Benjamin D. Maylor

Purpose: Individuals with paraplegia often have high levels of sedentary behaviour, which increases cardiovascular disease risk. This study assessed the acceptability, safety and feasibility of a sedentary behaviour intervention (REACH-SCI) in individuals with paraplegia. Materials and methods: A pre-post feasibility study was conducted in wheelchair-users with paraplegia (n=18) to evaluate an 8-week intervention comprised of a wearable activity tracker, educational booklet, goal setting, motivational support sessions, peer support and activity tools (Trial registration: NCT06957483 ). Acceptability was assessed via semi-structured interviews and questionnaires, safety via adverse events, pain and fatigue, and intervention feasibility via components being delivered as intended. Intervention engagement and preliminary efficacy were also assessed. Results: The intervention and study procedures were acceptable and safe, intervention components were delivered as planned, and the study was feasible in terms of participant recruitment, retention and data completion rates. Engagement with intervention components was generally high, except for peer support. The intervention demonstrated preliminary efficacy for improving sedentary behaviour, physical activity, cardiovascular disease biomarkers and psychosocial outcomes. Conclusions: An intervention to reduce sedentary behaviour in individuals with paraplegia was acceptable, safe and feasible. The effectiveness of REACH-SCI should be evaluated in a definitive trial to inform clinical care and public health guidelines.

BACKGROUND Using wearable devices to measure step count is important but step detection algorithms differ across device placements and are typically proprietary rather than open source. Moreover, due to a lack of openly available datasets, most validation studies of these algorithms have been conducted in laboratory settings, frequently with small sample sizes and without a ground-truth reference measure of step count. OBJECTIVE STEPCAM aims to 1) generate the largest openly available dataset of step count in a free-living environment to date; 2) evaluate step detection algorithms for wrist-, thigh-, and ankle-worn devices against camera-derived ground truth; and 3) examine differences in mean daily step count between different device placements. METHODS This will be a cross-sectional study, aiming to recruit 60 participants aged 18 years or older. Participants will wear devices on both wrists, dominant thigh, and dominant ankle for one full 24-hour period. During waking hours, a foot-facing video camera will also be worn to provide a ground-truth measure of step count. Model performance will be assessed using accuracy, recall, specificity, precision, F1 score, and balanced accuracy. RESULTS Recruitment began in May 2025, with twenty-four participants having completed data collection as of September 2026. The study is ongoing and expected to conclude in 2026. CONCLUSIONS STEPCAM will collect a dataset compromising more than 700 hours of free-living step count data from healthy adults, which will be made openly available upon study completion. Furthermore, this study will be the first to perform a thorough evaluation of step count algorithms from multiple device placements using a ground-truth reference measure.

Objective Wrist-worn accelerometers are common in large-scale epidemiological studies, but their ability to measure sedentary behaviour in free-living environments is unknown. We therefore aimed to evaluate the accuracy of openly-available methods to infer sedentary time from wrist-worn accelerometers. Methods We analysed data from 662 working-age adults in the SMART Work & Life study (20–70 years; mean age 45 years; 72% female) who concurrently wore wrist- and thigh-worn accelerometers for up to eight free-living days. Reference measurements of sedentary time were derived from the thigh accelerometer data using proprietary algorithms. Wrist accelerometer data were processed using widely used, publicly available activity recognition models. Performance was evaluated at 30-second epochs to generate per-participant metrics, alongside comparisons of mean daily sedentary time, mean daily number of prolonged sedentary bouts (≥ 30 minutes) and proportion of sedentary time in prolonged bouts. Model performance was examined across subgroups defined by age, sex, body mass index, season, recruitment centre, and in sensitivity analyses restricted to daytime hours (08:00–22:00). Results The best performing machine learning model (Actinet) accurately classified sedentary time from wrist-worn accelerometer data with a mean per-participant accuracy of 0.87 and F1 score of 0.85. Cut point-based approaches demonstrated lower accuracy of 0.80 (F1 score of 0.79). The ActiNet machine learning model showed strong agreement in daily sedentary time, daily number of prolonged sedentary bouts and proportion of sedentary time in prolonged bouts, all within 10% of the free-living thigh reference. Findings were consistent across subgroups and in analyses restricted to daytime hours. Conclusion Wrist-worn accelerometers can provide accurate measurements of sedentary behaviour in free-living settings, when assessed using current machine learning models, particularly ActiNet. This work provides confidence in future epidemiological research to examine sedentary behaviour patterns from wrist-worn accelerometers and their associations with health outcomes.

