PhD Scientific Days 2018

Budapest, April 19–20, 2018

Individual slow wave morphology is a marker of ageing

Przemyslaw Ujma, Péter

Péter P. Ujma1,2, Péter Simor3,4, Martin Dresler5, Róbert Bódizs1,2
1 Institute of Behavioural Sciences, Semmelweis University, Budapest, Hungary
2 National Institute of Clincical Neuroscience, Budapest, Hungary
3 Institute of Psychology, ELTE, Eötvos Loránd University, Budapest, Hungary
4 Nyírő Gyula Hospital, National Institute of Psychiatry and Addictions, Budapest, Hungary
5 Donders Institute of Brain, Cognition and Behaviour, Radboud University Medical Centre, Nijmegen, The Netherlands

Language of the presentation

English

Text of the abstract

Slow wave activity is a hallmark of NREM sleep which reflects a synchronized onset and cessation of cortical neuronal firing and which is strongly reduced with ageing. Gross slow wave activity can be assessed by several summary variables, such as the number of waves, mean wave amplitude or delta power. However, the shape of the slow wave is characterized by a stereotypical morphology, similar to evoked potentials, which may be more strongly related to physiological parameters such as ageing than amplitude or slopes which are lower-resolution descriptions of the wave shape. We recorded full-night polysomnography in 160 subjects (age 17-69) and detected slow waves using an automatic method. We triggered all slow waves to the maximum negative deflection and averaged the mean EEG of the electrodes F3 and F4 1-1 sec before and after these deflection points to obtain an individual average slow wave of each subject. We calculated the individual average slow wave amplitude, slope steepness and the total number of slow waves (summary parameters). We investigated if entering the individual slow wave morphology (with one variable at each EEG sampling point) in a least-square regression predicting age would yield additional accuracy beyond the effects of the summary parameters. We used LASSO regression in order to select a limited subset of predictors. Wave amplitude at multiple time points emerged as independent predictors of age in all models, beyond the effects of summary parameters. Younger subjects were characterized by more rapid and more prominent deflections on both the descending and ascending slope of slow waves. Our results suggest that the fine-resolution individual morphology of the slow wave is a more sensitive marker of ageing than summary parameters, and it may yield increased prediction accuracy as a clinical biomarker of age-related cognitive decline.

Data of the presenter

Associate Professor, Postdoctoral Researcher
Registers as a recipient of the ÚNKP-17-4 New National Excellence Programme of the Ministry of Human Capacities