I worry that the CMAJ’s editorial team erred when it promoted this article in its May 23, 2017 email to the readership. The second headline of that email read: “Research: Fructose sugars and diabetes — There is no evidence to support the hypothesis that fructose sugars particularly drive diabetes risk.”
This suggests that for all evidence linking fructose and diabetes, there is sufficient counter-evidence to conclude that there is no causation between the two.
However, the article in question1 was only a meta-analysis of prospective cohort studies. It did not look at observational studies, animal or in vitro studies, or interventional trials. Indeed, these were excluded from analysis as part of the study’s design (see Figure 1 in the original article). To say that there is no evidence linking fructose and type 2 diabetes is to ignore the potential merit of all the excluded studies (some of which are probably found in this review article).2
I will also point out that the authors of this meta-analysis stated directly that they have only “weak” confidence in their conclusion of null causation.
So in future, I would ask that the CMAJ editorial team take more care not to oversell an article’s conclusions just for the sake of a snappy headline.
Footnotes
Competing interests: Christina Quinlan states, “As a family doctor, I counsel my patients that too much sugar or fructose is bad for their health.”