Background : Exercise snacks are a novel strategy to promote vigorous activity in a time-efficient manner. However, quantifying exercise snacks in the real world poses a challenge. Methods : We performed secondary analyses of accelerometer data from an exercise snacks trial involving 37 participants (68% female/32% male, M _age ± SD : 54 ± 6 years) who were randomized to a progressive exercise snacks group ( n = 18) or a light-intensity comparator group ( n = 19) for 12 weeks. The intervention was delivered using an online platform that provided video instructions and notification reminders. Both groups were instructed to perform a minimum of three daily bouts lasting 30–60 s (depending on the phase of the intervention) on at least 3 days per week. Thigh-worn accelerometers captured activity at baseline (Week 0), Week 2, and Week 11. Raw accelerometer files were processed using GGIR to quantify daily average acceleration, intensity gradient, time accumulated in 40-m g incremental acceleration bins from 0 to >400 m g , and MX metrics, defined as the lowest acceleration above which the most active, but not necessarily continuous, X minutes of the day are accumulated. Results : At Week 11, when the progressive exercise snacks intervention included movements that were more vigorous, this group showed a higher value (effect estimate in milligravitational units) for the M1 (36 m g , 95% CI [1, 72], p = .044) metric when compared with the light-intensity group. Additionally, when using previously validated cut points, the exercise snacks group increased time (percent difference accumulated at vigorous intensities (3%, 95% CI [1, 5]) when compared with the light-intensity comparator group. Conclusions : Exercise snacks appear to elicit measurable changes in real-world activity metrics, as assessed using thigh-worn accelerometers.

Background: Wearable devices enable continuous measurement of physical activity, sedentary behavior, sleep, and heart rate under free-living conditions. However, most validation studies rely on small, homogeneous samples; are conducted under laboratory conditions; or lack gold standard ground-truth measurements, limiting the generalizability and accuracy of derived metrics. There is a pressing need for open-access, large-scale, free-living validation datasets that include multisensor data from diverse body locations and participant demographics to aid in model development. Objective: The Oxford Wearable ECG, Activity, Circadian Rhythm, and Sleep Validation Study (OxWEARS) aims to (1) validate accelerometer-based measurement of physical behaviors across 5 body sites against annotated camera data; (2) validate measurements of sleep and sleep staging from 5 different body sites against polysomnography; (3) validate wrist-worn photoplethysmography heart rate measurements against chest-worn electrocardiogram; and (4) generate a comprehensive, annotated, and anonymized dataset for open-access research use. Methods: This cross-sectional study will recruit approximately 160 adults (aged ≥40 years) stratified by age, sex, and BMI from the Oxford BioBank. Over 3 days and 4 nights, participants will wear sensors on the wrists, chest, hip, thigh, and ankle. Ground-truth measures will be obtained from a chest electrocardiogram patch for heart rate, a first-person camera for activity annotation, an ankle-worn accelerometer for step count, and at-home polysomnography for sleep. An under-mattress sensor will collect measures of sleep, respiration rate, and bedtime, and a subjective sleep diary will also be obtained. Signals from different wear locations will be compared against the ground truth using precision, recall, F ₁ -score, κ, and agreement metrics. Results: Recruitment commenced in November 2024, with 15 participants enrolled by May 2025. Overall, 50% of eligible adults contacted were happy to consent to the study, with excellent compliance with the protocol observed to date. Data collection is ongoing and expected to conclude in 2026, with the final annotated dataset made publicly available as soon as possible thereafter. Conclusions: The OxWEARS study will generate an openly accessible dataset containing more than 10,000 annotated hours from a stratified sample of adults. This will directly support scalable, generalizable human activity recognition efforts, while also enabling robust development and benchmarking of wearable-derived health metrics.

