A systematic review of existing datasets including fitness tests in children and adolescents was previously performed by Tomkinson et al and details of the search have been published.40 (link) These data were included in the FitBack dataset, with Monte Carlo simulation used to produce pseudo data (from reported means and SDs) when raw data were unavailable. In addition to this, the authors of the FitBack network conducted a centralised narrative search based on fitness terms to identify new datasets not included in the Tomkinson et al review.40 (link) For inclusion, valid data on sex, age and at least one of the ALPHA fitness tests (high-priority version) was required. In the previous study by Tomkinson et al, the age range was 9–17 year old, whereas in this study, we widened the age demographic to include subjects aged 6–18 years old. It is important to note that our search strategy was focused on fitness, and specific searches on adiposity, BMI or waist circumference were not conducted for pragmatic reasons (eg, the very large number of studies including these key words). Therefore, it is possible that we missed relevant anthropometry-specific datasets. This, together with the fact that other organisations are comprehensively monitoring paediatric obesity, is the reason why we primarily focused on CRF and muscular strength, and reported results for anthropometric measures (body height, body mass, BMI and waist circumference) as online supplemental material.
The FitBack network involved numerous experienced researchers working in paediatric fitness across Europe, which helped to identify unpublished fitness datasets that were pooled with gathered data. Moreover, large datasets from existing surveillance systems in Europe such as SLOfit,41 (link) NETFIT42 (link) and Fitescola43 (link) were also included. Further, we excluded older datasets if a more recent and more representative dataset was available for certain countries. The ambition was to use the most recent available data for each country, which in some cases was a single large dataset, while in others was the accumulation of several studies or datasets covering different geographical regions within a country. Sources used for generating the reference values are available on the FitBack website (www.fitbackeurope.eu/en-us/fitness-map/sources) as well as in online supplemental table 1.