Trends in Obesity Among Adults in Mississippi, 2017–2023

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Reviewer´s report:

Thank you for inviting me to review manuscript entitled: “Trends in Obesity among Adults in Mississippi, 2017-2023“.

The objective of this study is to analyse trends in obesity prevalence among adults in Mississippi from 2017 to 2023, considering five key health determinants: gender, education, age, race, and socioeconomic status. The manuscript provides a comprehensive trend analysis using Joinpoint Regression; however, several aspects could be refined to enhance transparency, clarity, and the overall quality of the analysis.

Major Concerns and Recommendations:

1. Consistency in Reporting p-Values

To ensure clarity and uniformity in statistical reporting, I recommend consistently presenting all p-values to three decimal places throughout the manuscript. Adopting this standardized approach improves the readability and comparability of the results.

2. Consideration of Summarizing APC Estimates in Tabular Format

To summarize the Annual Percentage Change (APC) estimates, including 95% confidence intervals (CI) and p-values, in a tabular format. Presenting these results in a table would enhance clarity, improve readability, and allow for easier comparison of trends across different health determinants.

3. Lack of Discussion on Potential Methodological Limitations

To improve the transparency and interpretability of the findings, I recommend including a dedicated discussion on the potential methodological limitations of the presented analysis. A structured summary of possible biases/confounding factors would strengthen the credibility of the study.

Conclusion

This study offers a valuable contribution to understanding obesity trends and their associated health determinants. Implementing these revisions would improve the study’s transparency and increase the robustness of the presented conclusions.

Author Response

Comments 1: To ensure clarity and uniformity in statistical reporting, I recommend consistently presenting all p-values to three decimal places throughout the manuscript. Adopting this standardized approach improves the readability and comparability of the results. 

Response: Thanks for feedback! The p-values have been distributed in the results section, and stated in the method section. Please see red highlighted sections of the paper.

Comment 2: To summarize the Annual Percentage Change (APC) estimates, including 95% confidence intervals (CI) and p-values, in a tabular format. Presenting these results in a table would enhance clarity, improve readability, and allow for easier comparison of trends across different health determinants. 

Response: Thanks for the feedback! Table 2 was added.

Comment 3: To improve the transparency and interpretability of the findings, I recommend including a dedicated discussion on the potential methodological limitations of the presented analysis. A structured summary of possible biases/confounding factors would strengthen the credibility of the study.

Response: Thanks for the feedback! The limitations and biases were added to the discussion section. Here is the excerpt from the manuscript: While the study is very methodogical sound, there was one limitation. By using Joinpoint regression trend analysis, the analysis may oversimplify trends. This may result in trends being perceived as less accurate. There is also bias associated with the study design. The bias is a result of the data collection process. The data was collected using convience sampling through telephone interviews, which results  in selection bias that can impact the data/analysis for the study.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

In this manuscript, the authors presented a study to assess changes in obesity in Mississippi adults in a 7-year interval. The topic is interesting, but the article has some clear weaknesses. In particular, there is a lack of a comprehensive view that considers the different determinants simultaneously: the authors merely examined one determinant at a time without considering their interactions.

Major concerns:

Materials and Methods:

This part needs to be supplemented. Key elements such as sample size (total, males, females) are missing. Also missing in Table 1 are the units of measurement and clarifications regarding the lack of data. I suggest reporting here what is indicated in the results (lines 89-93).

Results:

Figures: I suggest revising them by reporting whole years (without decimals) on the x-axis.

Discussion:

The authors do not discuss an important aspect: the impact that the COVID-19 outbreak may have had on the observed trends. I suggest integrating the discussion by examining this important aspect.

Minor concerns:

-Lines 31-33: As the topic is expanded later, I suggest ending the sentence with something like “as indicated in detail later”.

-Lines 71-73: the text seems to be repeated in subsequent lines 73-75. I suggest revising this entire section.

Author Response

Comment 1: In this manuscript, the authors presented a study to assess changes in obesity in Mississippi adults in a 7-year interval. The topic is interesting, but the article has some clear weaknesses. In particular, there is a lack of a comprehensive view that considers the different determinants simultaneously: the authors merely examined one determinant at a time without considering their interactions.

Response: We appreciate the reviewer's insightful comment regarding the examination of determinants simultaneously and considering their potential interactions. We agree that a comprehensive analysis exploring the interplay between different determinants of obesity would provide a richer and more nuanced understanding of the observed trends. Indeed, we recognize that obesity is a multifactorial issue, and these determinants likely do not operate in isolation.
However, due to inherent limitations within the BRFSS dataset, conducting a robust analysis of determinant interactions across the entire 2017-2023 period was not feasible for this study. Specifically, several factors prevented us from adopting this more complex approach:
1.    Inconsistent Data Reporting Periods and Cohorts: As the reviewer may have noted in the 'Race' and 'Income' sections of the results, the BRFSS data has inconsistencies in reporting across different determinants over the years. For instance, detailed income categories above $50,000 were only consistently reported from 2021 onwards. Similarly, data for certain racial/ethnic groups were not consistently available throughout the entire 2017-2023 period. This variability in reporting periods across determinants makes it statistically challenging and potentially misleading to conduct interaction analyses across the full 7-year span. Any attempt to analyze interactions across the entire period would necessitate using only the data available consistently for all determinants, which would significantly reduce the sample size and potentially introduce bias by excluding valuable data points specific to certain determinants in certain years.
2.    Data Structure and Sample Size Considerations: The BRFSS is a large, population-based survey, but it is designed to provide robust prevalence estimates for individual determinants within specific demographic categories. While we have substantial sample sizes for analyzing trends within each determinant separately, attempting to analyze interactions between multiple determinants simultaneously, across all years and subgroups, would drastically reduce the effective sample size. This is because interaction analysis requires complete data for all interacting variables for each individual included in the model. Given the survey nature of BRFSS and the varying response rates across different demographic questions, we cannot assume complete and consistent data linkage across all determinants for every respondent in every year. A simultaneous interaction analysis would likely be underpowered and might yield unreliable or unstable results due to reduced sample sizes in specific interaction strata.
3.    Focus on Describing Independent Trends as a First Step: Given these data limitations, we opted for a robust and statistically sound approach of analyzing trends for each determinant independently. This allowed us to maximize the use of the available data for each determinant and provide clear and interpretable trend estimates within each demographic category. We believe that understanding these independent trends for gender, age, education, income, and race is a crucial first step in characterizing the complex landscape of obesity in the US and D.C. These findings, while not exploring interactions directly, still provide valuable insights into how obesity prevalence is evolving differentially across key population subgroups and can inform targeted public health interventions.
4.    Limitations and Future Directions: We acknowledge that examining interactions is an important direction for future research. We have added a point to the discussion section explicitly mentioning this limitation and suggesting that future studies, ideally with datasets designed to facilitate interaction analysis or using advanced statistical methods suitable for complex survey data, could explore these important interrelationships. For this current study, given the data constraints, we believe that focusing on robustly describing the independent trends within each determinant provides the most scientifically sound and practically informative analysis.

