Three new studies on step counts agree on one thing: the number alone is not enough

The step count is the most widely read health number in the world, and it is the one consumer devices are built to maximise. Three studies published in the past ten days, all using wrist-worn accelerometers rather than self-report, converge on a less comfortable point: the daily total is a weak summary of what the underlying movement is doing.

Steps do not buy back sleep

The largest is a prospective analysis of UK Biobank participants who wore a wrist accelerometer for seven days between 2013 and 2015, published in eClinicalMedicine on 29 January [s1]. Open-source machine learning algorithms derived daily step count and overnight sleep duration; incident major adverse cardiovascular events — cardiovascular death, non-fatal myocardial infarction or stroke, or revascularisation — were identified through electronic health record linkage [s1].

Among 88,012 participants with a mean age of 62.2 years (SD 7.8), 3,817 experienced a MACE over a median 7.9 years of follow-up (IQR 7.3–8.4) [s1]. Because there are no universally accepted cut-points for device-measured step count or sleep, the authors divided each exposure into tertiles and rounded the boundaries: step count as low (under 7,500), intermediate (7,500–11,000) and high (over 11,000); sleep as short (under 6.5 hours), intermediate (6.5–7.5 hours) and long (over 7.5 hours) [s1].

Low step count and short sleep were each independently associated with higher MACE risk, and there was no evidence of interaction between them (P for interaction = 0.42) [s1]. Compared with the reference group — high step count and intermediate sleep — the highest risk was in participants with both low step count and short sleep, at a hazard ratio of 1.84 (95% CI 1.62–2.10) [s1].

The authors' interpretation is the useful part: a higher daily step count does not fully attenuate the elevated cardiovascular risk associated with short sleep [s1]. Two exposures, additive rather than compensatory. This is an observational cohort, so it establishes association rather than causation, and the accelerometer wear period was a single week used to characterise years of subsequent risk.

The bursts inside the total

A separate question is whether the intensity distribution hidden inside a step count carries information the total discards. A wearables-based national cohort study published in the International Journal of Behavioral Nutrition and Physical Activity on 30 January examined vigorous intermittent lifestyle physical activity — short bursts of vigorous movement accrued through ordinary daily life rather than structured exercise [s2].

The sample was 3,293 US adults from the 2011–14 National Health and Nutrition Examination Survey who reported no participation in structured exercise (52.3% female, mean age 50.7, SD 16.6) [s2]. Bouts lasting up to one minute were measured with a wrist-worn accelerometer, and 290 deaths occurred over a mean 6.7 years (SD 1.4) [s2].

The dose–response curve was L-shaped. Relative to zero bouts per day, the sample median of 5.3 bouts per day was associated with a 44% lower risk of all-cause mortality (HR 0.56; 95% CI 0.39–0.82), with the curve flattening beyond roughly 8 bouts per day (HR 0.46; 95% CI 0.28–0.77) [s2].

The authors are unusually direct about the weakness of their own finding. When adults with prevalent cardiovascular disease or cancer at baseline were excluded (n = 2,731, 152 events), the dose–response relationship was similar but the confidence intervals crossed unity across most of the curve — for example, at the median frequency, HR 0.68 (95% CI 0.41–1.11) [s2]. Their stated conclusion is that the observed relationships may be driven at least partly by reverse causation, in which existing disease reduces both movement and survival, and that longer follow-up and larger samples are needed [s2].

In older adults, the number mostly reflects age

The third study is the smallest and the most deflationary about the metric itself. Published in Epidemiologia on 5 February, it followed 668 adults in Spain aged 60 to 100 (mean 71.33, SD 8.11) who wore a consumer accelerometer — a Xiaomi Mi Band 4 — continuously for seven days, alongside physical and cognitive testing and questionnaires on depression, quality of life and physical activity [s3].

Men averaged 8,919.08 (SD 4,455.65) steps per day, significantly more than women at 7,855.46 (p = 0.002), and a moderate negative correlation was found between age and daily step count (r = −0.460) [s3]. Meeting age-specific step recommendations was associated with better anthropometric, psychosocial and cardiometabolic markers — but, in the authors' words, many of those differences disappear after adjusting for age and sex [s3].

Their conclusion is that daily step count may not adequately assess health in older adults, and that walking intensity should be considered for accurate evaluation [s3]. This is a cross-sectional study in a single country with a modest sample, so it describes a measurement problem rather than proving one.

What this changes for a reader

Not much about walking, which all three studies treat as beneficial. What shifts is the interpretation of the number a watch displays. The UK Biobank analysis indicates that the step figure cannot be read as compensation for a short night [s1]. The NHANES analysis indicates that a total which contains occasional hard minutes may not be equivalent to the same total accrued at a uniform amble — while conceding that its own effect sizes may partly reflect illness rather than protection [s2]. And the Spanish cohort indicates that in older adults, much of what distinguishes a high step count from a low one is age and sex rather than health status [s3].

None of these is a training instruction, and none of them establishes cause. They are three arguments, from three designs, that the summary statistic on the wrist is a lossy compression of the thing it is standing in for.

What to watch

Whether device-measured cut-points start replacing the self-report-derived thresholds that public health guidance still uses — a gap the UK Biobank authors flag explicitly as their reason for constructing tertiles rather than adopting existing targets [s1]. And whether intensity metrics accumulate enough follow-up to separate protection from reverse causation [s2].

Sources

1. [s1] Joint associations of device-measured step count and sleep duration with incident major adverse cardiovascular events: prospective analysis of the UK Biobank. eClinicalMedicine, 29 January 2026. https://doi.org/10.1016/j.eclinm.2026.103769
2. [s2] Vigorous intermittent lifestyle physical activity (VILPA) and mortality risk among US adults: a wearables-based national cohort study. International Journal of Behavioral Nutrition and Physical Activity, published online 30 January 2026. https://doi.org/10.1186/s12966-026-01876-2
3. [s3] Limitations of Daily Step Count for Assessing Health in Older Adults: The Need to Consider Walking Intensity. Epidemiologia, published online 5 February 2026. https://doi.org/10.3390/epidemiologia7010024