With the help of apps and wearables, people are collecting more data about themselves than ever before. This can not only help them personally, it can also be very helpful for research. A conversation with complexity researcher Dirk Brockmann, who learns a lot from projects for which healthy and sick people donate the data from their fitness trackers.
SZ: When I collect data about my body in everyday life or during sports, Health collect, why should I share it with you?
Dirk Brockmann: Because the data can be of much greater use than just to you alone. Ordinary citizens become citizen scientists and can advance research.
What research is this, for example?
Our biggest project of this kind to date was the Corona data donation app. During the pandemic, half a million people took part within a very short space of time and made their data available, which they had obtained via fitness trackers. This really helped us to better understand many details of the pandemic.
But data and consumer advocates often warn people against sharing their data or storing it outside of their personal devices at all.
I understand the reservations, and they are justified. After all, almost all fitness tracker data collected in Germany is stored on the manufacturers’ servers in the USA. However, through our projects, donors can retrieve their data from the USA. The data is also extremely well protected here, because our research is subject to high standards.
What insights did you gain?
During the pandemic, we saw that many people’s heart rhythm changed after a corona infection – and for a long time, even four months later. People were also less active weeks later than before their infection. Their daily step count was reduced by more than a thousand steps, and their average resting heart rate was increased for more than 120 days. It was noticeable that the vital data of people who had been vaccinated changed less; they returned to normal levels more quickly. For example, the resting heart rate of vaccinated people recovered after just three to six weeks, not eleven weeks as in unvaccinated people. On average, vaccinated people were as active as before their infection after four weeks, while unvaccinated people took six to eleven weeks to do so.
:A “German” Corona variant is spreading
Even more infectious: The XEC variant, which originated in Germany, is displacing other coronaviruses. What is known about it.
How meaningful is such data anyway? The donors are not exactly representative…
That’s right, they were not selected representatively. Anyone who wants to donate is welcome. And yet, these are incredible treasures of data because the number of participants is so large. Such large numbers of test subjects who donate such amounts of data over a long period of time cannot be achieved with a randomized study. These types of experiments must be seen as a supplement and extension of traditional studies. In addition, such data from fitness trackers or smartwatches is extremely high-resolution; we often have data accurate to the minute or second over several years. From this, you can quickly gain exciting insights that would previously have taken many years. That is why the projects are infinitely valuable.
In the meantime, you have also seen the Sleep of people. What did you discover?
We have evaluated the quality of sleep of people in Germany. This has also revealed some astonishing things. In the past, the sun had a huge influence on when people got up and went to bed. Light pollution in cities has now prevented this. But we have found that in rural areas there is still an unexpectedly high east-west dependence when it comes to sleeping. In the east, where the sun rises and sets earlier, people actually go to bed earlier. We also see some things that we cannot explain. For example, the quality of sleep of southern Germans is apparently better than that of northern Germans. We do not yet know why.
Do you often find yourself surprised by the turn your work takes?
Yes, and that’s what’s so exciting about these exploratory studies: They always give rise to new questions. We gain insights that we hadn’t even thought of before. With the Corona data donation app, for example, we wanted to understand the pandemic. It was only later that we realized that the data was also extremely helpful for understanding long Covid, because we already had people’s movement data before they got sick. Ultimately, they are natural experiments. We don’t design them; the people who donate their data to us help design them.