Daylight Saving Time Doesn’t Change Total Daily Steps — But Shifts When People Walk

Changing the clocks for daylight saving time does not alter people’s total daily step counts, but it does shift when they are active during the day, according to a new study analyzing longitudinal Fitbit data from the National Institutes of Health All of Us Research Program.

The research, published in Nature, used a natural difference-in-differences design comparing Arizona—which does not observe daylight saving time—to neighboring Mountain Time states that do. This approach allowed researchers to isolate the causal effects of the time change without relying on strict modeling assumptions.

Contrary to common belief, the study found no net change in total daily steps associated with either the spring or fall transitions. Instead, daylight saving time caused a reallocation of physical activity across different times of day.

During the fall transition, when clocks are set back, morning step counts increased by an average of 202 steps (confidence interval: [78, 326], P = 0.001), while evening step counts decreased by 180 steps (confidence interval: [−263, −97], P < 0.001). The spring transition showed the opposite pattern, with activity shifting from morning to evening.

These shifts in activity timing varied significantly by demographic group and by individuals’ inherent activity patterns. The researchers identified three data-driven phenotypes: ‘morning walkers,’ ‘neutral walkers,’ and ‘evening walkers.’ The extent to which people adjusted their activity schedules depended on these classifications, suggesting that structural factors such as rigid work schedules or perceived safety may limit some people’s ability to flexibly adapt to time changes.

Resting heart rate also exhibited subtle intraday shifts that mirrored the behavioral changes in activity timing. However, these physiological differences were deemed clinically insignificant by the study authors.

The findings indicate that abolishing daylight saving time and choosing either permanent standard time or permanent daylight time would unlikely have a meaningful effect on overall population activity levels. Any behavioral impact would primarily concern the timing of activity rather than the total amount.

This large-scale analysis provides causal evidence from real-world wearable data, offering insights that can inform ongoing policy debates about the health and behavioral effects of seasonal time changes. The study underscores how longitudinal monitoring through devices like Fitbits can reveal nuanced patterns in population health behaviors that might be missed in shorter-term or self-reported studies.