Investigating the rigour of research findings in experimental studies assessing the effects of breaking up prolonged sitting – extended scoping review

• Published in: Revista brasileira de fisioterapia (São Carlos (São Paulo, Brazil)) Brazil)), 2021-01, Vol.25 (1), p.4-16
• Main Authors: English, Coralie, Weerasekara, Ishanka, Carlos, Anjelica, Chastin, Sebastien, Crowfoot, Gary, Fitzsimons, Claire, Forster, Anne, Holliday, Elizabeth, Janssen, Heidi, Mackie, Paul, Mead, Gillian, Dunstan, David
• Format: Article
• Language: English
• Subjects: Databases, Factual; Humans; Physical activity; Prospective Studies; Publication bias; Registries; Sedentary behaviour; Sitting Position; Sitting-time; Systematic Review
• ISSN: 1413-3555; 1809-9246
• DOI: 10.1016/j.bjpt.2020.04.007

•Only 50% of the published trials were registered.•Unregistered trials may have collected other measures that remain unreported.•Many additional outcomes not appearing in trial registries were published.•Some reported positive findings are likely spurious due to unadjusted multiple comparisons. Sedentary behaviour research is a relatively new field, much of which has emerged since the widespread acceptance of clinical trial registration. The aim of this study was to investigate the trial registration and related issues in studies investigating the effect of frequent activity interruptions to prolonged sitting-time. Secondary analysis of a scoping review including systematic searches of databases and trial registries. We included experimental studies investigating the effects of frequent activity interruptions to prolonged sitting-time. We identified 32 trials published in 45 papers. Only 16 (50%) trials were registered, with all 16 trials being completed and published. Of the unregistered trials, we identified three (19%) for which similarities in the sample size and participant demographics across papers was suggestive of duplicate publication. Identification of potential duplicate publications was difficult for the remaining 13 (81%). Results from 53 (76%) of the 70 registered outcomes were published, but 11 (69%) registered trials reported results from additional outcomes not prospectively registered. A total of 46 different outcomes (out of 53 reported outcome measures, similar measures were collated) were reported across all trials, 31 (67%) of which were collected in ≤2 trials. We found direct evidence of trial registration issues in experimental trials of breaking up sitting-time. The lack of prospective registration of all trials, and the large number of outcomes measured per trial are key considerations for future research in this field. These issues are unlikely to be confined to the field of sedentary behaviour research.