It happens every year, and nobody decides it. Somewhere in late September the step counter starts sliding. We blame the early sunset, the clock change, a vague loss of willpower. The objective data tells a different story: the strongest driver of your yearly step pattern is temperature. Not the calendar, and not day length.
That distinction sounds academic. It is not, because it completely changes what you can actually do about it.
The seasonal gap is real, and it has been measured
Let us establish the fact first. The winter dip is not a feeling. It has been documented for decades with pedometers and accelerometers rather than questionnaires.
The most-cited study on adults is still Hamilton, Clemes and Griffiths, published in 2008 in Annals of Human Biology. Ninety-six British adults wore a pedometer for four weeks in summer, then four weeks the following winter. The result: 10,417 steps per day on average in summer against 9,132 in winter, a gap of 1,285 steps per day. The difference held on every day of the week, weekends included.
This is not an isolated finding. In 2021, a scoping review in the International Journal of Behavioral Nutrition and Physical Activity pulled together 110 studies, 118,189 participants, 30 countries. Both total activity volume and moderate-to-vigorous activity were higher in summer than winter, and sedentary time was higher in winter than in spring or summer. The phenomenon is robust.
Its magnitude, however, varies wildly from one population to another. Using data from connected scales and watches, Withings published a far tighter seasonal ranking: 7,053 steps per day in summer, 6,843 in spring, 6,656 in winter and 6,541 in fall. Roughly 500 steps between the best and worst season, where the British study found 1,285.
And look closely at that ordering. In this dataset, the emptiest season is not winter but fall — 115 steps below winter. If daylight were the only engine, December should be the floor. It is not.
The peak sits near 20 °C, and heat costs more than cold
A Japanese study gives the clearest mechanism. Kimihiro Hino, Jung Su Lee and Yasushi Asami used pedometer data from the municipal Yokohama Walking Point program, covering 24,625 adults tracked across a full year, and matched every day against local weather.
The relationship is not a straight line. It is a bell curve, peaking between 19.4 and 20.7 °C (roughly 67 to 69 °F). Below that peak, each additional degree Celsius buys you between 46 and 53 steps a day. Above it, each additional degree costs between 98 and 188.
In other words, the heat penalty per degree runs two to four times steeper than the cold one. That explains why summer is not always the yearly high point, and why the true optimum looks like a pleasant October afternoon rather than the second week of August.
The curve also explains the spread between studies. A northern European winter sits a dozen degrees below the peak. A Mediterranean or southern Californian one sits roughly half that far away. Same mechanism, radically different magnitudes.
The clock change does not give your steps back
Here is the most stubborn belief, and it has just been tested properly. In April 2026, a Duke University team published an analysis of daylight saving transitions in Nature Health, using 1,157 Fitbit wearers from the All of Us research program. The design is elegant: Arizona does not observe daylight saving time, so it serves as a control group against neighboring states that do.
The verdict on totals is unambiguous. Clock changes do not move your daily step count. Not in spring (β = 31.9, p = 0.799), not in fall (β = −9.5, p = 0.950). Nothing.
What they do move is the distribution. After the fall transition, the authors measure 202 more steps in the morning and 180 fewer in the evening. In spring, the mirror image. Your steps relocate within the day; they do not vanish.
"Overall, changing times does not seem to be the public health benefit that some think that it is, at least in terms of activity levels," says Jessilyn Dunn, who led the study.
So if your monthly average collapses in November, the one-hour shift is not the explanation. What surrounds it is: temperature, rain, and appetite for going outside.
Daylight matters less than you would guess
The most striking natural experiment comes from Iceland. In the AGES-Reykjavik study, Nanna Arnardóttir and colleagues fitted 138 adults averaging 80 years old with accelerometers, once in summer and once in winter. At that latitude, the gap in day length between the two measurement windows approaches seven and a half hours.
Seven and a half hours of daylight, gone. You would expect a collapse. What the researchers observed was total activity about 19% higher in summer (111 versus 93 thousand counts per day), and lifestyle physical activity dropping from 36 minutes a day in summer to 27 in winter. Nine minutes.
Real, measurable, and far more modest than the doomsday scenario implies. The authors themselves note that a spectacular contrast in light produces relatively contained shifts in behavior.
What you can actually do with this
Enough diagnosis. Here is what these numbers license in practice.
