Food sources of fructose-containing sugars and glycaemic control: systematic review and meta-analysis of controlled intervention studies

In this article by Choo and colleagues (BMJ 2018;363:k4644, doi:10.1136/bmj.k4644, published 21 November 2018), the 2008 study by Puglisi and colleagues was incorrectly categorised as fruit when it should have been dried fruit for the analysis of the effect of fructose-containing sugars on fasting blood glucose in addition studies. The correct categorisation of the Puglisi study results in the following changes:

• In the last row of table 1 under “Addition studies”, the entry should begin: “Fruit=9; dried fruit=1; . . .”

• Under the “Addition studies” heading in figure 3, the row for “Fruit” should be replaced by two rows for “Fruit” and “Dried fruit” with revised data.

• Under the heading “Fasting blood glucose” in the results, the third and fourth sentences should begin: “We saw a significant interaction by food source in addition studies (P=0.01): . . .” and “However, fruit (six study comparisons), dried fruit (one study comparison) . . .”, respectively.

• The supplementary table 2 and figure 11 have also been revised accordingly.

Additional corrections to the paper include:

• In figure 2, the overall P value for “Added sweeteners” under “Addition studies” should read 0.57.

• In figure 3, 95% confidence intervals for all substitution and addition studies have been revised. Additionally, the P value for heterogeneity for “Fruit drinks” under “Addition studies” should read <0.01.

• In figure 4, the overall P value for “Mixed sources” under “Substitution studies” should read <0.01.

• Under the heading “Fasting blood insulin” in the results, the third sentence should begin: “We saw no effect in substitution studies (70 studies, substantial heterogeneity (I²=61%, P<0.001)) . . .”

• In supplementary figures 18-21, the units for insulin should read pmol/L.

The article and supplementary files will be updated in due course.