Saltar al contenido
Dispositivos y mediciónEstudio de cohorte5 min read

Activity–stroke and heart attack curves differ by accelerometer processing method

Colectivo editorial MoveVitals · Study published · Reading updated

Original paper:Investigating the comparability of wearable accelerometer methods in the association between physical activity and cardiovascular disease: a cohort study using UK Biobank · Preventive Medicine · 2026 · Artículo original ↗

Idea principal

Among 90,237 UK Biobank adults free of cardiovascular disease at baseline, followed until December 2022 for 1,298 strokes and 2,031 myocardial infarctions, the shape of the link between moderate-vigorous physical activity (MVPA) and each outcome depended on how the accelerometer data were processed. For stroke, hazard fell in a straight line only with the machine-learning metric; for myocardial infarction, two metrics gave a curved decline and activity counts a straight one. The abstract gives no hazard ratios, so it shows that methods disagree, not which one is right, and the findings are observational associations.

Study design

Estudio de cohorte

Reading basis

Published abstract

Reading focus

Association & confounding

Esta explicación editorial está disponible en inglés, chino simplificado y chino tradicional. Esta interfaz muestra la lectura en inglés.

Pregunta de investigación

Wearable accelerometers let researchers relate objectively measured physical activity to disease, but different processing methods turn the same recording into different amounts of moderate-vigorous activity. The authors state that the choice of method may influence the shape of the dose-response association between wearable-measured activity and health outcomes. They set out to compare how three accelerometer-generated MVPA metrics relate to incident stroke and to incident myocardial infarction (heart attack).

Población y método

This was a cohort study in the UK Biobank accelerometer sub-cohort, recruited in the UK between 2013 and 2015. For each participant the authors computed daily MVPA duration three ways: with Low-pass Filtered Euclidean Norm Minus One (LFENMO), with a machine-learning method, and with activity counts. They analysed 90,237 participants who were free of cardiovascular disease at baseline.

The outcomes were incident stroke and incident myocardial infarction, followed up until December 2022; there were 1,298 incident strokes and 2,031 myocardial infarctions. For each of the three MVPA metrics the authors used Cox regression with a restricted cubic spline to estimate the dose-response association with each outcome. Full details are not in the abstract: it does not report the accelerometer model or wear period, participant age and sex, which factors were adjusted for, how events were identified, or any hazard ratios.

Resultados

  1. 1
    The three metrics gave different shapes for stroke. With the machine-learning metric the hazard ratio fell in a straight line as MVPA rose, but not with LFENMO or activity counts, and the abstract does not say what shape those two showed.
  2. 2
    For myocardial infarction the shapes also differed. The machine-learning and LFENMO metrics gave a curvilinear decrease in hazard ratios, meaning the decline was not constant across the range of MVPA, whereas activity counts gave a linear decrease.
  3. 3
    In total the analysis followed 90,237 people free of cardiovascular disease at baseline and counted 1,298 incident strokes and 2,031 myocardial infarctions, so the three comparisons rest on the same people and events.
  4. 4
    The authors conclude that the dose-response associations between MVPA and cardiovascular disease varied markedly across the three accelerometer-derived metrics, and that research using a single accelerometer metric should interpret its association with caution.

Límites e incertidumbre

  1. 1
    Our observation: the abstract gives no hazard ratios, confidence intervals or P values, so the size of the differences between metrics, and whether they exceed chance, cannot be judged. Curves called linear or curvilinear depend on the spline and on the number of events, which differ for stroke and myocardial infarction.
  2. 2
    Our observation: the analysis shows that the metrics disagree but has no reference standard, so it cannot say which metric measures true activity best or which curve is the correct one.
  3. 3
    Our observation: this is an observational cohort of people who volunteered for an accelerometer sub-cohort, so the associations may not describe the wider population and do not show that adding MVPA minutes lowers stroke or heart attack risk. The abstract does not mention confounding or follow-up length beyond December 2022.
  4. 4
    Our observation: this reading is based on the abstract only, so adjustments, definitions of events and wear-time rules are not covered.

