The present study found an international consensus over the actions required to achieve high- quality national and international surveillance of the movement behaviours (physical activity, sedentary behaviour, sleep) in children and adolescents across diverse countries around the world. The consensus was global, extended across levels of economic development, and across surveillance expertise in early childhood, later childhood and adolescence.

Increased funding at the national level for surveillance was rated as a top priority, highlighting the significant underfunding of movement behaviour surveillance in countries across all levels of economic development, not just in LMICs. While methodological improvements for surveillance were rated as high priorities these are also dependent on increased funding to a large extent. The inadequate size and representativeness of current surveillance systems could be addressed by improved funding. Increased funding will also be required to deal with the highest priority methodological problems, notably the need to develop and validate new ways of measuring screen time: 98% of the sample rated this as either extremely or very important. Respondents prioritised the need for larger and more inclusive surveys so that a proper understanding of inequalities (e.g. among children and adolescents with disability, urban versus rural populations) can be achieved. Surveillance that includes important sub-groups, such as those living with disability, indigenous populations and individuals from LMICs, can help identify and address health inequalities. There was also very high level of agreement on the need to improve measurement of a number of other variables, e.g. physically active play, compliance with muscle-strengthening guidance (Supplementary File). The need for standardised surveillance protocols that can be used across countries was also a high priority.

How the findings of the present study should be implemented is a matter for stakeholders to decide, but a number of priority actionable steps have been identified: research effort/funding should be directed to improved methodology for measurement of screen time, physical activity play and other variables (Supplementary File); advocacy should be used to increase funding for surveillance; regional or global networks for surveillance need to be formed or existing networks need long-term funding so they can be expanded/sustained; a framework for ethical use of government and non-government funding in surveillance should be developed; global capacity in movement behaviour research and surveillance should be increased by greater availability of education and training opportunities and greater opportunities to work with existing surveillance collaborations; new standardised protocols for surveillance should be agreed upon; and advocacy and partnership building at a national level should be undertaken so that the importance of the movement behaviours (for NCD prevention, for example) is understood more widely, and so that awareness of national and international movement behaviour guidelines and strategies is increased.

The current model of movement behaviour surveillance within national health surveys in many countries requires re-evaluation and upgrading [1, 3]. The common practice of assessing the time spent in movement behaviours and intensity of physical activity by using non-validated, or even demonstrably invalid [1], questionnaire items within national health surveys has many limitations. There is empirical evidence that this approach has failed in some countries, e.g. by producing misleading data on physical activity levels, with unintended adverse consequences for public health policy [1]. Alternative national and international surveillance approaches are likely to be needed. More device-based surveillance will be useful [24, 25] but device-based measurement is not a simple solution. For example, choices made when processing accelerometry data can have profound consequences for apparent levels of physical activity [3]. For some countries, informative device-based surveillance of physical activity with accelerometers is already well-established [3], and accelerometry can also measure time spent sedentary. Accelerometry cannot measure screen time but can measure time spent sleeping and accelerometer-measured sedentary time may be a proxy for screen time in the absence of more direct and accurate measures. A discussion of the advantages and disadvantages of device-based surveillance of the movement behaviours is beyond the scope of the present manuscript. In summary, devices may be essential for accurate measurement of overall time use and intensity of physical activity. Devices are useful to validate self-reports or parent-proxy reports of the movement behaviours [1, 3] but inherently lack the ability to capture qualitative information regarding the context, type, or purpose of the activity undertaken, restraining the depth of insights available for informing the development of physical activity policies. Devices alone are unlikely to provide comprehensive methods for all surveillance requirements. Devices might also complement subjective methods in surveys.

A risk of the growth of device-based surveillance is that the increased cost relative to subjective methods could lead to increased inequality in global surveillance, with reduced access to devices in LMICs relative to HICs. International movement behaviour partnerships like the SUNRISE study [20] can make device-based surveillance much more equitable. SUNRISE depends on the sharing of accelerometry equipment and expertise in accelerometer use. A major barrier to equipment sharing in the SUNRISE study, reflected by respondents in our surveys, has been the difficulties (cost, time, customs clearance, loss of or damage to devices) associated with shipping equipment across international borders and that is a practical problem future regional or global surveillance hubs would face.

