Evaluating the performance of raw and epoch non-wear algorithms using multiple accelerometers and electrocardiogram recordings

Acute Response of Different High-Intensity Interval Training Protocols on Cardiac Auto-Regulation Using Wearable Device

The purpose of this study was to compare different high-intensity interval training (HIIT) protocols with different lengths of work and rest times for a single session (all three had identical work-to-rest ratios and exercise intensities) for cardiac auto-regulation using a wearable device. With a randomized counter-balanced crossover, 13 physically active young male adults (age: 19.4 years, BMI: 21.9 kg/m2) were included. The HIIT included a warm-up of at least 5 min and three protocols of 10 s/50 s (20 sets), 20 s/100 s (10 sets), and 40 s/200 s (5 sets), with intensities ranging from 115 to 130% Wattmax. Cardiac auto-regulation was measured using a non-invasive method and a wearable device, including HRV and vascular function. Immediately after the HIIT session, the 40 s/200 s protocol produced the most intense stimulation in R-R interval (Δ-33.5%), ln low-frequency domain (Δ-42.6%), ln high-frequency domain (Δ-73.4%), and ln LF/HF ratio (Δ416.7%, all p < 0.05) compared to other protocols of 10 s/50 s and 20 s/100 s. The post-exercise hypotension in the bilateral ankle area was observed in the 40 s/200 s protocol only at 5 min after HIIT (right: Δ-12.2%, left: Δ-12.6%, all p < 0.05). This study confirmed that a longer work time might be more effective in stimulating cardiac auto-regulation using a wearable device, despite identical work-to-rest ratios and exercise intensity. Additional studies with 24 h measurements of cardiac autoregulation using wearable devices in response to various HIIT protocols are warranted.

A review of the evidence for the utility of physical activity monitor use in patients with idiopathic inflammatory myopathies

Few proven therapies exist for patients with idiopathic inflammatory myopathies (IIMs), partly due to the lack of reliable and valid outcome measures for assessing treatment responses. The current core set measures developed by the International Myositis Assessment and Clinical Studies group were developed to standardize assessments of disease activity and treatment effect. None of the current measures address functional improvement in muscle weakness. Therefore, supplemental measures to more objectively assess physical activity levels and fatiguability in free-living settings are needed to assess disease activity more comprehensively. Validated physical activity monitors (PAMs) have the potential to serve as an objective functional outcome measure in clinical trials and observational studies. This review examines the current evidence for the use of body-worn PAMs in clinical settings with IIM patients. A practical overview of methods for PAM use in clinical patient populations (including measurement details and data processing) that focuses on IIM patients is also presented.

Individual and school correlates of body mass index and cardiorespiratory fitness in primary school children from the REACT project: A multivariate multilevel analysis

OBJECTIVE

This paper examines the relationship between body mass index (BMI) and cardiorespiratory fitness (CRF) using a multivariate multilevel approach and investigates the links between individual and school-related correlates with children's BMI and CRF.

METHODS

This cross-sectional sample included 1014 children (6-10 years) from 25 Portuguese primary schools. BMI was calculated, and CRF was assessed with the PACER test. Fundamental movement skills (FMS) included five object control tasks. Moderate-to-vigorous physical activity (MVPA), sleep, and sedentary time were assessed with the ActiGraph wGT3X-BT accelerometer. Socioeconomic status (SES) and school variables were also obtained. A multivariate multilevel model was used, and alpha was set at 5%.

RESULTS

BMI and CRF systematically increased with age. Most of the joint variance (94.4%) was explained at the child level, and BMI and CRF were correlated at this level (ρ = -.37). More active children demonstrated higher CRF levels and had lower BMI levels; sedentary and sleep time were not significantly associated with BMI or CRF. FMS were positively associated with CRF but were not significantly associated with BMI. Children at higher SES were more fit and had lower BMI than their peers of lower SES. Finally, school-level variables were not significantly related to BMI and CRF.

CONCLUSION

BMI and CRF had a low but statistically significant negative correlation in this sample of children. Most of the variation in BMI and CRF was explained by child-level characteristics.

Is there an association between proficiency in fundamental movement skills and mderate-to-vigorous physical activity in childhood on weekdays and weekends? The REACT project

OBJECTIVE

The present study probes into the association between children's fundamental movement skills (FMS) and moderate-to-vigorous physical activity (MVPA) during weekdays and weekends.

