In what scientists say is the largest study of human movement ever conducted, researchers collected physical activity data from smartphone users in 111 countries – and discovered what they think could be a significant unrecognised public health risk.

Analysing the level of activity of more than 700,000 people, the team found it wasn't the average level of physical activity in each country that proved the biggest predictor of obesity – it was the amount of "activity inequality" within each population that increased the prevalence of obesity.

When they compiled the data, the team noticed that in countries with little obesity, people tended to walk about the same amount each day. In other countries, though, where there were bigger gaps between those who walked a lot and those who walked only a little, obesity was significantly more of a problem.

This correlation persisted even when the average level of steps taken was the same across countries – showing that while national physical activity averages might seem like a reasonable way of comparing countries' level of fitness, by themselves they aren't at all reliable for gauging obesity prevalence.

Of the 46 countries represented with results from at least 1,000 app users, the five countries with the highest levels of activity inequality were Saudi Arabia, then Australia, Canada, Egypt (all equal second), and the US.

At the other end of the scale – where more evenly distributed patterns of physical activity correlate with lower prevalence of unhealthy weight gain – China leads the list, followed by Sweden, South Korea, and Czechia.