Alternative models of movement behaviour surveillance that do not depend on national health surveys might also be very informative. For example, the FitBack initiative https://www.fitbackeurope.eu/en-us/ has provided valuable national and international surveillance of child and adolescent physical fitness, based on measurements made mainly in physical education classes [1, 29]. This model might be considered more widely for countries where school attendance is universal. Other non-traditional approaches to surveillance might also be helpful, e.g. greater use of ‘Citizen Science’ [30]. Citizen Science surveillance initiatives might involve community members using smartphone apps to track and report their own or their child’s physical activity, sedentary behaviour or sleep patterns. These approaches harness the collective power of diverse perspectives and large-scale data collection, potentially offering real-time, contextualized insights into movement behaviours within specific communities, combining device-measured quantitative information and self/proxy-reported qualitative information on the setting, contex and type of behaviour/activity. The validity and reliability of citizen science-derived results are bolstered through meticulous validation protocols, rigorous data quality checks and collaborative efforts between scientific experts and participants, and would require actions to develop standardised methodologies, cross-verification processes and quality assurance [30].

Regardless of the model chosen, surveillance systems should generate good quality information that can drive the development of effective actions in policy and practice [31]. A greater focus on the surveillance of specific domains and contextual factors (e.g. active transportation, active/outdoor play) might be more practical than surveillance of time spent in moderate-to-vigorous-intensity physical activity (MVPA). That type of approach would be informative for population-level monitoring of trends over time and understanding inequalities and responses to changes in policy. A shift in emphasis to surveillance of domains and contexts would also allow countries to focus on the domains that are of highest priority to them. Our respondents emphasised the importance of avoiding a narrow focus, on surveillance of MVPA only for example. They recommended that surveillance is not restricted to individual data, but that actions are required to expand the measurement and monitoring of the ‘upstream’ influences on the movement behaviours (e.g., parent/peer/school and policy environment) that is lacking in many national and international movement behaviour surveillance systems and initiatives at present [1, 3].

Respondents recommended establishing more effective national partnerships to promote child and adolescent movement behaviours in the public health and sports science/sports medicine agendas. This would include increasing policymaker and practitioner awareness and use of WHO guidelines and strategies [4, 5, 10] and more emphasis on policy implementation in relation to the movement behaviours in childhood and adolescence. These improvements will need advocacy for movement behaviours by researchers and an increase in knowledge exchange between experts in surveillance and child and adolescent movement behaviours and policymakers and practitioners in health and other relevant areas (e.g. education, childcare, transport, planning). The AHKGA Global Matrix initiative has shown that physical activity policy development is generally positive, but policy implementation and evaluation are lacking in many countries [11]. The present study shows the high priority placed by experts in surveillance on the need for greater policymaker and practitioner awareness of the importance of child and adolescent movement behaviours, particularly on issues already high on the global public health agenda: NCD prevention, child and adolescent development and the right of children to play [1].The WHO Global Status Report on Physical Activity in 2022 highlighted the need for well-implemented policy on physical activity in all countries, underpinned by robust surveillance [32], anchored to the Sustainable Development Goals.

Substantive changes in surveillance are clearly required globally, but these might be more achievable than one might think with a harmonised approach that builds on international partnerships. For example, existing international surveillance initiatives for the movement behaviours and/or related health outcomes in childhood and adolescence (e.g., SUNRISE, the AHKGA Global Matrix, FitBack) already include many of the elements required for high quality surveillance [1]. These initiatives make use of international collaboration to share expertise and measurement equipment, build human capacity in surveillance and discipline expertise and they support regional and national surveillance via regional or international knowledge exchange/research hubs/peer support networks. Initiatives of this kind could be expanded and made more sustainable by long-term funding and partnership with key existing organisations (e.g., WHO Geneva, regional WHO offices) to sustain surveillance.

The WHO ‘STEPwise approach to surveillance’ (STEPS) [33] does not use methods which would be applicable to movement behaviour surveillance in the age groups covered by the present study, but the approach taken by WHO STEPS to develop a global surveillance system might prove useful as a model for future global surveillance in children and adolescents. STEPS is a globally comparable, standardised and integrated surveillance tool by which countries can collect, analyse and disseminate core information on NCDs across a wide range of environmental and cultural settings. STEPS has succeeded largely via a combination of funding, and support/co-ordination from regional and international WHO offices, providing a good example of the kind of shared international infrastructure and collaboration required for more informative global surveillance of the movement behaviours of children and adolescents [32, 33]. Developing a STEPS-like international system will be much more challenging in some countries than others as engagement with existing WHO-based surveillance of movement behaviours and NCD risk factors varies greatly between countries, but success in some countries provide hope [34] https://www.who.int/teams/noncommunicable-diseases/surveillance/data.