METHODS

This cross-sectional sample included 1014 children aged 6-10 years from the REACT project. Physical activity was assessed with accelerometry, and five FMS (stationary dribble, kick, catch, overhand throw, and underhand roll) were assessed with the digital platform Meu Educativo®. Three groups were formed based on the frequency of FMS that each child mastered: group 1 (wizard level in four or five FMS); group 2 (wizard level in two or three FMS); and group 3 (wizard level in at most one skill). Multilevel models were used to analyze the data treating children (level-1) nested within schools (level-2).

RESULTS

Compared to group 1, groups 2 (-12.9 ± 2.3 min day-1) and 3 (-23.9 ± 2.4 min day-1) were less physically active during weekdays and weekends (group 2: -14.7 ± 2.7 min day-1 and group 3: -22.4 ± 2.9 min day-1), independent of age and sex. There was a decline in MVPA during the weekend. Further, on average, boys were more active than girls, and with increasing age, children were less active.

CONCLUSION

On average, children with higher FMS levels are generally more physically active than their peers with lower FMS levels. Even though MVPA tends to decline on weekends, FMS proficiency is a significant factor in maintaining 60 min of MVPA on weekends.

Associations of obesity, movement behaviors, and socioeconomic status with fundamental movement skills in children: Results from the REACT project

OBJECTIVE

To investigate the relationship of biological characteristics (age, sex, and obesity), movement behaviors (physical activity and sedentary time), and family socioeconomic status with fundamental movement skills (FMS) in primary school children.

METHODS

This cross-sectional study sampled 1014 children (537 girls) aged 6 to 10 years from 25 primary schools in Matosinhos, north of Portugal. Five object control skills (dribbling, kicking, catching, throwing, and underarm rolling) were assessed with a categorical scale using the Meu Educativo® platform. Body Mass Index (BMI) was calculated and transformed into z-scores. Moderate-to-vigorous physical activity (MVPA) and sedentary time were monitored with accelerometry (ActiGraph wGT3X-BT) for seven consecutive days. Family socioeconomic status (SES) was obtained from the Portuguese social support system. Ordinal multilevel logistic regression was used to analyze the associations of weight status, MVPA, sedentary time and SES with FMS, adjusted for sex and age.

RESULTS

Boys (odds ratio (OR) = 6.54; 95% CI: 5.13-8.36) and older children (OR = 2.04; 95% CI: 1.85-2.26) were more likely to achieve higher FMS scores. Children with obesity (OR = 0.60; 95% CI: 0.45-0.80), those less active (OR = 0.56; 95% CI: 0.42-0.75) and children with more sedentary time (OR = 0.86; 95%CI: 0.77-0.97) were less likely to score high on FMS. Family SES was not significantly associated with FMS scores.

CONCLUSION

Primary school children's FMS are significantly related to biological and behavioral factors but not to family SES. These findings highlight the need for suitable strategies to enhance children's FMS proficiency, considering differences in these characteristics. Fostering adequate motor skill proficiency levels will assist in establishing a robust foundation for healthy lifestyles of all children.

Sleep and fundamental movement skills in primary schoolchildren: The REACT project

OBJECTIVE

Whether sleep is related to fundamental movement skills (FMS) in the pediatric population is largely unknown. The objective of this study was to examine the association between sleep characteristics (duration, efficiency, regularity) and FMS proficiency levels in school-aged children.

METHODS

This cross-sectional study included 996 children (mean age: 8.3 ± 1.2 years) from 25 of the 32 primary schools in Matosinhos, northern Portugal. Data collection occurred between January and June 2022. Sleep was assessed using an ActiGraph wGT3X-BT accelerometer worn on the wrist for 7 consecutive days. FMS proficiency levels were assessed in the schools with a new digital platform (Meu Educativo®) that evaluated five object control skills (dribble, kick, catch, throw, and underhand roll), with a total score ranging between 5 and 15. Multilevel ordinal logistic regression was used to test the associations between sleep characteristics and FMS proficiency levels. Covariates included age, sex, body mass index z-score, socioeconomic status, and moderate-to-vigorous physical activity.

RESULTS

The results showed that sleep characteristics (duration, efficiency, and regularity) were not related to FMS proficiency. Being a boy, older age, and higher moderate-to-vigorous physical activity levels were all significantly associated with better FMS proficiency levels. There were no significant sex-by-age interactions.

