Document resource
To date, public health guidance on physical activity, sedentary behaviour and sleep has been primarily informed by studies that relied on self-report methods (mostly recall questionnaires) to characterise and quantify behaviour. This class of evidence has established physical inactivity as an important risk factor for non-communicable diseases that is served by guidelines, interventions and policy. However, it is widely recognised that self-reported data suffer from numerous limitations, including recall and social desirability biases as well as poor validity and precision. These limitations can lead to an underestimation of effect sizes. In addition, questionnaires on physical activity and sedentary behaviour restrict data coverage to limited parts of the 24-hour day, typically to structured and planned activities. Self-reported sleep information also suffers from low-resolution, often reported in hourly increments with no indication of sleep quality or timing. These limitations reduce the utility of such data for characterising the full 24-hour profile for integrated holistic behavioural interventions and precision medicine. Wearable devices, such as accelerometers, represent a significant leap forward as they overcome many of these limitations. Despite wearables being introduced in cohort studies over 20 years ago, 1 their contribution to public health guidelines has been very limited to date. For example, such research contributed only a small fraction (estimated to be 5–10%) of the evidence synthesised in the WHO 2020 Physical Activity and Sedentary Behaviour Guidelines,2 prompting the Guidelines Development Group to flag the need for a gradual transition from self-report to wearables-based guidelines.3 The surprisingly small contribution of wearables-based evidence to guidelines is, in part, due to methodological heterogeneity within the evidence base and the slow progress with developing universal standards. Except for two children’s consortia, the International Children’s Accelerometry Database and the Sleep and Activity Database for the Early Years, which were limited to devices worn on the waist,4 5 no other epidemiological consortium of studies with wearables data existed until recently.