Advances in surveillance will allow regional, national and global stakeholders to confidently identify and respond to trends in movement behaviours and inequalities in these behaviours [32, 35]. Improved surveillance would also allow stakeholders to assess the impact of policy change or societal change, for example, the impacts of climate change, economic crises, political instability and pandemic mitigation measures [11, 36]. Greater collaboration between experts in movement behaviours and policymakers might also prevent unintended harms from policy, or at least identify unintended policy consequences. For example, in Canada, a well-meaning policy, the ‘Child Fitness Tax Credit’, a $500 tax credit for registering a child in an eligible physical activity programme, had limited impact other than increased socio-economic inequalities in health [37], also illustrating the need for more robust evaluation of movement behaviour-related policy. The process of developing higher quality surveillance systems is not ‘just’ about monitoring as a passive data collection exercise. Surveillance tends to reveal previously unnoticed public health problems, which in turn tends to lead to new policy or more informed policy [1]. Researchers could also develop much better-informed research questions if surveillance was of higher quality.

Figure 4 illustrates how the present study findings might lead to actions which improve national and international surveillance. The foundation of the conceptual model in Fig. 4 is increased funding for surveillance. The model is not a simple linear pyramid but incorporates feedback loops. For example, more effective researcher advocacy is likely to be needed to achieve increased funding for surveillance, and to ensure that improved surveillance translates to meaningful changes in policy and practice. More advocacy for movement behaviour funding and surveillance may be uncomfortable for movement behaviour researchers, but merits greater effort. A number of the priorities identified in the present study might be addressed by more effective advocacy, e.g. greater policymaker awareness of WHO guidelines and strategies, and of the linkage between the movement behaviours and NCD prevention. There is increasing expectation in HICs that research funding will include knowledge exchange, and that the work of the researcher does not end when the paper is published [38]. Capacity for knowledge exchange in public health is limited in LMICs at present, but there are useful examples [39], and guidance on developing more effective knowledge exchange is available [40]. Successful movement behaviour surveillance research and knowledge exchange can occur within the same project [37] and greater effort in knowledge exchange can have impact in public health policy and practice, including in physical activity and health [41, 42]. One likely reason for public underfunding of child and adolescent movement behaviour surveillance and policy action, even in HICs with well-established surveillance systems such as Canada, is that these behaviours are being perceived as the responsibility not of government or society, but of the individual/family [43]. Future advocacy and knowledge exchange might therefore emphasise the importance of ‘upstream’ socio-ecological influences, above the level of the individual/family, and the key role for public policy in shaping the socio-ecological environment [44].

Conceptual model for improving national and international surveillance and achieving a society with more active children and adolescents

Increased and sustained funding is the foundation of improved national and international movement behaviour surveillance, but funding was also the only survey item with marked disagreement between participants, both in their ratings and comments. While 93% of respondents agreed that greater public funding was either ‘very important’ or ‘extremely important’, only 26% felt that it was either very or extremely important to seek funding from private/commercial sources. A number of strongly held concerns were expressed over accepting funding from commercial organisations with interests counter to public health such as tobacco, sugar, infant formula and soft drinks. Funding from commercial organisations may provide an opportunity to advance the current inadequate state of national and international movement behaviour surveillance, but brings a risk of bias [45, 46]. Private funding can lead some to dismiss the findings of research or surveillance funded by commercial organisations, regardless of the efforts made to minimise bias. ‘Big Food’ has undoubtedly tried to delay or even avoid progressive public health policies [47]. There are less obvious vested interests in movement behaviours (‘Big Tech’ or ‘Big Cars’) [48], which might encourage an excessive focus on food as the main source of our major public health problems.

The reality for many respondents to the present study is that private/commercial funding of surveillance or surveillance research is the only option, and this does not just apply to LMICs. There may be no choice between commercial and non-commercial funding, but a decision about whether to conduct surveillance or surveillance research or not. There is potential for commercial vested interests to take advantage of gaps in funding for movement behaviour surveillance and research. Accepting funding for movement behaviour surveillance research should not be in conflict with other public health objectives (e.g. diet or substance use related). It seems possible to carry out successful movement behaviour surveillance projects funded by private donors which do not lead to public health conflicts [49]. However, sustainable surveillance is likely to require government funding in most countries. Some respondents in the present study commented that striking a balance between academic and commercial interests that do not disadvantage public health is what is desirable and felt that striking such a balance was realistic. Intolerance of movement behaviour research and surveillance that is funded from commercial sources may also enforce a Westernised lens on a global issue. Countries have unequal availability of public and non-governmental organisation funding for research and failing to accept commercial funding if it is the only source of funding might lead to increasing inequality in global movement behaviour research and surveillance. Finally, some respondents noted that public/governmental funding brings its own problems such as pressure to avoid criticism of government policy or lack of government policy, or loss of editorial independence.

