Personal virtual coaches for obese people need to improve self-confidence

Reviewed by James Ives, M.Psych. (Editor)Jun 24 2019

Just give overweight people an app that monitors their movement patterns and that stimulates them to be active and exercise. This could be more than a simple pedometer. Technology advanced to the level of personal 'virtual coaches'. But despite all progress, it isn't that simple. The systems are not yet capable of analyzing and influencing patients' behavior and enhancing their self-confidence. For long term success, this is necessary, says researcher Reinoud Achterkamp of the University of Twente and Roessingh Research and Development

The number of people that are overweight or even obese shows a worrying rise: in The Netherlands, half of the adults is overweight, the percentage grows depending on age. Among children, it is about 12 percent. At the same time, technology could help, with the fast growing number of apps helping people to maintain a healthy lifestyle. Of course there are simple pedometers or GPS-trackers, but new trends include virtual coaches that monitor your movements and can send you personal recommendations. It is not that simple to set people in motion and develop new patterns in life, researcher Reinoud Achterkamp says. The key is self-confidence or self-efficacy.

For his PhD research, Achterkamp equipped 30 test persons with a 3D movement sensor and a smartphone giving feedback and instructions, for a period of several weeks. Does the patient show a lack of movement for some period of time, he or she will get the advice to have a walk or bike ride. The app is 'context-aware': it can take the weather into account or the physical surroundings. For COPD patients this approach was used before; it included air quality for that and could also give the advice to take a rest. Another feature of the system is that a tone of voice is chosen that the patient feels comfortable with.

To turn this in a truly effective coach, the system should be able to recognize behavior. A good example is the 'state of change' the patient is currently at. This determines the level of motivation for changing behavior. Is doing exercise already part of the daily pattern or does the patient feel resistance in some way? Striking outcome is that all patients, no matter at what 'state of change' they are, show a pattern of diminishing activity during the day. "

Reinoud Achterkamp, Researcher, University of Twente

The current approach, according to Achterkamp, lacks the type of feedback that gives the patient the feeling 'I can do this' and that works stimulating. What could help is sharing results with other patients, just like runners or cyclists share their latest results on social media.

Gamification could also help, rewarding the patient for behavior in the same way as the apps do that stimulate people to go to work by bike. Emphasizing the unhealthy aspects of their lifestyle won't help: they are fully aware of that, and this will only give rise to resistance. Mentioning the advantages, in relation to the latest results, is much more motivating. In this era of Big Data, it could be attractive to add much more information, like details about the food intake during the day. This again is an action that could negatively influence motivation. According to Achterkamp, using artificial intelligence for learning about the patient's behavior and for giving adequate feedback will add a lot to current systems: introducing psychology into the system. Apart from that, a form of blended care will remain necessary, in which the patient discusses progress with a therapist or fellow-sufferers, next to the interaction with the virtual coach.

The research is part of the Dutch national project Smart reasoning systems for well-being at work and well-being at home (SWELL), one of the projects of the COMMIT ICT program. Achterkamp did his research at Roessingh Research and Development - the R&D division of Roessingh rehab center in Enschede. Achterkamp did this together with the Biomedical Signals and Systems group, the 'Personalised eHealth technologies' program of UT's TechMed Centre.