Abstract:Accelerometers are widely used to measure physical activity behaviour, including in children. The traditional method for processing acceleration data uses cut points to define physical activity intensity, relying on calibration studies that relate the magnitude of acceleration to energy expenditure. However, these relationships do not generalise across diverse populations and hence they must be parametrised for each subpopulation (e.g., age groups) which is costly and makes studies across diverse populations and over time difficult. A data-driven approach that allows physical activity intensity states to emerge from the data, without relying on parameters derived from external populations, offers a new perspective on this problem and potentially improved results. We applied an unsupervised machine learning approach, namely a hidden semi-Markov model, to segment and cluster the raw accelerometer data recorded (using a waist-worn ActiGraph GT3X+) from 279 children (9–38 months old) with a diverse range of developmental abilities (measured using the Paediatric Evaluation of Disability Inventory–Computer Adaptive Testing measure). We benchmarked this analysis with the cut points approach, calculated using thresholds from the literature which had been validated using the same device and for a population which most closely matched ours. Time spent active as measured by this unsupervised approach correlated more strongly with PEDI-CAT measures of the child's mobility (R 2 : 0.51 vs 0.39), social-cognitive capacity (R 2 : 0.32 vs 0.20), responsibility (R 2 : 0.21 vs 0.13), daily activity (R 2 : 0.35 vs 0.24), and age (R 2 : 0.15 vs 0.1) than that measured using the cut points approach. Unsupervised machine learning offers the potential to provide a more sensitive, appropriate, and cost-effective approach to quantifying physical activity behaviour in diverse populations, compared to the current cut points approach. This, in turn, supports research that is more inclusive of diverse or rapidly changing populations. Physical activity participation in young children has often been measured using parent reports. Accelerometry provides a more objective measurement but the traditional methods used to quantify this require calibration and struggle to generalise to diverse or rapidly changing populations such as young children. In recent years unsupervised machine learning methods have been shown to be able to segment and cluster accelerometry, allowing categories of activity intensity to emerge from a data-driven process. Here we show that an unsupervised machine learning technique (the hidden semi-Markov model) can be used to estimate categories of activity intensity in accelerometry data recorded from a diverse population of children age 9–36 months. We also show that this approach better captures the variance of movement abilities in the population than the traditional cut points approach. The hidden semi-Markov model approach provides a more effective approach for processing and analysing accelerometer data in rapidly changing and diverse populations such as young children, compared to the more traditional cut points approach. As it does not require calibration studies to incorporate new populations it has the potential to facilitate inclusion of unrepresented populations in research, as well as being less resource intensive.

Validity of Low-Cost Measures for Global Surveillance of Physical Activity in Pre-School Children: the SUNRISE Validation Study.

Prevalence and Correlates of Adherence to the Global Total Physical Activity Guideline Based on Step Counting Among 3- to 4-Year-Olds: Evidence From SUNRISE Pilot Studies From 17 Countries

Validating device-based physical activity indicators with observation physical activity indicators in Swiss pre-school children

International Comparison of the Levels and Potential Correlates of Objectively Measured Sedentary Time and Physical Activity among Three-to-Four-Year-Old Children

Cross-validation of Cut-Points in Preschool Children Using Different Accelerometer Placements and Data Axes

Examining Accelerometer Validity for Estimating Physical Activity in Pre‐schoolers During Free‐living Activity

Convergent Validity of Four Accelerometer Cutpoints With Direct Observation of Preschool Children's Outdoor Physical Activity

Pooled Analysis of Physical Activity, Sedentary Behavior, and Sleep Among Children From 33 Countries

Mobile Device Accuracy for Step Counting Across Age Groups.

An investigation into the minimum accelerometry wear time for reliable estimates of habitual physical activity and definition of a standard measurement day in pre-school children.

Objectively measured physical activity and sedentary time in youth: the International children's accelerometry database (ICAD)

Levels and Correlates of Objectively Measured Sedentary Behavior in Young Children: SUNRISE Study Results from 19 Countries

Cross‐validation of Actigraph Derived Accelerometer Cut‐points for Assessment of Sedentary Behaviour and Physical Activity in Children Aged 8‐11 Years

Objective Measurement of 24-Hour Movement Behaviors in Preschool Children Using Wrist-Worn and Thigh-Worn Accelerometers

Using unsupervised machine learning to quantify physical activity from accelerometry in a diverse and rapidly changing population

Exploring Factors Associated With Accelerometer Validity Among Ethnically Diverse Toddlers

Low-Cost Consumer-Based Trackers to Measure Physical Activity and Sleep Duration Among Adults in Free-Living Conditions: Validation Study

Cross-sectional examination of 24-hour movement behaviours among 3- and 4-year-old children in urban and rural settings in low-income, middle-income and high-income countries: the SUNRISE study protocol

Urban Versus Rural Differences in Meeting 24-H Movement Behaviour Guidelines among 3-4-year-olds: an Analysis of SUNRISE Pilot Study Data from 10 Low- and Middle-Income Countries.

Conducting Accelerometer-Based Activity Assessments in Field-Based Research

Accelerometer Thresholds for Estimating Physical Activity Intensity Levels in Infants: A Preliminary Study