With increasing age, the performance of athletes decreases – that much seems to be certain. But when exactly this happens and what the specific processes are, has hardly been researched. A group of scientists led by Bergita Ganse, professor of surgery at Saarland University in Homburg, together with teams at RWTH Aachen University and TU Darmstadt, have now developed a method that uses artificial intelligence to predict the decline in physical performance of senior athletes .
One asked oneself whether it would be possible to “predict the performance of an athlete into senior age with a single measurement,” said Ganse. They were also interested in early predictions. “Are, for example, athletes who performed better than others in their younger years, also more efficient in old age?” Says the Werner Siemens endowed professor for innovative implant development.
100 years of information
Of the Scientist’s Algorithm is based – as is so often the case nowadays in AI – on machine learning. “We have shown that it is possible to predict the future decline in performance of an athlete on the basis of machine learning with only one starting point more precisely than with previous methods,” says Ganse. The researchers’ data set comes from Sweden: the team from Aachen, Darmstadt and Homburg used the information collected from around 5,500 athletes from the so-called Swedish Veteran Athletics database. A total of over 20,000 data points are recorded here.
However, only one discipline was examined: running. The reason: Here the sporting rules have hardly changed, which makes the data, which comes from a period of 100 years, comparable. Runners ran 100, 200, 800 meters, regardless of whether they are 23, 40 or 70 years old, according to the researchers.
Database trains algorithm
With the database it was possible to train an algorithm that could predict age values ​​based on earlier values. Interestingly, it turned out that people who show comparatively poor athletic performance at a young age also had a lower drop in performance later on. Very high-performing, young athletes were more affected by a drop in performance, as were older athletes with low output levels. The lowest rate of decline was recorded in high-performing athletes with a high starting age.
Ganse & Co. believe that high athletic performance with a high starting age stands for “continuous, lifelong involvement in other sports” – or a combination of healthy nutrition and a good genetic constitution. The central thesis of the researchers: Those who run a top time in their discipline at an advanced age will remain more productive than their peers even at an older age. “So it is still worthwhile to start exercising at an advanced age.”
(bsc)