- Measure your own gap, not the literature's. Between 500 and 1,300 steps depending on the cohort: that range is far too wide to serve as a personal benchmark. Last year's history, on the other hand, is exact. If you already track steps with a watch or a connected scale, the history is already sitting there — you just need to read it by season instead of by week.
- Compare year over year, not month over month. A drop from October to December carries no information: it happens to everyone. December 2026 against December 2025 is a genuine signal.
- Aim for the window, not the hour. If the peak sits near 20 °C, the best slots from October through March are the mildest hours of the day, often early afternoon. In summer the logic inverts: early morning or late evening, because past the peak every degree gets expensive.
- In fall, relocate rather than resign. The Duke study shows mornings naturally absorbing what evenings lose. Working with that drift — a walking commute in the morning instead of a 6 p.m. walk in the dark — takes less willpower than fighting it.
- Do not confuse today's weather with the season. One rainy week says nothing about your trend. The correlation only means something across months.
This is exactly the kind of correlation Kantise is built to surface: the weather connector places temperature and rainfall next to your step count and lets the curve appear on its own. We explored the same mechanics in our piece on correlations between sport and weather, and the seasonal effect is its most visible extension.
One last point worth making. Walking less in winter is not a moral failure. It is a rational response to a less welcoming environment, shared by more than a hundred thousand people measured across thirty countries. And if you want to offset it, the best-documented approach is also the dullest: go outside anyway, even for less time. The benefits of outdoor time do not obey the calendar, as the literature on time in nature and green spaces makes clear.
FAQ
How many steps do people actually lose in winter?
It depends heavily on climate and population. The 2008 British study measured 1,285 fewer steps per day between summer and winter; Withings user data shows roughly 500 steps between the best and worst season. Your own gap sits somewhere in that range, and only your history can pin it down.
Is temperature or daylight the bigger factor?
Both matter, but temperature shows the cleaner relationship. The Yokohama study describes a curve peaking between 19.4 and 20.7 °C. And the Icelandic study shows that a seven-and-a-half-hour swing in daylight produces only about a 19% difference in activity — considerably less than intuition suggests.
Does the fall clock change reduce my step count?
Not the total. The study published in Nature Health in April 2026 found no significant effect on daily step counts, in either direction of the transition. It did measure a shift within the day: roughly 202 more steps in the morning and 180 fewer in the evening after the fall change.
Why would fall be worse than winter?
It is not true everywhere, but it is what the Withings seasonal ranking shows, with fall landing 115 steps below winter. One plausible explanation the authors offer: New Year's resolutions revive activity in January and February, while fall has no motivational equivalent. The broader lesson is that the ordering of seasons is not universal.
Should I hold the same step goal all year?
A fixed target turns a normal seasonal swing into four consecutive months of failure. Comparing each season against the same season a year earlier is both fairer and more encouraging — and it is the only comparison that genuinely separates your effort from the thermometer.
Reading your own curve
The season is not a fate, but it is not a character flaw either. It is a variable. Once you place it next to your step count, it stops being a vague excuse and becomes a usable reference point.
Ready to find out what your year really says about how you move? Kantise brings your activity data and your local weather into a single dashboard, without asking you to spend your evenings on it.
Sources
- Hamilton S. L., Clemes S. A., Griffiths P. L. (2008). "UK adults exhibit higher step counts in summer compared to winter months." Annals of Human Biology, 35(2). doi.org/10.1080/03014460801908058
- Hino K., Lee J. S., Asami Y. (2017). "Associations between seasonal meteorological conditions and the daily step count of adults in Yokohama, Japan." Preventive Medicine Reports, 8. pubmed.ncbi.nlm.nih.gov/28831368
- Turrisi T. B. et al. (2021). "Seasons, weather, and device-measured movement behaviors: a scoping review from 2006 to 2020." International Journal of Behavioral Nutrition and Physical Activity, 18(24). PMC7863471
- Arnardóttir N. Y. et al. (2017). "Comparison of Summer and Winter Objectively Measured Physical Activity and Sedentary Behavior in Older Adults: AGES-Reykjavik Study." International Journal of Environmental Research and Public Health, 14(10). PMC5664769
- Jeong H., Katta S., Wang W. K., Volfovsky A., Dunn J. (2026). "Impact of daylight saving time on physical activity patterns." Nature Health. PMC13249386
- Withings (2015). "How do our activity levels shift throughout the seasons?" Withings Blog. withings.com
Kantise is an observation tool, not a medical device. The information in this article is provided for general purposes and does not replace the advice of a healthcare professional.