Background: Controlled laboratory studies have demonstrated that breaking up sitting can reduce postprandial glucose in South Asian adults. This study examined the effects of substituting sitting with standing and walking on interstitial glucose in South Asian individuals under free-living conditions. Methods: South Asian adults ( n = 14 [50% male]; body mass index 26.5 ± 0.8 kg·m ⁻² ) aged 41 ± 3 years completed two, 4-day regimens in a counter-balanced order: (1) SIT (restrict walking and standing to ≤ 1 h/day each) and (2) SITless (substitute ≥ 5 h/day of sitting with ≥ 3 h of standing and ≥ 2 h of walking, and interrupt sitting every 30 min). Interstitial glucose was measured using Flash glucose monitoring. Sitting and physical activity were measured with the activPAL3. Outcomes were compared between regimens using linear mixed models. Results: Interstitial glucose net incremental area under the curve (iAUC) for waking hours was lower by − 9.2 mmol L ⁻¹ ·16 h ⁻¹ (95% Confidence Interval [CI]: − 18.1, − 0.3) in SITless than SIT ( p = 0.04), while lunch postprandial glucose iAUC was significantly lower by -1.0 mmol L ^−1. 2 h ⁻¹ (95% CI − 1.8, 0.2) in SITless ( p = 0.02). There were no significant differences in other 24 h or 16 h glucose metrics ( p ≥ 0.06). Compared to SIT, sitting was lower by − 3.6 h/day (95% CI − 4.9, − 2.3) in SITless ( p < 0.01). Standing and stepping time were higher by 1.9 h/day (95% CI 0.6, 3.2) and 1.6 h/day (95% CI 1.2, 2.1) in SITless ( p ≤ 0.01). Conclusions: Substituting sitting with standing and walking under free-living conditions can be used to effectively attenuate glycaemia during waking hours, but not across 24 h, in South Asian adults. Clinical trial registration: NCT04645875..

Background Many adults in high-income countries carry a device capable of measuring physical- activity behaviour. Thus, there is public health need to understand whether such data can enhance prediction of future health outcomes. We aimed to investigate whether device-measured daily-step count and walking cadence improve the prediction of mortality beyond traditional risk-factors. Methods Risk models were developed to predict five-year all-cause mortality using data from the UK Biobank accelerometer sub-study, with external validation in the US 2011-2014 National Health and Nutrition Examination Survey (NHANES). Median daily-step count and peak one-minute walking cadence were derived using self-supervised machine learning models from seven-day wrist-worn accelerometer data. Cox models were used to develop a baseline model incorporating traditional risk- factors, and a baseline model plus accelerometer data (i.e. daily-steps and walking cadence). Changes in model performance were assessed using Harrell’s C-index, net reclassification index (NRI; 10% threshold), and the Nam-D’Agostino calibration test. Findings Among 79,717 UK Biobank participants, 1,640 died within 5-years. Adding accelerometer data to the baseline model modestly improved risk discrimination and classification with a change in c- index of 0.008 (95% confidence interval [CI] 0.005-0.011) and 3.3% NRI (95%CI 2.1%-4.5%). Greatest improvements in prediction were observed in participants with prior disease at baseline, showing a change in c-index of 0.028 (95%CI 0.019-0.039) and 5.9% NRI (95%CI 3.1%-8.6%). In the NHANES external validation cohort (n=4,713; deaths=378), similar improvements in prediction were observed (change in c-index: 0.015, 95%CI 0.007-0.025; NRI: 4.0%, 95%CI 0.7%-7.4%). All models were well calibrated (Nam-D’Agostino χ² range: 6.8-13.2). Interpretation Device-measured daily-step count and walking cadence consistently demonstrated modest improvements in predicting mortality risk beyond traditional risk-factors, with the most significant enhancements seen in individuals with prior disease. These findings suggest that incorporating information from wearables does provide important new ways to improve risk stratification for targeted intervention in high-risk individuals.