CONCLUSION

Sleep was not found to be related to FMS performance in children. This finding suggests that sleep is not a good correlate of FMS proficiency levels in school-aged children, and attention should be dedicated to other more important factors such as skill-learning-specific physical activity.

A multivariate multilevel approach to unravel the associations between individual and school factors on children's motor performance in the REACT project

OBJECTIVE

The aim was to (1) estimate the relationship between physical fitness (PF) and object control fundamental movement skills (FMS), (2) identify child characteristics that relate with PF and FMS, and (3) examine associations between the school environment, PF, and FMS.

METHODS

The sample included 1014 Portuguese children aged 6-10 years from the REACT project. PF was assessed via running speed, shuttle run, standing long jump, handgrip, and the PACER test. Object control FMS were assessed with stationary dribble, kick, catch, overhand throw, and underhand roll. Test performances were transformed into z-scores, and their sum was expressed as overall PF and FMS. Child-level variables included body mass index (BMI) z-scores, accelerometer-measured sedentary time and moderate-to-vigorous physical activity, and socioeconomic status (SES). School size, physical education classes, practice areas, and equipment were also assessed.

RESULTS

Approximately, 90% of the variance in object control PF and FMS was at the child level, and 10% at the school level. The correlation between PF and object control FMS was .62, which declined to .43 with the inclusion of covariates. Older, more active, and higher SES children had higher object control PF and FMS, and boys outperformed girls. BMI was negatively associated with PF but not with object control FMS. Sedentary time and number of physical education classes were not significant predictors. Most school predictors did not jointly associate with PF and object control FMS.

CONCLUSION

PF and object control FMS z-scores were moderately related. Not all child characteristics were associated with both PF and object control FMS, and their effect sizes were different. School characteristics only explained 10% of the total variation in PF and object control FMS.

Accelerometer-Measured Physical Activity, Fitness and Indicators of Cardiometabolic Risk among Rural Adolescents: A Cross-Sectional Study at 15-Year Follow-up of the MINIMat Cohort

BACKGROUND

Little is known about the relationship of physical activity (PA) and fitness with cardiometabolic risk among rural adolescents in low- and middle-income countries. Thus, we examined the associations of PA and fitness with selected cardiometabolic indicators along with potential gender-based differences in a birth cohort of rural adolescents from southeast Bangladesh.

METHODS

We utilized data from the 15-year follow-up of Maternal and Infant Nutrition Interventions in Matlab (MINIMat) cohort (n = 2253). Wrist-worn ActiGraph wGT3x-BT accelerometers were used to estimate sedentary time (ST) and PA. Fitness was assessed using: handgrip strength, standing long jump, and Chester Step Test. Anthropometric parameters, systolic blood pressure (SBP), and fasting lipid, insulin and glucose levels were measured. We calculated insulin resistance using the Homeostasis Model Assessment equation (HOMA-IR). Linear regression and isotemporal substitution models were fitted.

RESULTS

The adolescents spent 64 min/day (inter-quartile range: 50-81) in moderate-to-vigorous physical activity (MVPA). A 10-minute-per-day higher vigorous PA (VPA) was associated with: 4.9% (95% confidence interval (CI): 2.9-6.8%) lower waist circumference (WC), 3.2 mmHg (95% CI: 1.5-4.8) lower SBP, 10.4% (95% CI: 2.9-17.3%) lower TG, and 24.4% (95% CI: 11.3-34.9%) lower HOMA-IR. MVPA showed similar associations of notably smaller magnitude. Except for WC, the associations were more pronounced among the boys. Substituting ST with VPA of equal duration was associated with lower WC, SBP, triglyceride and HOMA-IR. Grip strength was favorably associated with all indicators, displaying considerably large effect sizes.

CONCLUSION

Our findings indicated beneficial roles of PA- particularly VPA- and muscular fitness in shaping cardiometabolic profile in mid-adolescence. VPA and grip strength may represent potential targets for preventive strategies tailored to adolescents in resource-limited settings.