Comments 2: This part needs to be supplemented. Key elements such as sample size (total, males, females) are missing. Also missing in Table 1 are the units of measurement and clarifications regarding the lack of data. I suggest reporting here what is indicated in the results (lines 89-93).

Response: Thanks for your feedback! However, key elements/units of measurements have been Inaccessible since day 1 of the new administration (Website message: CDC's website is being modified to comply with President Trump's Executive Orders.) but even if the website were functioning, I believe the data is not available for the males and females, just the entire sample size is available and as mentioned in the data availability notes: there are a lot of missing entries from the entire sample size]

Comment 3: Figures: I suggest revising them by reporting whole years (without decimals) on the x-axis.

Response: I added the whole years on the x-axis for each figure.

Comment 4: The authors do not discuss an important aspect: the impact that the COVID-19 outbreak may have had on the observed trends. I suggest integrating the discussion by examining this important aspect.

Response: Thanks for your feedback! I added the impact of the COVID-19 outbreak to the discussion section. Here are the excerpts from the manuscript: This study showed that while obesity continues to rise among certain demographic groups—particularly males, middle-aged adults, and higher-income individuals—there are also promising signs of decline among females, high school-educated individuals, and Hispanic populations. Current policies have inadequately addressed overweight and obesity [11]. Throughout the COVID-19 pandemic, factors such as physical inactivity, sedentary behaviors, and poor dietary choices emerged as leading contributors to obesity. Moreover, unhealthy eating habits, heightened behavioral stress, depression, anxiety, low mood, as well as age, gender, and belonging to ethnic minority groups, were also recognized as significant risk factors for obesity during this period [11]. Before the COVID-19 pandemic, many of the risk factors for obesity were already well-established, such as physical inactivity, sedentary lifestyles, unhealthy eating habits, and behavioral stress. These factors were consistent contributors to obesity across various populations [11]. Without significant reforms, projected trends will have severe consequences at both individual and population levels, with the associated disease burden and economic costs continuing to rise [11]. [Excerpt 1]

An existing study found that highly educated individuals were able to adapt by engaging in healthier behaviors, such as maintaining regular physical activity at home, following balanced diets, and utilizing online fitness programs or virtual health consultations. Their awareness of the risks associated with obesity and COVID-19 likely motivated them to prioritize their health [13]. [Excerpt 2]

Metabolic and bariatric surgeries are most commonly performed on adults aged 18 to 65 years, with the majority of patients typically falling between their 30s and 50s [27]. This age group often seeks these procedures due to long-term struggles with obesity and related health conditions, such as type 2 diabetes or hypertension [27]. However, in recent years, there has been a growing recognition of the benefits of these surgeries for ado-lescents with severe obesity, particularly when other treatments have proven ineffective [27]. [Excerpt 3]

While age, gender, and belonging to ethnic minority groups were recognized as in-fluencing obesity risk before the pandemic, the pandemic highlighted and exacerbated these disparities. For example, ethnic minority groups faced heightened challenges due to systemic inequities, which were further magnified during the pandemic [30]. [Excerpt 4]

Comment 5: Lines 31-33: As the topic is expanded later, I suggest ending the sentence with something like “as indicated in detail later”.

Response: Thanks for your feedback! The comment was addressed. Here is the excerpt from the manuscript: Characterized by excessive fat accumulation, obesity is often associated with a range of chronic conditions, including heart disease, type 2 diabetes, and certain cancers as indicated in detail later. 

Comment 6: Lines 71-73: the text seems to be repeated in subsequent lines 73-75. I suggest revising this entire section. 

Response: Thanks for your feedback! The comment was addressed. Here is the excerpt from the manuscript: In this study, we investigated trends in obesity prevalence among adults in Mississippi from 2017 to 2023 across five determinants: race, gender, age group, household income, and highest education attained. Table 1 presents the data for these indicators. We obtained annual crude prevalence of obesity from the CDC’s Behavioral Risk Factor Surveillance System (BRFSS) Prevalence & Trends Data [9]. We then calculated standard errors using Microsoft Excel and performed trend analysis using Joinpoint Regression (version 5.3.0.0, National Cancer Institute) [10].

 

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I appreciated the changes made by the authors. The manuscript is now publishable.