Conduct and reporting of formula milk trials: systematic review

Importance: Formula milk is consumed by most infants in Europe and North America and global consumption is rising. New formula products need to be tested in clinical trials, but concerns have been raised that such trials are biased and may undermine breastfeeding.Objective: To systematically review the conduct and reporting of formula trials. Trial primary outcomes were evaluated for risk of bias, and reported trial procedures were evaluated for risk of undermining breastfeeding in participants.Data sources: MEDLINE, EMBASE and CENTRAL databases were searched from 1st January 2006 to 31st December 2020. Study selection: Intervention trials comparing at least two formula products in children under 3 years were included, but not trials of human breastmilk or breastmilk fortifiers. Data extraction and synthesis: Data were extracted in duplicate and primary outcome data were synthesised for meta-analysis using a random-effects model weighted by the inverse variance method. Main outcomes and measures: Risk of bias was evaluated using the Cochrane Risk of Bias 2.0 tool, risk of undermining breastfeeding was evaluated using published consensus guidance. Primary outcomes of included trials were identified from clinical trial registries, protocols or trial publications. Results: We screened 22,201 titles and identified 307 trials published between 2006 and 2020, of which seventy-three (24%) trials in 13,197 children were prospectively registered. A further 111 unpublished but registered trials in 17,411 children were identified. Detailed analysis was undertaken for the 125 trials (23,757 children) published since 2015. Seventeen (14%) of these recent published trials were conducted independently from formula companies, 26 (21%) were prospectively registered with a clear aim and primary outcome, and authors or sponsors shared prospective protocols for 11 (9%). Risk of bias was low in five (4%) and high in 100 (80%) recent published trials, mainly due to inappropriate exclusions from analysis and selective reporting. For 68 recently published superiority trials we calculate a pooled standardised mean difference of 0.51 (range -0.43 to 3.29) with an asymmetrical Funnel plot (Egger’s test P=0.0001), which reduced to 0.19 following correction for asymmetry. Eighty-six (69%) study primary outcomes were reported by authors as favourable and 115 (92%) abstract conclusions were favourable. One of 38 (3%) trials in partially breastfed infants reported adequate breastfeeding support and 14 of 87 (16%) trials in non-breastfed infants confirmed the decision not to breastfeed was firmly established before enrolment. Conclusions and relevance: Formula trials lack independence or transparency and published outcomes are biased by selective reporting.Registration: PROSPERO 2018 CRD4201809192