Remote monitoring of amyotrophic lateral sclerosis using wearable sensors detects differences in disease progression and survival: a prospective cohort study

BACKGROUND

There is an urgent need for objective and sensitive measures to quantify clinical disease progression and gauge the response to treatment in clinical trials for amyotrophic lateral sclerosis (ALS). Here, we evaluate the ability of an accelerometer-derived outcome to detect differential clinical disease progression and assess its longitudinal associations with overall survival in patients with ALS.

METHODS

Patients with ALS wore an accelerometer on the hip for 3-7 days, every 2-3 months during a multi-year observation period. An accelerometer-derived outcome, the Vertical Movement Index (VMI), was calculated, together with predicted disease progression rates, and jointly analysed with overall survival. The clinical utility of VMI was evaluated using comparisons to patient-reported functionality, while the impact of various monitoring schemes on empirical power was explored through simulations.

FINDINGS

In total, 97 patients (70.1% male) wore the accelerometer for 1995 days, for a total of 27,701 h. The VMI was highly discriminatory for predicted disease progression rates, revealing faster rates of decline in patients with a worse predicted prognosis compared to those with a better predicted prognosis (p < 0.0001). The VMI was strongly associated with the hazard for death (HR 0.20, 95% CI: 0.09-0.44, p < 0.0001), where a decrease of 0.19-0.41 unit was associated with reduced ambulatory status. Recommendations for future studies using accelerometery are provided.

INTERPRETATION

The results serve as motivation to incorporate accelerometer-derived outcomes in clinical trials, which is essential for further validation of these markers to meaningful endpoints.

FUNDING

Stichting ALS Nederland (TRICALS-Reactive-II).

Does Wrist-Worn Accelerometer Wear Compliance Wane over a Free-Living Assessment Period? An NHANES Analysis

PURPOSE

Accelerometers are used to objectively measure physical behaviors in free-living environments, typically for seven consecutive days or more. We examined whether participants experience "wear fatigue," a decline in wear time day over day, during typical assessment period acquired in a nationally representative sample of 6- to 80-yr-olds in the United States.

METHODS

Participants were instructed to wear an ActiGraph GT3X+ on their nondominant wrist continuously for seven consecutive days. Participants with seven complete days of recorded data, regardless of wear status, were included in the analyses ( N = 13,649). Wear was scored with the sleep, wake, and nonwear algorithm.

RESULTS

Participants averaged 1248 ± 3.6 min·d -1 (mean ± SE) of wear over the assessment, but wear time linearly decreased from day 1 (1295 ± 3.2 min) to day 7 (1170 ± 5.3 min), resulting in a wear fatigue of -18.1 ± 0.7 min·d -1 ( β ± SE). Wear fatigue did not differ by sex but varied by age-group-highest in adolescents (-26.8 ± 2.4 min·d -1 ) and lowest in older adults (-9.3 ± 0.9 min·d -1 ). Wear was lower in evening (1800-2359 h) and early morning (0000-0559 h) compared with the middle of the day and on weekend days compared with weekdays. We verified similar wear fatigue (-23.5 ± 0.7 min·d -1 ) in a separate sample ( N = 14,631) with hip-worn devices and different wear scoring. Applying minimum wear criteria of ≥10 h·d -1 for ≥4 d reduced wear fatigue to -5.3 and -18.7 min·d -1 for the wrist and hip, respectively.

CONCLUSIONS

Patterns of wear suggest noncompliance may disproportionately affect estimates of sleep and sedentary behavior, particularly for adolescents. Further study is needed to determine the effect of wear fatigue on longer assessments.

Effects of ActiGraph's filter, epoch length and non-wearing time algorithm on step counts in people with COPD

The influence of the ActiGraph® processing criteria on estimating step counts in chronic obstructive pulmonary disease (COPD) remains uncertain. This study aimed to assess the influence of filters, epoch lengths and non-wearing time (NWT) algorithms on steps/day in people with COPD. ActiGraph GT3X+ was worn on the waist for seven days. Steps were detected using different filters (normal and low-frequency extension [LFE]), epoch lengths (15s and 60s), and NWT algorithms (Choi and Troiano). Linear mixed-effects model was applied to assess the effects of filter, epoch length, NWT algorithm on steps/day. Lin's concordance correlation and Bland-Altman were used to measure agreement. A total of 136 people with COPD (107 male; 69 ± 8 years; FEV1 51 ± 17% predicted) were included. Significant differences were found between filters (p < 0.001), but not between epoch lengths or NWT algorithms. The LFE increased, on average, approximately 7500 steps/day compared to the normal filter (p < 0.001). Agreement was poor (<0.3) and proportional bias was significant when comparing steps/day computed with different filters, regardless of the epoch length and NWT algorithm. Filter choice but not epoch lengths or NWT algorithms seem to impact measurement of steps/day. Future studies are needed to recommend the most accurate technique for measuring steps/day in people with COPD.

