Exploration of sleep as a specific risk factor for poor cardiometabolic and mental health & the comparison of subjective and objective assessments of sleep

Abstract

Numerous studies have attempted to evaluate the impact of sleep on cardiometabolic and mental health, although most of the population-based studies utilised self-reported sleep assessment and health status. Therefore, the main aim of this project is to explore the relationship between accelerometer measured sleep and cardiometabolic and mental health amongst the UK Biobank participants. The UK Biobank collected extensive information of the general UK population. They have also collected accelerometry data allowing the extraction of sleep duration and quality. Disease status was obtained from their primary care record. Out of the 84,411 participants with available processable accelerometry data, 17.3% slept <5 hours/night, 25.9% slept 5-6 hours/night, 33.5% slept 6-7 hours/night, 18.7% slept 7-8 hours and 4.6% slept >8 hours/night. Short sleep duration was significantly associated with the male gender, older age, high body mass index, social deprivation and ethnic minority group (p<0.001). A significant ‘U-shaped’ association was found between sleep duration and metabolic disease status. Both short and long sleep durations were also associated with negative mood and worse cognitive performances including slower reaction time and worse visual memory (p<0.001). These findings showed the importance of sleep in maintaining health. However, sleep misperception was found to be common leading to a discrepancy between subjective and objective measurements of sleep. A cross-sectional analysis was conducted on 28 human participants (11 controls and 17 patients). Sleep was assessed using a paper sleep diary, wrist-worn tri-axial accelerometer and laboratory-based polysomnography. The level of cortisol, melatonin, mitochondrial DNA damage and gene expression was measured using saliva, urine, skin swab and hair samples, respectively. An overestimation of sleep duration was observed in this study which is consistent with the UK Biobank analysis. Patients were found to have a longer sleep duration, but a lower sleep efficiency. Moreover, patients were found to have a lower level of melatonin and cortisol. A ‘U-shaped’ association was found between sleep duration and mitochondrial DNA damage level. Finally, circadian rhythm and mitochondria-related pathways have been identified in the gene expression analysis. However, these associations were not found to be statistically significant. Therefore, it is proposed that larger sample size should be considered in future studies informed in part by further power calculations based upon the findings presented in the current thesis.