Relevancia práctica

This finding is mainly a caution for researchers and for anyone who compares active minutes across apps and devices: minutes of moderate-vigorous activity from different algorithms are not interchangeable, and the apparent shape of the link between activity and stroke or heart attack changes with the algorithm. For myocardial infarction, all three metrics showed hazard ratios falling as MVPA rose; for stroke, only the machine-learning metric showed a clear straight-line decline.

Do not conclude that one metric is the right one, that a given number of minutes is a threshold, or that the minutes on your watch map onto the minutes in this study. These are associations in an observational cohort, not proof that adding activity changes your own risk, and decisions about heart or stroke risk belong with your clinician.

Tus datos

The active minutes your device or app reports come from the maker's own algorithm, which may be based on activity counts or on a model, so they are not directly comparable with another brand or with research figures. Stick to one device and one app and follow your own trend over weeks. If you change device, app or data source, mark the date, because an apparent jump or drop in MVPA may come from the method rather than from you.

Cómo interpretar la evidencia

Exposure precedes follow-up outcomes, but confounding and reverse causation can remain. A hazard ratio is not an individual absolute risk or a demonstrated intervention effect.

Fuentes y atribución

Investigating the comparability of wearable accelerometer methods in the association between physical activity and cardiovascular disease: a cohort study using UK Biobank

Preventive Medicine · 2026-08

DOI: 10.1016/j.ypmed.2026.108603 ↗PubMed · 42142761 ↗

Interpretación editorial de MoveVitals · contenido educativo · sin revisión clínica individual registrada

Fuentes verificadas 2026-11-01

Guías prácticas

Leer en contexto

·Estudio/revisión de validación

Smartwatch data as forensic evidence: heart rate is best validated, the rest uneven

A systematic review and meta-analysis of studies on smartwatches from four major brands, framed around digital forensics, found that heart rate was both the most studied and the most accurate measure. Step count, distance, energy expenditure, activity classification and sleep were less consistently validated, and results varied by brand, protocol and reporting. The abstract gives no pooled numbers, and most studies were run in healthy people in laboratories, so it cannot say how a particular watch performs for a particular person in daily life.

Forensic Science International

Explicaciones detalladasLeer estudio →
·Revisión de revisiones

Activity trackers: about 1,800 more steps a day, but small effects on other outcomes

An umbrella review of 39 systematic reviews and meta-analyses, covering 163,992 participants, found that activity trackers were associated with more physical activity (SMD 0.3–0.6), equating to about 1,800 extra steps and 40 minutes more walking per day, and with a bodyweight reduction of about 1 kg. Effects on blood pressure, cholesterol, glycosylated haemoglobin, quality of life and pain were typically small and often non-significant. The abstract does not say what else came with the trackers in these programs or how much the reviews overlap, so the gains cannot be credited to the device alone.

The Lancet Digital Health

Explicaciones detalladasLeer estudio →
·Estudio/revisión de validación

Seven sleep trackers: good at spotting sleep, weak on wake and sleep stages

In a sleep laboratory, 34 healthy young adults wore a subset of seven consumer sleep trackers for three nights. Against polysomnography every device detected sleep very well (sensitivity 0.93 or higher) but caught wake poorly (specificity 0.18–0.54), and five of the seven matched or beat research actigraphy (specificity 0.39). Sleep-stage estimates were inconsistent, and the participants were healthy adults aged 18 to 35 in a fixed 8-hour bed window, with each device tested in only 11 to 20 of them, so the results say little about people with sleep problems.

Sleep

Explicaciones detalladasLeer estudio →
·Estudio/revisión de validación

Seven wrist devices: heart rate mostly within 5% error, calorie estimates far off

Sixty adults wore seven wrist-worn devices while sitting, walking, running and cycling in a laboratory. Median heart rate error against an electrocardiogram (ECG) ran from 1.8% on the bike to 5.5% walking, but no device achieved a calorie error below 20 percent (device medians from 27.4% to 92.6%). These were consumer models on sale at the time, tested in a lab, so the results do not rank today's watches or show how they perform in daily life.

Journal of Personalized Medicine

Explicaciones detalladasLeer estudio →
Dispositivos y medición →