no longer significantly associated with OS (HR, 1.10[0.97; 1.24]). Daytime dysfunction and long sleep duration were significantly associated with lower OS in CRC survivors, also after adjustment for depressive symptoms. Conclusion: Cancer patients reporting clinically low sleep quality had a lower OS. However, this might be partly explained by patients’ depressive symptoms. In CRC survivors, daytime dysfunction and long sleep duration were, independent of depressive symptoms, related to lower OS.