A standardized workflow for long-term longitudinal actigraphy data processing using one year of continuous actigraphy from the CAN-BIND Wellness Monitoring Study

Monitoring sleep and activity through wearable devices such as wrist-worn actigraphs has the potential for long-term measurement in the individual's own environment. Long periods of data collection require a complex approach, including standardized pre-processing and data trimming, and robust algorithms to address non-wear and missing data. In this study, we used a data-driven approach to quality control, pre-processing and analysis of longitudinal actigraphy data collected over the course of 1 year in a sample of 95 participants. We implemented a data processing pipeline using open-source packages for longitudinal data thereby providing a framework for treating missing data patterns, non-wear scoring, sleep/wake scoring, and conducted a sensitivity analysis to demonstrate the impact of non-wear and missing data on the relationship between sleep variables and depressive symptoms. Compliance with actigraph wear decreased over time, with missing data proportion increasing from a mean of 4.8% in the first week to 23.6% at the end of the 12 months of data collection. Sensitivity analyses demonstrated the importance of defining a pre-processing threshold, as it substantially impacts the predictive value of variables on sleep-related outcomes. We developed a novel non-wear algorithm which outperformed several other algorithms and a capacitive wear sensor in quality control. These findings provide essential insight informing study design in digital health research.

Using GPS Technologies to Examine Community Mobility in Older Adults

BACKGROUND

Objective measures of community mobility are advantageous for capturing movement outside the home. Compared with subjective, self-reported techniques, global positioning system (GPS) technologies leverage passive, real-time location data to reduce recall bias and increase measurement precision. We developed methods to quantify community mobility among community-dwelling older adults and assessed how GPS-derived indicators relate to clinical measures of physical and cognitive performance.

METHODS

Participants (n = 149; M ± standard deviation [SD] = 77.1 ± 6.5 years) from the program to improve mobility in aging (PRIMA) study, a physical therapy intervention to improve walking ability, carried a GPS device for 7 days. Community mobility was characterized by assessing activity space, shape, duration, and distance. Associations between GPS-derived indicators and cognition and physical function were evaluated using Spearman correlations.

RESULTS

In adjusted models, a larger activity space, greater duration (eg, time out-of-home), and greater distance traveled from home were correlated with better 6-Minute Walk Test performance (ρ = 0.17-0.23, p's < .05). A more circular activity shape was related to poorer performance on the Trail Making Test, Part A (ρ = 0.18, p < .05). More time out-of-home and a larger activity space were correlated with faster times on the Trail Making Test, Part B (ρ = -0.18 to -0.24, p's < .05). Community mobility measures were not associated with global cognition, skilled walking, or usual gait speed.

CONCLUSION

GPS-derived community mobility indicators capture real-world activity among older adults and were correlated with clinical measures of executive function and walking endurance. These findings will guide the design of future interventions to promote community mobility.

Generalizability and performance of methods to detect non-wear with free-living accelerometer recordings

Wearable physical activity sensors are widely used in research and practice as they provide objective measures of human behavior at a low cost. An important challenge for accurate assessment of physical activity behavior in free-living is the detection non-wear. Traditionally, heuristic algorithms that rely on specific interval lengths have been employed to detect non-wear time; however, machine learned models are emerging. We explore the potential of detecting non-wear using decision trees that combine raw acceleration and skin temperature, and we investigate the generalizability of our models, traditional heuristic algorithms, and recently developed machine learned models by external validation. The Decision tree models were trained using one week of data from thigh- and hip-worn accelerometers from 64 children. External validation was performed using data from wrist-worn accelerometers of 42 adolescents. For non-wear episodes longer than 60 min, the heuristic algorithms performed the best with F1-scores above 0.96. However, regarding episodes shorter than 60 min, the best performing method was the decision tree model including the six most important predictors with F1 scores above 0.74 for all sensor locations. We conclude that for classifying non-wear time, researchers should carefully select an appropriate method and we encourage the use of external validation when reporting on machine learned non-wear models.