Background: A novel ‘whole day’ approach that could motivate the public to be more physically active is Snacktivity™. The Snacktivity™ approach encourages individuals to accumulate 150 min of physical activity in short 2–5-min ‘snacks’ of moderate-vigorous intensity physical activity (MVPA) throughout the day/week. Method: A randomised controlled trial to assess the feasibility/acceptability of a Snacktivity™ intervention and trial processes was conducted. The trial aimed to recruit 80 physically inactive adults from healthcare services and via social media. Participants were randomised to the Snacktivity™ intervention or usual care and followed up at 12 weeks. The intervention was predominately delivered by health professionals within consultations. Assessment of whether the Snacktivity™ intervention and trial methods were acceptable to participants, adherence to Snacktivity™ (assessed by Fitbit) and physical activity (assessed by accelerometer), and retention were considered according to traffic light stop-go progression criteria (green-amber-red). Results: Seventy-two participants ( n = 37 Snacktivity™ intervention; n = 35 usual care) were recruited across 14 months (72/80, 90%, (green) 95% CI: 83% to 97%). Snacktivity™ adherence was achieved in 12/37 participants (32%, (red) 95% CI: 17% to 48%). Physical activity adherence was achieved in 17/37 participants (46%, (amber) 95% CI: 30% to 62%). Seven participants (10%, (green) 95% CI: 3% to 17%) withdrew from follow-up and 25/72 (35%, (amber) 95% CI: 24% to 46%) had no accelerometer data at follow-up (retention). Conclusion: The Snacktivity™ intervention may be feasible and acceptable to implement. Findings can inform subsequent research that seeks to investigate whether Snacktivity™ based approaches are effective in promoting physical activity in the population. Trial Registration: ISRCTN: 64851242. Registration date: 31/01/21.

Objectives: To describe age-related differences in the absolute and relative intensity of physical activity (PA) and associations with mortality.Methods: UK Biobank participants with accelerometer-assessed PA (mg) and fitness data (N=11 463; age: 43–76 years) were included. The intensity distribution of PA was expressed in absolute and relative terms. The outcome was mortality.Results: PA volume (average acceleration) and absolute intensity were lower with increasing age (~−0.03 to −0.04 SD of mean value across all ages per year; p<0.001) but differences in relative intensity by age were markedly smaller in women (−0.003 SD; p<0.184) and men (−0.012 SD; p<0.001). Absolute intensity was higher in men, but relative intensity higher in women (p<0.001). Over a median (IQR) follow-up of 8.1 (7.5–8.6) years, 121 (2.4 per 1000-person-years) deaths occurred in women and 203 (5.0 per 1000-person-years) in men. Lower risk of mortality was observed for increasing absolute or relative intensity in women, but for absolute intensity only in men. In men, the lowest risk (HR 0.62, 95% CI 0.43, 0.91) was observed in those with high absolute intensity (80th centile), but low relative intensity (20th centile). Conversely, in women, the lowest risk was associated with high levels (80th centile) of both absolute and relative intensity (HR 0.59, 95% CI 0.41, 0.86).Conclusion: Absolute PA intensity dropped with age, while relative intensity was fairly stable. Associations between PA intensity and mortality suggest that prescribing intensity in absolute terms appears appropriate for men, while either absolute or relative terms may be appropriate for women.

This study aimed to determine the impact of cab noise when driving Heavy Goods Vehicles (HGV) on cut-point estimated moderate-to-vigorous physical activity (MVPA) from wrist-worn accelerometers. First, we investigated the impact of cab noise on accelerometer output during HGV driving and then on cut-point estimated MVPA in HGV drivers. A GENEActiv accelerometer was located beneath the seat in six HGVs for 8 days. Acceleration recorded during driving lay predominantly (∼94%) within the sedentary range (g). HGV drivers (N = 386, 47.9 ± 9.3 years) wore a wrist-worn GENEActiv and a thigh-worn activPAL simultaneously for 8 days covering workdays and nonworkdays. MVPA recorded by the activPAL excludes seated transport, thus provided the criterion. Wrist accelerometer MVPA was classified using two cut-points approximating 3 metabolic equivalents (MVPA_100mg) and 4.3 metabolic equivalents (indicative of brisk walking, MVPA_{BRISK_WALK}). Acceleration classified as MVPA_100mg or MVPA_{BRISK_WALK} during activPAL-determined seated transport was considered erroneous. Across all-days, activPAL MVPA was 15 (interquartile range: 9, 26) min/day. Compared with activPAL, MVPA_100mg was 100 min/day higher (95% limits of agreement ±53 min), but MVPA_{BRISK_WALK} similar (mean bias = −2 min/day, 95% limits of agreement ±15). On workdays, 23 (interquartile range: 11, 52) min of MVPA_100mg and 2 (1, 7) min of MVPA_{BRISK_WALK} were erroneous. However, on nonworkdays, only 4 (3, 14) and 0.4 (0, 1) min, respectively, were erroneous. In conclusion, MVPA may be erroneously captured using cut-point analyses of accelerometer data in HGV drivers. However, this was substantially reduced by using an MVPA cut-point indicative of brisk walking, which also approximated activPAL estimated MVPA.