Title: Identification of sleep-wake-cycle parameters associated with dementia subtypes
Lab: Inserm U1153
« EpiAgeing » Team
Location: Université de Paris, Inserm U1153, 10 avenue de Verdun, 75010 Paris
Team director: Archana Singh-Manoux, Séverine Sabia
Supervisors: Clémence Cavaillès, Ian Meneghel Danilevicz, Séverine Sabia
Topic: The sleep-wake cycle changes with age and is frequently altered in individuals with
neurocognitive disorders. These alterations may differ depending on the type of disease, but their
ability to characterize and differentiate neurocognitive disorder subtypes remains poorly studied,
particularly using objective and multidimensional measures.
Accelerometers allows for the objective and continuous assessment of various aspects of these
behaviours. In the CIRCAME cohort, more than 1,000 patients recruited from Parisian memory clinics
wore a wrist-worn accelerometer continuously for nine days (9 × 24 hours). These data are used to
derive a wide range of parameters of physical activity (intensity, duration, and distribution throughout
the day), sleep (duration, quality, and fragmentation), rest-activity rhythms (using both parametric and
non-parametric approaches), and chronotype, which represent the main dimensions of the sleep-wake
cycle.
The aim of this internship is to identify the accelerometer-derived parameters that best discriminate
the main neurocognitive disorder subtypes, including Alzheimer’s disease, dementia with Lewy bodies,
vascular dementia, and frontotemporal dementia.
To address this objective, specific supervised machine learning methods will be applied. The detailed
collection of sociodemographic, lifestyle, and health-related data will allow analyses to be adjusted for
a wide range of other disease markers, in order to assess the added value of accelerometer-derived
measures compared to established risk factors.
Requirements:
Excellent programming skills (in R) and a strong background in machine learning methods are essential
to successfully conduct these analyses.
Contact : Ian Meneghel Danilevicz ian.meneghel-danilevicz@inserm.fr, Clémence Cavaillès
clemence.cavailles@inserm.fr, Séverine Sabia, severine.sabia@inserm.fr
