Measuring physical behaviour in children with varying abilities: From differences to data

Abstract

Physical activity is crucial for children’s physical and mental health and social development. Children who are physically active early in life are more likely to remain active later on. Therefore, the World Health Organization (WHO) developed the 24-Hour Movement Guidelines. Despite this knowledge, global data show that few children meet these recommendations. The WHO recommends using activity monitors to assess physical activity, yet little research has examined their reliability and validity in paediatric populations, particularly among children with developmental disabilities. Age-related differences and atypical movement patterns complicate assessment, and children who rely on assistive devices are often excluded from studies. Paediatric physiotherapists aim to help children be as physically active as possible within their own context. Activity monitoring can offer valuable insights into a child’s physical behaviour, but many physiotherapists struggle to interpret and apply these data in clinical reasoning and goal setting. This thesis aims to improve the assessment of physical behaviour in children with and without developmental disabilities by combining generic and specific tools with strong psychometric properties and advanced accelerometry in real-world settings. This approach supports clinical reasoning, eventually enables tailored and objective treatment, and facilitates progress tracking focused on functioning and meaningful participation. The thesis consists of two parts. Part One investigates the reliability and validity of activity monitors. A systematic review and meta-analysis (Chapter Two) on children aged 2–19 years with atypical gait showed high to very high reliability and moderate to high validity in controlled settings, but limited evidence in free-living contexts. Quality of results was moderate, and results varied depending on age, type of activity, and measurement conditions. Two validation studies (Chapters Three and Four) further evaluated an activity monitor prototype (AM-p) in real-world conditions. In ambulatory children with and without developmental disabilities (n=93), the ankle-worn AM-p distinguished stationary behaviour, cycling, and locomotion activities with 82% accuracy, performing better in older children and those without disabilities. In a second study among children using a manually propelled wheelchair (n=37), the single ankle-worn AM-p performed well in detecting stationary behaviour and moderately in distinguishing locomotion and active wheelchair use. A dual-sensor setup (ankle and arm) improved accuracy and allowed more specific classification of movement types, showing promise for use in children with diverse mobility patterns. Part Two explored user perspectives and clinical application. Semi-structured interviews (Chapter Five) with paediatric physiotherapists, parents, and children (6–19 years) emphasized user-friendliness, clear information, child-friendly design, shared decision-making, and practical applicability. Using a co-design approach, an AM-p Toolkit was developed, containing physical and digital tools and training materials to support physiotherapists in practice. Feasibility testing (Chapter Six) showed perceived benefits but also a need for technical improvements, better wearing comfort, and more efficient, clinically oriented training. This thesis provides insights into integrating activity monitoring in paediatric physiotherapy practice. Future research should refine activity categories, validate in free-living settings, and improve accessibility and interdisciplinary collaboration to maximize benefits for children’s health and participation in daily living.