What is your role in the frame of the i-PROGNOSIS project?

Karolinska Institutet (KI) brings theoretical background in behavioural neuroscience and its expertise in experimental analysis and modelling of task-dependent human behaviour and perceptual response. This expertise is exploited mainly in the field of meal-dependent behavioural analysis.

Additionally, KI is also actively involved in the overall behavioural modelling of the collected datasets in order to help in the development of predictive behavioural models of Parkinson’s disease based on non-intrusive data collection from mobile phones and smartwatches.

Why is it important to explore eating behaviours of people living with early Parkinson’s?

Weight changes (both gain and loss) have been observed among Parkinson’s disease (PD) patients. Potential explanations for these observations might be the progressive changes in motor and non-motor symptoms that occur during disease development. These changes are probably associated with eating behavior – i.e. handling food on the plate, transporting food into the mouth, manipulating food in the mouth as well as swallowing food – potentially leading to changes in long-term food intake.

Since: 1) both weight gain and weight loss might result in negative long-term health consequences such as early death, and 2) the fact that no objective studies have been conducted to assess eating behaviour among PD patients, we believe that novel studies are urgently needed in this area. We therefore intend to establish the first steps in an effort to bring more evidence to these important issues during the i-PROGNOSIS project.    

Furthermore, in the setting of early Parkinson’s diagnostics, we are hopeful that small behavioural changes during in-meal behaviours are both: a) present early in the disease’s progress and b) detectable through non-invasive smartwatch-based data collection. If these two assumptions hold, then we might have an additional behaviour that can be monitored in a non-invasive fashion, in order to add power to the multi-modal behavioural predictive models that are developed in i-PROGNOSIS. 

Do you have preliminary findings to share?

At this point in time, we are past the mid-point of the collection of in-meal behavioural data from Parkinson’s disease patients and matched controls. Right now our collaborators in TUD, using data acquisition tools developed by AUTH, have recorded (in great detail) meals from 32 PD patients and 10 control volunteers in Germany.

Once data collection is complete (March 2019), we are expecting that we will be able to share exciting new data about the in-meal characteristics in Parkinson’s disease. Note that this study will be the first ever in this field, allowing us to better analyse the real life data that are already being collected through the iPrognosis SData application.

Read all interviews here: