In this paper by Chiavaroli and colleagues (BMJ 2021;374:n1651, doi:10.1136/bmj.n1651, published 5 August 2021), a significant positive dose response gradient for absolute dietary glycaemic index and systolic blood pressure was not included in the conclusion of the abstract and the summary box, and should be listed among the secondary outcomes that showed reductions.
Further, in the first paragraph of the discussion, the direction of the linear dose response should be described as “positive” rather than “inverse.” Lastly, in figure 2, the subheading “Publication bias” underneath “Upgrades” should read “Dose response.”
The article will be updated in due course.