How people wake up is associated with previous night's sleep together with physical activity and food intake

How people wake up and regain alertness in the hours after sleep is related to how they are sleeping, eating, and exercising. Here, in a prospective longitudinal study of 833 twins and genetically unrelated adults, we demonstrate that how effectively an individual awakens in the hours following sleep is not associated with their genetics, but instead, four independent factors: sleep quantity/quality the night before, physical activity the day prior, a breakfast rich in carbohydrate, and a lower blood glucose response following breakfast. Furthermore, an individual's set-point of daily alertness is related to the quality of their sleep, their positive emotional state, and their age. Together, these findings reveal a set of non-genetic (i.e., not fixed) factors associated with daily alertness that are modifiable.

Detecting accelerometer non-wear periods using change in acceleration combined with rate-of-change in temperature

Background

Accelerometery is commonly used to estimate physical activity, sleep, and sedentary behavior. In free-living conditions, periods of device removal (non-wear) can lead to misclassification of behavior with consequences for research outcomes and clinical decision making. Common methods for non-wear detection are limited by data transformations (e.g., activity counts) or algorithm parameters such as minimum durations or absolute temperature thresholds that risk over- or under-estimating non-wear time. This study aimed to advance non-wear detection methods by integrating a ‘rate-of-change’ criterion for temperature into a combined temperature-acceleration algorithm.

Methods

Data were from 39 participants with neurodegenerative disease (36% female; age: 45–83 years) who wore a tri-axial accelerometer (GENEActiv) on their wrist 24-h per day for 7-days as part of a multi-sensor protocol. The reference dataset was derived from visual inspection conducted by two expert analysts. Linear regression was used to establish temperature rate-of-change as a criterion for non-wear detection. A classification and regression tree (CART) decision tree classifier determined optimal parameters separately for non-wear start and end detection. Classifiers were trained using data from 15 participants (38.5%). Outputs from the CART analysis were supplemented based on edge cases and published parameters.

Results

The dataset included 186 non-wear periods (85.5% < 60 min). Temperature rate-of-change over the first five minutes of non-wear was − 0.40 ± 0.17 °C/minute and 0.36 ± 0.21 °C/minute for the first five minutes following device donning. Performance of the DETACH (DEvice Temperature and Accelerometer CHange) algorithm was improved compared to existing algorithms with recall of 0.942 (95% CI 0.883 to 1.0), precision of 0.942 (95% CI 0.844 to 1.0), F1-Score of 0.942 (95% CI 0.880 to 1.0) and accuracy of 0.996 (0.994–1.000).

Conclusion

The DETACH algorithm accurately detected non-wear intervals as short as five minutes; improving non-wear classification relative to current interval-based methods. Using temperature rate-of-change combined with acceleration results in a robust algorithm appropriate for use across different temperature ranges and settings. The ability to detect short non-wear periods is particularly relevant to free-living scenarios where brief but frequent removals occur, and for clinical application where misclassification of behavior may have important implications for healthcare decision-making.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-022-01633-6.

24-hour movement behaviour profiles and their transition in children aged 5.5 and 8 years - findings from a prospective cohort study

BACKGROUND

Time spent in movement behaviours, including physical activity (PA), sedentary behaviour (SB) and sleep, across the 24-h day may have distinct health consequences. We aimed to describe 24-h movement behaviour (24 h-MB) profiles in children and how profile membership changed from age 5.5 to 8 years.

METHODS

Children in the Growing Up in Singapore Towards healthy Outcomes (GUSTO) cohort were asked to wear an accelerometer (ActiGraph-GT3X+) on their wrist for seven consecutive days at ages 5.5 and 8 years to measure 24 h-MB patterns. Time spent in night sleep, inactivity (proxy for SB), light PA, moderate PA (MPA), and vigorous PA (VPA) per day were calculated using the R-package GGIR 2.0. Using latent profile analyses (n = 442) we identified 24 h-MB profiles, which were given animal names to convey key characteristics. Latent transition analyses were used to describe the profile membership transition from ages 5.5 to 8 years. Associations with sex and ethnicity were examined.

