A data science team tapped into smartphones of more than 717,000 people worldwide to capture the number of steps they take each day as an indicator of physical activity. Findings from the NIH-funded study by researchers at Stanford University in Palo Alto, California appear in yesterday’s issue of the journal Nature (paid subscription required).
Smartphone data from more than 68 million days of activity by 717,527 individuals across 111 countries reveal variability in physical activity around the world. (T Althoff et al, Stanford University)
The Stanford team led by computer science professor Jure Leskovec is seeking techniques for harnessing the wealth of data produced by chips on mobile phones called accelerometers that sense movement of the device, for example when answering a call or when the screen orientation changes. Those same sensors respond to stepping motions, and can capture the number of steps taken by phone users.
Leskovic, with bioengineering professor Scott Delp, analyzed data on numbers of steps taken recorded by a popular wellness app Argus, by mobile software company Azumio, in Palo Alto. The data, provided with identifying details removed, offer insights into the “global pandemic of physical inactivity” as the researchers call it, which according to a 2012 study results in some 5.3 million deaths per year. Yet as the authors also point out, up to now there are few if any large-scale measurements of physical activity on a global scale.
The team analyzed activity data from 717,527 individuals in 111 countries for a total of 68 million days of activity, but reported results from 46 countries with 1,000 or more participants. Nine in 10 of these participants came from 32 high-income countries, as indicated in a World Bank index, while the remainder were from 14 middle-income nations. Worldwide, participants averaged about 5,000 steps per day, but the researchers found the range from sedentary to active levels was wider in countries with higher levels of obesity. The researchers call this gap “activity inequality.”
The authors rank activity inequality of the 46 countries, and found Hong Kong at the top with the smallest inequality range and Saudi Arabia last on the list with the largest. The United States is close to the bottom, about equal to Egypt, Canada, and Australia. Another key factor in activity inequality is gender, where lower activity rates among women are associated with activity inequality, which contribute to the low rankings of the U.S. and Saudi Arabia.
“When activity inequality is greatest,” says Leskovec in a university statement, “women's activity is reduced much more dramatically than men's activity, and thus the negative connections to obesity can affect women more greatly.”
The team also analyzed a subset of data from app users in 69 cities in the U.S., and compared data from participants with the cities’ rankings on “walkability” or pedestrian friendliness. The authors found an association between a city’s walkability rank and lower inactivity equality scores, with women comparatively less active than men in cities less friendly to pedestrians.
Making use of the popularity of smartphones for public health monitoring is the subject of a current challenge sponsored by Centers for Disease Control and Prevention. CDC is seeking proposals for innovative ways of using mobile devices to collect data on health-related behavior, with the deadline for initial proposals set for 4 August 2017.
Originally posted in Science & Enterprise