Poster F100, Tuesday, March 28, 8:00 – 10:00 am, Pacific Concourse
Associations between sleep duration and structural and functional brain MRI measures in the UK Biobank cohort
Claire Sexton1, Kai Spiegelhalder2, Stephen Smith1, Heidi Johansen-Berg1, Debbie Lawlor3, Martin Rutter4, Simon Kyle1; 1University of Oxford, 2University of Freiburg, 3University of Bristol, 4University of Manchester
While both short and long sleep duration have consistently been shown to be associated with adverse cognitive outcomes, brain magnetic resonance imaging (MRI) studies have often been limited in size and yielded varied results. UK Biobank is a population-based study of 500,000 individuals; within this cohort, both short and long sleep duration have been associated with reduced performance in tests of reaction time, reasoning, and numeric, visual and prospective memory. Here, we report associations between sleep duration and 2,501 imaging-derived phenotypes (IDPs) spanning five modalities (T1-weighted MRI, susceptibility-weighted MRI, diffusion MRI, task functional MRI (fMRI) and resting fMRI) in the first release of neuroimaging data. Individuals reporting a diagnosis of a neurological illness or sleep disorder were excluded from analyses, resulting in a sample of 5,250 participants (mean age 62.2 ± 7.5 years, 53% female, sleep duration 7.2 ± 1.0 hours). All variables were first passed through a rank-based inverse Gaussian transformation and eight confounds (age, age2, sex, age x sex, age2 x sex, head motion during tfMRI, head motion during rfMRI, and headsize) regressed out. Correlation analyses were then performed between sleep duration squared and IDPs. For each modality, the significance level was set at 0.01 divided by the number of IDPs. No significant associations were detected between sleep duration and IDPs. Therefore, we did not find evidence to support a quadratic relationship between sleep duration and MRI measures of brain structure and function. Voxelwise analyses may provide more sensitive markers of the underlying causes of cognitive deficits.
Topic Area: METHODS: Neuroimaging