RESULTS

We identified four profiles, "Rabbits" (very high-MPA/VPA, low-inactivity and average-night-sleep), "Chimpanzees" (high-MPA, low-inactivity and average-night-sleep), "Pandas" (low-PA, high-inactivity and high-night-sleep) and "Owls" (low-PA, high-inactivity and low-night-sleep), among children at both time points. At ages 5.5 and 8 years, the majority of children were classified into profiles of "Chimpanzees" (51 and 39%, respectively) and "Pandas" (24 and 37%). Half of the sample (49%), particularly "Rabbits", remained in the same profile at ages 5.5 and 8 years: among children who changed profile the predominant transitions occurred from "Chimpanzees" (27%) and "Owls" (56%) profiles to "Pandas". Sex, but not ethnicity, was associated with profile membership: compared to girls, boys were more likely to be in the "Rabbits" profile (adjusted OR [95% CI]: 3.6 [1.4, 9.7] and 4.5 [1.8, 10.9] at ages 5.5 and 8 years, respectively) and less likely to be in the "Pandas" profile (0.5 [0.3, 0.9] and 0.4 [0.2, 0.6]) at both ages.

CONCLUSIONS

With increasing age about half the children stayed in the same of four 24 h-MB profiles, while the predominant transition for the remaining children was towards lower PA, higher inactivity and longer sleep duration. These findings can aid development and implementation of public health strategies to promote better health.

STUDY REGISTRATION

This study was registered on 4th August 2010 and is available online at ClinicalTrials.gov: NCT01174875 .

Feasibility of a continuous, multi-sensor remote health monitoring approach in persons living with neurodegenerative disease

BACKGROUND

Remote health monitoring with wearable sensor technology may positively impact patient self-management and clinical care. In individuals with complex health conditions, multi-sensor wear may yield meaningful information about health-related behaviors. Despite available technology, feasibility of device-wearing in daily life has received little attention in persons with physical or cognitive limitations. This mixed methods study assessed the feasibility of continuous, multi-sensor wear in persons with cerebrovascular (CVD) or neurodegenerative disease (NDD).

METHODS

Thirty-nine participants with CVD, Alzheimer's disease/amnestic mild cognitive impairment, frontotemporal dementia, Parkinson's disease, or amyotrophic lateral sclerosis (median age 68 (45-83) years, 36% female) wore five devices (bilateral ankles and wrists, chest) continuously for a 7-day period. Adherence to device wearing was quantified by examining volume and pattern of device removal (non-wear). A thematic analysis of semi-structured de-brief interviews with participants and study partners was used to examine user acceptance.

RESULTS

Adherence to multi-sensor wear, defined as a minimum of three devices worn concurrently, was high (median 98.2% of the study period). Non-wear rates were low across all sensor locations (median 17-22 min/day), with significant differences between some locations (p = 0.006). Multi-sensor non-wear was higher for daytime versus nighttime wear (p < 0.001) and there was a small but significant increase in non-wear over the collection period (p = 0.04). Feedback from de-brief interviews suggested that multi-sensor wear was generally well accepted by both participants and study partners.

CONCLUSION

A continuous, multi-sensor remote health monitoring approach is feasible in a cohort of persons with CVD or NDD.

A framework for handling missing accelerometer outcome data in trials

Accelerometers and other wearable devices are increasingly being used in clinical trials to provide an objective measure of the impact of an intervention on physical activity. Missing data are ubiquitous in this setting, typically for one of two reasons: patients may not wear the device as per protocol, and/or the device may fail to collect data (e.g. flat battery, water damage). However, it is not always possible to distinguish whether the participant stopped wearing the device, or if the participant is wearing the device but staying still. Further, a lack of consensus in the literature on how to aggregate the data before analysis (hourly, daily, weekly) leads to a lack of consensus in how to define a "missing" outcome. Different trials have adopted different definitions (ranging from having insufficient step counts in a day, through to missing a certain number of days in a week). We propose an analysis framework that uses wear time to define missingness on the epoch and day level, and propose a multiple imputation approach, at the day level, which treats partially observed daily step counts as right censored. This flexible approach allows the inclusion of auxiliary variables, and is consistent with almost all the primary analysis models described in the literature, and readily allows sensitivity analysis (to the missing at random assumption) to be performed. Having presented our framework, we illustrate its application to the analysis of the 2019 MOVE-IT trial of motivational interviewing to increase exercise.

A novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks

To date, non-wear detection algorithms commonly employ a 30, 60, or even 90 mins interval or window in which acceleration values need to be below a threshold value. A major drawback of such intervals is that they need to be long enough to prevent false positives (type I errors), while short enough to prevent false negatives (type II errors), which limits detecting both short and longer episodes of non-wear time. In this paper, we propose a novel non-wear detection algorithm that eliminates the need for an interval. Rather than inspecting acceleration within intervals, we explore acceleration right before and right after an episode of non-wear time. We trained a deep convolutional neural network that was able to infer non-wear time by detecting when the accelerometer was removed and when it was placed back on again. We evaluate our algorithm against several baseline and existing non-wear algorithms, and our algorithm achieves a perfect precision, a recall of 0.9962, and an F1 score of 0.9981, outperforming all evaluated algorithms. Although our algorithm was developed using patterns learned from a hip-worn accelerometer, we propose algorithmic steps that can easily be applied to a wrist-worn accelerometer and a retrained classification model.

Discrimination of wear and non-wear in infants using data from hip- and ankle-worn devices

INTRODUCTION

A key component to analyzing wearable sensor data is identifying periods of non-wear. Traditionally, strings of consecutive zero counts (e.g. >60-minutes) are identified indicating periods of non-movement. The non-movement window length is then evaluated as wear or non-wear. Given that non-movement is not equivalent to non-wear, additional criteria should be evaluated to objectively identify periods of non-wear. Identifying non-wear is especially challenging in infants due to their sporadic movement, sleep frequency, and proportion of caregiver-generated movement.

PURPOSE

To use hip- and ankle-worn ActiGraph wGT3X-BT (wGT3X-BT) data to identify non-wear in infants.

METHODS

Fifteen infant participants [mean±SD; age, 8.7±1.7 weeks (range 5.4-11.3 weeks); 5.1±0.8 kg; 56.2±2.1 cm; n = 8 females] wore a wGT3X-BT on the hip and ankle. Criterion data were collected during two, 2-hour directly observed periods in the laboratory. Using raw 30 Hz acceleration data, a vector magnitude and the inclination angle of each individual axis were calculated before being averaged into 1-minute windows. Three decision tree models were developed using data from 1) hip only, 2) ankle only, and 3) hip and ankle combined.

RESULTS

The hip model classified 86.6% of all minutes (wear and non-wear) correctly (F1 = 75.5%) compared to the ankle model which classified 90.6% of all minutes correctly (F1 = 83.0%). The combined site model performed similarly to the ankle model and correctly classified 90.0% of all minutes (F1 = 80.8%).

CONCLUSION

The similar performance between the ankle only model and the combined site model likely indicates that the features from the ankle device are more important for identifying non-wear in infants. Overall, this approach provides an advancement in the identification of device wear status using wearable sensor data in infants.

Patient-centred measurement of recovery from day-case surgery using wrist worn accelerometers: a pilot and feasibility study

This pilot and feasibility study evaluated wrist-worn accelerometers to measure recovery from day-case surgery in comparison with daily quality of recovery-15 scores. The protocol was designed with extensive patient and public involvement and engagement, and delivered by a research network of anaesthesia trainees. Forty-eight patients recruited through pre-operative assessment clinics wore wrist accelerometers for 7 days before (pre-operative) and immediately after elective surgery (early postoperative), and again at 3 months (late postoperative). Validated activity and quality of recovery questionnaires were administered. Raw accelerometry data were archived and analysed using open source software. The mean (SD) number of valid days of accelerometer wear per participant in the pre-operative, early and late postoperative periods were 5.4 (1.7), 6.6 (1.1) and 6.6 (1.0) days, respectively. On the day after surgery, Euclidian norm minus one (a summary measure of raw accelerations), step count, light physical activity and moderate/vigorous physical activity decreased to 57%, 47%, 59% and 35% of baseline values, respectively. Activity increased progressively on a daily basis but had not returned to baseline values by 7 days. Patient questionnaires suggested subjective recovery by postoperative day 3 to 4; however, accelerometry data showed that activity levels had not returned to baseline at this point. All activity measures had returned to baseline by 3 months. Wrist-worn accelerometery is acceptable to patients and feasible as a surrogate measure for monitoring postoperative recovery from day-case surgery. Our results suggest that patients may overestimate their rate of recovery from day-case surgery, which has important implications for future research.