A Consumer Cohort Yields Accurate Step Estimates in Young and Middle Aged Adults

Public health officials face a persistent challenge: knowing exactly how active a nation's population truly is. For decades, the standard method has been to ask people to recall their own exercise habits in surveys. While convenient, these self-reports are notoriously unreliable; people often forget details or unconsciously present themselves in a more active light than reality. To get a clearer picture, scientists have turned to devices that count steps automatically. However, a new problem has emerged. The people who buy and wear these consumer gadgets tend to be younger, wealthier, and more active than the average citizen. This creates a skewed sample, leading experts to wonder if data from millions of wearable users can ever accurately reflect the physical activity of an entire country, or if the bias is simply too great to ignore.

A team of researchers set out to test whether this massive, non-representative pool of data could still offer a truthful window into national health trends. They focused on step counts, a simple and universal measure of movement, gathered from over seven million Fitbit users across 28 countries. The team compared these digital records against official government statistics from Japan and the United Kingdom, where national health surveys have long tracked physical activity using different methods. In Japan, the government has used waist-worn pedometers since the mid-1990s, while a major study in England recently equipped older adults with wrist-worn sensors similar to consumer devices. By aligning these different data sources, the researchers could see if the consumer numbers matched the official counts, and if the trends over time moved in the same direction.

The results revealed a story of surprising accuracy mixed with clear limitations. When the researchers looked at younger and middle-aged adults, the consumer data aligned closely with national estimates. In Japan, for adults between the ages of 20 and 50, the average daily step count from Fitbit users was almost identical to the national average, differing by less than one percent. Similarly, in England, the data for adults aged 50 to 69 showed a very small gap, with the consumer numbers hovering just slightly above the official counts. This suggests that for the bulk of the working-age population, the collective movement of millions of gadget users can serve as a reliable proxy for the nation's activity levels, even if the users themselves are not a perfect mirror of the general public.

However, the accuracy dropped significantly when the researchers examined older adults. In both Japan and England, the step counts from the consumer devices were substantially higher than the national estimates for people over 70. In Japan, the estimate for those aged 70 and older was nearly 16 percent higher than the official count, and in England, the figure for those over 80 was more than 20 percent too high. This discrepancy likely stems from the fact that older adults are less likely to own or wear these devices, and those who do wear them may be more active than their peers who do not. The researchers also noted that the type of device mattered; when the national survey in England used a wrist-worn sensor similar to the Fitbit, the gap between the two data sources narrowed, suggesting that differences in how the devices are worn and how they count steps contribute to the error.

Beyond comparing single snapshots of data, the team investigated whether these consumer records could track changes in activity over time. They looked at trends across 28 countries, comparing the rise and fall of daily steps recorded by Fitbit users against long-term shifts in how many people reported meeting physical activity guidelines. The two sets of data moved in lockstep. In countries where the number of people meeting activity goals increased over a twelve-year period, the average daily steps recorded by the devices also rose. Where activity levels stagnated or fell, the device data reflected that decline as well. This strong correlation indicates that even if the absolute number of steps is slightly off for certain groups, the direction of the trend is accurate enough to signal whether a population is becoming more or less active.

The study concludes that while consumer wearable data cannot yet replace official national surveys for every demographic, it offers a powerful tool for monitoring health trends in younger and middle-aged adults. The researchers found that in countries with high device adoption, these continuous streams of data could fill the gaps between expensive, infrequent government surveys. If a nation waits seven years between official health checks, the consumer data could provide a more accurate estimate of current activity levels than simply assuming the numbers from a decade ago are still valid. However, the study also highlights a significant equity gap: the countries where this method works best are predominantly in the Global North, where device ownership is high. In many other parts of the world, the lack of consumer data leaves a blind spot in global health monitoring, reminding us that while technology offers new ways to see the world, it does not yet see everyone.