We appreciate the opportunity to answer the letter of Peña et al. [1] regarding our paper “Effectiveness of a minimally processed food-based nutritional counselling intervention on weight gain in overweight pregnant women: a randomized controlled trial” published on September 2022 in the European Journal of Nutrition [2].

Our study aimed to evaluate the effectiveness of a nutritional counselling intervention based on encouraging the consumption of unprocessed and minimally processed foods, rather than ultra-processed products, and the practice of physical activities to prevent excessive gestational weight gain in pregnant women with overweight. The study was a two-armed, parallel, randomized controlled trial conducted in primary health units of a Brazilian municipality, from 2018 to 2021. Adult pregnant women with overweight (n = 350) were randomly assigned to the control (CG) or intervention (IG) groups. The intervention consisted of three individualized nutritional counselling sessions based on encouraging the consumption of unprocessed and minimally processed foods rather than ultra-processed products, following the NOVA food classification system [3, 4], and the practice of physical activities [2].

In their letter, Peña et al. [1] questioned the legitimacy of our data based on three main issues. Regarding their first issue, they mentioned “undisclosed outcome switching”, in which the authors argue that the primary outcome described in the published study protocol [5] and the clinical trial registration (adequacy of weight gain) differs from that described in our present paper (excessive weight gain).

While investigating the effect of an intervention on gestational weight gain (GWG), defining the study's primary outcome is a significant challenge. All weight gain categories (insufficient weight gain, appropriate weight gain, and excessive weight gain) should be included as study outcomes as they are equally relevant. After the clinical trial registration of our study, we observed that recent studies that investigated the effect of interventions on weight gain in women with overweight or obesity used excessive weight gain as the primary outcome [6, 7]. We found that this alteration in our study would not affect the previously described sample size calculation. Therefore, we chose to change the primary outcome considering the recent evidence. Our mistake was not having clarified the reasons for this change in the article. However, in our article [2], the effect of the intervention on all categories of weight gain was explored, with these all being described in the results section of the manuscript.

Since all the outcomes on maternal weight gain were reported in our manuscript, we disagree with Peña and colleagues’ [1] comment that “When taken together, the statements in the manuscript, the clinical trials registration, and the published protocol are consistent with research misconduct involving changing or omitting data or results such that the research is not accurately represented in the research record”.

The second point raised by Peña et al. [1] is related to “inappropriate rounding”. According to the authors, “in the body of the manuscript, the p-value for the modified intent to treat analysis is listed as 0.049. In contrast, in the abstract, the p-value was listed as 0.04”.

Regarding this concern, as reported in the abstract of our paper, “In modified intention-to-treat analysis, there was a lower chance of the IG women having excessive GWG [OR 0.56 (95% CI 0.32, 0.98), p = 0.04], when compared to the CG. No between-group differences were observed for the other maternal outcomes investigated”. The p-value presented in the abstract is for the adjusted logistic regression model investigating the effect of the intervention on the outcome of excessive weight gain. This value is the same as in Table 2 of our manuscript for this outcome. The p-value of 0.049, mentioned by Peña et al. [1], is for the Chi-Square test for differences between groups, which was not described in the abstract due to the word limit.

In this way, the comment of Peña et al. [1] “This is inconsistent with standards about numerical rounding, and a second example of misleading reporting that exaggerates results” is entirely unappropriated.

The third issue raised by Peña et al. [1] is related to “switching from non-significant preplanned analysis to statistically significant alternative”. In this point, the authors mentioned that our published study protocol [5] predicted only the intention-to-treat analysis, and there was no mention of the modified intention-to-treat analysis.

Our published study protocol [5] did not account for the modified intention-to-treat analysis. However, face-to-face interviews were suspended between March and October 2020 due to the COVID-19 pandemic, which led to a substantially higher loss to follow-up than was predicted, as described in our manuscript. Information on maternal weight was obtained through medical records at each prenatal visit, and the data on the outcome was available. However, some IG women did not receive any nutritional counselling. The trial was designed to investigate the effect of the intervention on the outcomes, and these women were not exposed to any intervention.

As described in the CONSORT 2010 [8], “the term modified intention-to-treat is quite widely used to describe an analysis that excludes participants who did not adequately adhere to the protocol, in particular, those who did not receive a defined minimum amount of the intervention”. Only women who did not attend any nutritional counselling sessions were excluded from the modified intention-to-treat analysis in our study. Both results for the intention-to-treat and modified-intention-to-treat analysis are described in the results section of our paper. However, considering the word count for the abstract, it was not possible to describe both results there.

Taking all these points together, we totally disagree with the accusations of Peña and colleagues [1] that our paper “fails to meet best reporting practices of parallel group randomized controlled trials”.

As previously declared, the authors of our paper have no conflict of interest. The study was conducted according to the guidelines of the Declaration of Helsinki and followed the CONSORT statement. A correction to the paper will address the missing information.