A smart phone has measured the walking and movements of more than a hundred thousand adults in a British study. The collected data can be used to predict a person’s risk of death from a six-minute walk.
The prediction is as accurate as complicated surveys that have analyzed a person’s death predictions, the study claims.
Wing so put your smartphone in your pocket and go for a walk. Six minutes is enough.
During that time, the various sensors of the mobile phone measure your movements. They tell an enlightened researcher what your risk of dying in the next five years is.
That’s what he says health researcher and professor Bruce Schatz from the University of Illinois at Urbana–Champaign, USA. He did research at the Carl R. Woese Genomics Institute.
Previously the risk of death has been measured by examining health and the level of everyday physical activity in many ways.
For this, wearable motion detectors and, for example, sports watches have been needed.
Sports watches have grown in popularity. But most of us already own a smartphone. They basically have the same sensors as expensive sports watches.
However, it is difficult to collect all activity data from a smartphone, says Schatz.
Schatz therefore checked with his team the data collected from the mobile phones of more than a hundred thousand, exactly 100,655 participants.
The data had been collected in the United Kingdom Biobank. The statistics contained information on the health information of middle-aged Britons who have lived in Britain for more than 15 years.
As part of a larger study, subjects wore motion sensors on their wrists for a week. About two percent of the participants had died in the following five years.
Schatz studied this information using a machine learning model. His group developed an algorithm that evaluates the data from the sensors. The sample was taken from about a tenth of the people in the original material.
Algorithm evaluate the information of a six-minute walk from the material. They were assembled into the sensors of the smartphone, which measure movement and acceleration.
Especially for those suffering from heart or lung diseases, the algorithm produced a good prediction model.
Sick people first slow down their walking pace when they are out of breath. Then they pick up the pace again briefly, explains Schatz.
The researchers then tested their model on data from another group.
Researchers assigned to this job the so-called c-index points. This metric is commonly used to evaluate the statistical accuracy of the sample and results in biostatistics.
The value of the C-index was 0.72 for Schatz. The reading is on par with other assessment methods used when researchers measure human life expectancy.
The prognosis is just as good as traditional methods that assess risk factors, compares biostatistics researcher Ciprian Crainiceanu from Johns Hopkins University from Maryland, USA.
Originally the study used motion detectors on the wrists. However, the accelerometers of a standard smartphone can do the same during walks, says Schatz.
Now he is planning a big study with smartphones. It could even be used to make weekly forecasts.
With motion sensors, you can get sufficiently accurate results of walking pace and steps. The method can be applied to national health surveys, Schatz says in the university’s press release.
He told about the research New Scientist -magazine. Its published by PLoS Digital Health.
#Health #research #ordinary #smartphone #give #good #prediction #life #expectancy #based #sixminute #walk
