Twice a year, clock changes jolt the routines of millions, yet their effect on physical activity has been hard to measure at scale. Researchers in Jessilyn Dunn's laboratory at Duke, working with colleagues including statistician Alexander Volfovsky, turned to large-scale Fitbit data from the US All of Us Research Program to find out, publishing their analysis of daylight saving time and activity patterns in the journal Nature Health.
The headline result is reassuringly subtle: total daily steps did not change after the transitions. What shifted was timing. After the autumn clock change, morning activity rose while evening activity fell; after the spring change, the pattern reversed. The effects varied across individuals and demographic groups, with work schedules and daily routines appearing to shape how readily people adjust their movement to the new clock.
Beyond the finding itself, the study by Hayoung Jeong, Srikar Katta, Will Ke Wang and colleagues makes a methodological argument that matters for public health: consumer wearables, worn by millions of people in their everyday lives, offer a practical way to assess how policies and environmental changes ripple through behaviour, effectively turning a biannual national ritual into a well-powered natural experiment.