Regardless of the source of funding, unbiased population health research is highly beneficial to society. The level of disagreement over funding sources observed in the present study suggests that there may be a need to establish a globally accepted framework for future surveillance and surveillance research. This could include a written agreement that the funder will not influence the design and methods, will not be involved in data interpretation and reporting, will not own the data and will not use the funding of surveillance to promote harmful products. Some WHO guidance on the use of commercial funding is available [50] and useful models for managing funding from private/commercial sources for nutrition research are also emerging [51, 52]. Movement behaviour researchers have a responsibility to keep advocating for increased funding for surveillance, to share their research findings widely with relevant stakeholders, and to demonstrate the economic and other benefits which will derive from movement behaviour surveillance. Future research could examine which commercial sources are most acceptable to surveillance researchers (e.g., food versus banking versus insurance), while also establishing best practice guidelines when engaging in private/commercially sponsored research, similar to recent efforts from nutrition researchers [52].

Our study presents strengths that include novelty, timeliness and importance of the topic–particularly in the wake of the adverse impact of the COVID-19 pandemic on the movement behaviours of children and adolescents [14,15,16,17,18,19] and with relatively recent evidence-based 24-h guidelines for the movement behaviours from birth to the end of adolescence [5,6,7,8,9,10]. High-quality, evidence-based movement behaviour guidelines are not matched by high quality global surveillance at present [3]. A further strength is the wide global reach in terms of geographical representation and level of economic development, almost equal numbers of respondents from LMICs/HICs and equal numbers of women and men, which is very rare in this field of research [53]. Respondents drew on expertise across the age range from early childhood to the end of adolescence and response rates from the two international initiatives were fairly high. Lastly, our findings suggest that there are many important research and surveillance gaps, including the need for a framework with which to use private/commercial funding and the need for better surveillance measures of screen time and physically active play.

Weaknesses of the present study include the restriction of inclusion to SUNRISE and AHKGA Report Card participants. This decision was taken largely on practical grounds given that these two global initiatives provided an expert constituency that was easily identifiable and relatively easy to contact and would prevent the over-representation from any participating countries or regions. Since both initiatives are global, the inclusion of respondents from SUNRISE and AHKGA ensured wide geographical reach and avoided the under-representation of LMICs, typical of most biomedical research. SUNRISE and AHKGA focus on 3–17-year-olds, so different priorities could have been identified for toddlers and infants. However, many of the improvements suggested by respondents in the present study (e.g. improved human capacity, better understanding of the movement behaviours and how to measure them, increased funding of surveillance, more sophisticated methods) would likely apply to surveillance in infants and toddlers too [1, 3]. In addition, many of the survey respondents also have expertise in movement behaviour surveillance in infants and toddlers. Restricting inclusion to those involved in the SUNRISE and AHKGA initiatives may have limited generalisability though; for example, responses may have differed from researchers who work outside the initiatives, including those who work exclusively in physical activity (not the other movement behaviours), or in adults rather than children and adolescents.

While the sample size for the present study was large relative to the global pool of paediatric movement behaviour surveillance experts, and for Delphi studies based on international expertise, the sample size for statistical comparisons was small. This may have limited our ability to detect differences, but Delphi Studies aim to identify consensus and statistical inferences are usually of secondary importance [20,21,22, 54]. We used a modified Delphi Study with items offered to participants that were identified by the international leadership group of the study prior to the round 1 survey rather than being completely open at the start of the process. The modified Delphi design is now relatively common [25] and was intended to produce a more efficient process than a traditional Delphi design in the present study, but it may have constrained survey responses. However, participants were invited to suggest additional items to those provided to them at round 1, and only one new item was suggested and added (the linkage of movement behaviours to academic attainment, categorised under other issues). There was also possible acquiescence bias in responses as the study leadership group included those who lead the SUNRISE and AHKGA initiatives, though participants were informed that individual survey responses would be treated as anonymous and not be shared with the present study leadership group in an effort to minimise this source of bias.