PREreview of Effects of physical activity and sociodemographic aspects on the mental health of the Brazilian population at different times of the COVID-19 pandemic

Summary:

This observational study primarily examined the mental health trends of Brazilians before and after lockdown from COVID-19. Data concerning depression, anxiety, stress, and physical activity(PA) was collected in April-May 2020, called T1, and in October-November 2021, called T2. Data was collected through an online questionnaire sent in T1 to participants from Brazil, Spain, and Chile. A subsequent form was sent in T2 only to Brazilian participants. Analysis attempted to explain data trends with other studies on the social and demographic context of Brazil at the time.

Strengths:

• Introduction and abstract clearly demonstrate objectives and relevant context for understanding the study. Overall, the paper is well-organized. 

• T1 and T2 sample size lend statistical power and generalizability, and the contextual definitions of both periods adds clarity to analysis. 

Suggestions/Improvements:

• Abstract

  • Stylistically, the phrase PA should be introduced as “physical activity (PA)” for greater clarity. 

• Introduction

  • Occasional inclusion of information pertaining to periods such as January 2025 or the 2017 WHO study can be removed as they are not directly relevant to analysis in the COVID-19 time period of the study, However, the inclusion of this information is personal preference. 

• Methodological procedures

  • A stronger connection is needed between the mention of Google Forms and CHERRIES, as the latter hasn't been sufficiently elaborated upon to show its relevance or application in this context. 

  • The application of STROBE and CHERRIES are not adequately explained; they are merely mentioned. 

  • Both the snowball sampling in T1 and convenience sample in T2 result in a strong sampling bias and low generalizability, despite the sample numbers. A strong justification of benefits in comparison to limitations of these methods would aid this section. This is present in the conclusion, but earlier mention would be more ideal.  

• Results

  • Although T1 included 3386 participants, T2 only included 1006 participants. This is a significant decrease in participation, and represents a significant source of error and bias in the study.

  • The T2 sample was skewed toward women (66.8%) in comparison to men (33.2%). While this might reflect actual survey response trends, it limits the generalizability of findings to Brazilian men or gender-diverse populations. Furthermore, there might have been a response bias resulting in more women responding than men, skewing the data.

• Discussion

  • Studies or reviews are mentioned with a citation number. It is preferable to directly name the study or authors or use an (Author Year) in text citation. Additionally, conglomerating citations in this manner increases the difficulty of verifying whether the references are truly supporting the claim. 

  • The study mentions limitations of the convenience sampling method, but does not explain how this was factored into the analysis of research data. 

  • There is mention of “response bias” but more specific examples may enhance detail.

Competing interests

The authors declare that they have no competing interests.