Authors retract two studies on high blood pressure and supplements after realizing they’d made a common error

A group of researchers from Iran, Italy and the UK have retracted two meta-analyses on supplements and high blood pressure after making what a statistics expert calls a common error.

Both papers were originally published in the Journal of Human Hypertension. Here’s the retraction notice for “Elevated blood pressure reduction after α-lipoic acid supplementation: a meta-analysis of randomized controlled trials:”

The Authors have retracted this article because there are fundamental errors in the data presented that undermine the conclusions drawn. Standard errors were used instead of standard deviations when using data from one of the studies (Koh et al) included in the meta-analyses. Numerous inconsistencies were noted from the numbers used in the meta-analyses when compared to the original studies.

All authors agree to this retraction.

Here’s the notice for “Effect of L-citrulline supplementation on blood pressure: a systematic review and meta-analysis of randomized controlled trials:”

The authors have retracted this article [1] because there are fundamental errors in the data presented that undermine the conclusions drawn.

Standard error was used instead of the standard deviation from two trials included in the meta-analysis. This has seriously affected the overall effect. There have also been other inconsistencies noted in the values used in the calculations from the original studies.

All authors agree to this retraction.

The paper has been cited twice, according to Clarivate Analytics’ Web of Knowledge.

For more on the difference between standard error and standard deviation, see this post from the Stats Geek.

Sadegh Jafarnejad of Kashan University of Medical Sciences, the corresponding author of both papers, told Retraction Watch:

As the retraction notices emphasize, both papers were retracted due to authors request. After some human errors (honest errors) were found in presented data, we tried to correct the papers by letter to editor/erratum. However after consulting with co-authors and Editor, we decided to retract the papers and submit them in the journal after rectifying and revising the data (one is submitted in the same journal and another one will be submitted very soon). 

Common errors

Indiana University’s David Allison, who along with his graduate students, postdocs, and faculty colleagues has been trying to alert journals to statistical problems in the literature, told Retraction Watch:

Errors in meta-analyses are common. In my experience, perhaps the most common error entails miscalculation of the effect size and what seems to be the most common source of this miscalculation is misapprehension of some measure of variance. Mistaking a standard error for a standard deviation is common and seems to often result from a combination of insufficient carefulness on the part of the meta-analysts and unclear or incorrect labeling of standard errors vs standard deviations in the original articles being meta-analyzed. Unfortunately, such errors can invalidate a meta-analysis.

In 2016, John Ioannidis of Stanford warned that “there is massive production of unnecessary, misleading, and conflicted systematic reviews and meta-analyses.”

Quoting a 1990 paper, Allison said:

As Ingram Olkin stated years ago, “Doing a meta-analysis is easy…Doing one well is hard.”

Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].