Ultra-Processing of Study Results in Nutrition

Expert opinion holds that ultra-processed food is not a good thing. So it’s quite natural to expect that helping people resist the convenience and taste of this ubiquitous type of food will help with dietary health outcomes. So natural that it’s quite easy to dismiss inconvenient, unexpected findings. A little ultra-processing of study results in nutrition can almost effortlessly bring such findings back into line with expectations.

Without any intent to mislead, it’s quite easy to rationalize modified analyses that will better support prevailing wisdom about ultra-processed foods.

Preventing Excessive Weight Gain in Pregnant Women

A case in point can be found in the European Journal of Nutrition. We noted this study, an RCT, back in 2022, when Daniela Sartorelli et al reported “unprecedented” effectiveness for counseling pregnant women with a BMI between 25 and 30 regarding ultra-processed foods to prevent excessive weight gain. The claim of unprecedented effectiveness did not seem to line up with the details of statistical analysis in the paper.

So Armando Peña, Mike Gibney, Kevin Hall, Ted Kyle, Andrew Brown, and David Allison wrote to the journal, raising questions about undisclosed changes in analyses and outcomes, inappropriate rounding, and selective reporting in this paper. Collectively, they said, these problems set up a stark contrast between an objective analysis of results and the authors’ description of an “unprecedented” demonstration of effectiveness.

Disagreement Without Being Disagreeable

It seems fair to say that the authors did not agree with all of these concerns. In fact, they concluded their response saying “we totally disagree.”

Notwithstanding that statement of disagreement, those authors issued a correction to explicitly disclose the discrepancies between pre-specified outcomes and analysis versus what they originally published. Peña notes that correction, calls it laudable, and we entirely agree.

We might disagree on the potency of educating people about ultra-processed of food for changing health outcomes. But we certainly agree on the need for transparency and rigor in reporting nutrition research.

Click here for the original paper, here for the letter from Peña et al, here for the response, and here for the correction.

Tortilla Chip, photograph by Neil Tackaberry, licensed under CC BY-ND 2.0

Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.

May 29, 2023