Commentary on “A meta-analysis but not a systematic review: an evaluation of the Global BMI Mortality Collaboration”

Abstract

•We conducted a participant level meta-analysis on the BMI mortality association. •All-cause mortality was minimal at 20·0–25·0 kg/m2. •Both overweight and obesity were associated with higher risk of mortality. •Our results were broadly consistent across the four continents studied. A meta-analysis but not a systematic review: an evaluation of the Global BMI Mortality CollaborationJournal of Clinical EpidemiologyVol. 88PreviewSystematic reviews and meta-analyses of individual participant data (MIPDs) are recognized as a gold standard approach for evidence synthesis [1]. However, not all meta-analyses are based on systematic reviews. Determining whether an MIPD is based on a rigorous systematic review is important for its appraisal [2]. Full-Text PDF A meta-analysis of individual participant data constructed to align with prior expert views: comments on Bhupathiraju et al.Journal of Clinical EpidemiologyVol. 88PreviewThe response by Bhupathiraju et al. ignores the major point of our article [1], which was to evaluate the extent to which the Global Body Mass Index Mortality Collaboration (GBMC) publication [2] was based on data from a systematic search. As our article documented, 238 of the 239 data sets used by the GBMC were previously known to the senior author, and several GBMC members are on the steering committee of the one remaining study, the UK Biobank. As described by Bhupathiraju et al., the GBMC was created in 2013, motivated by the publication of Flegal et al. Full-Text PDF