Dietary Inflammatory Index of Northern Mexican Indigenous Adults and Its Association with Obesity: Cross-Sectional Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Introduction

 

Line 58 – Please explain briefly  what is  higher modernity index

Line 58 – Are there studies relating “prudent”/”non prudent”  dietary habits with abdominal obesity? They should be included.

Line 77 – Any relation with neurodegenerative diseases? If yes, it should be included

Line 86 – Since BMI e WC was the measured antropometric  parameters (two parameters), I believe thet the word “multiple” leads to think that there are more parameters;

 

Methods

Line 99  - Where is written “published elsewhere”, a reference should be provided;

Line 120 – Where is written “Abdominal obesity was classified using the International Diabetes Federation cutoff points for South Asians: WC ≥80 cm in women and ≥90 cm in men”, since the studied population is from Mexico and not form Asia, if there is another reference available which could be more adequate, it should be included.

Line 130 -  “Individuals whose dietary data demonstrated inconsistencies were excluded from the analysis” should be in 2.2. Participants;

Line 132 – “Nutrient intake was adjusted for energy using the residual regression method” should have a brief explanation;

Line 176 – “tertile 1 (most anti-inflammatory diet) and tertile 3 (most pro-inflammatory diet)” is confusing; is there a tertile 2? How was denominated?   

 

Disussion

Line 296 -   Where is written “showing that there is a risk of increased obesity markers with each additional point in the DII score, regardless of the individual's position on the DII scale” please explain better “regardless of the individual's position on the DII scale”

 

Line 320 - “traditional” and “prudent” dietary patterns” would become better as “traditional” or “prudent” dietary patterns ;

The questions of the questionaire must be included as suplementary material. Moreover, any relation between the questions, diatary patterns and its trnaslation to nutrientes is explained or easily available for the reader.

Author Response

Dear Reviewer 1,

 

We sincerely thank you and the other reviewers for the time and effort dedicated to evaluating our manuscript titled "Dietary Inflammatory Index of Northern Mexican Indigenous Adults and Its Association with Obesity: Cross-Sectional Study."

We appreciate the constructive feedback and valuable suggestions, which have helped us improve the quality and clarity of our work. Below, we provide a point-by-point response to your comments. We hope the revisions satisfactorily address the concerns raised.

To facilitate tracking of changes, the manuscript's new additions, based on the reviewers’ suggestions, are highlighted in yellow. In the response letter, line numbers refer to the latest version of the manuscript.

 

Reviewer 1

 

Introduction

Line 58 – Please explain briefly  what is  higher modernity index

Response:

A higher score on the modernity index reflects a greater number of household goods and services, such as a fan, a blender, a gas stove, internet, a DVD, an oven, an iron, a refrigerator, air conditioning, a TV, a telephone, a radio, a washing machine, a cell phone, electricity, a car, and a wood stove. We have made a modification in this line for better understanding (line 57):

“In contrast, a higher modernity index or having a greater number of household goods and services was associated with a higher BMI.”

 

Line 77 – Any relation with neurodegenerative diseases? If yes, it should be included

Response:

We haven’t included it, and we found it very important to add this information. The corresponding line was modified, and you can find it at line 79:

“High DII scores have been consistently associated with NCDs [17], including diabetes [18], metabolic syndrome [19,20], cardiovascular disease [18,21], neurodegenerative diseases [22,23], and cancer [24,25].”

 

 

Line 86 – Since BMI e WC was the measured antropometric  parameters (two parameters), I believe thet the word “multiple” leads to think that there are more parameters;

Response:

Thank you for this important comment; we definitely agree. Therefore, we decided to delete the word “multiple” to avoid that misunderstanding. The corresponding paragraph (lines 86-88) now reads:

“Given this context, the objective of the present study is to assess the relationship between DII scores and measures of obesity, as well as their prevalence among indigenous adults from Sonora (IAS).”

 

 

Methods

Line 99  - Where is written “published elsewhere”, a reference should be provided;

Response:

We appreciate this important observation. We added the references to the new version of the manuscript.

Line 101:

“The sampling procedures for both studies have been published elsewhere [5,7];”

 

Line 120 – Where is written “Abdominal obesity was classified using the International Diabetes Federation cutoff points for South Asians: WC ≥80 cm in women and ≥90 cm in men”, since the studied population is from Mexico and not form Asia, if there is another reference available which could be more adequate, it should be included.

Response:

According to the IDF criteria, based on the Alberti and colleagues (2006), shows that the cut-off points for “Ethnic South and Central Americans” must “use South Asian recommendations until more specific data are available” (Table 6). Indeed, the Nutrition and Health National Survey (ENSANUT, for its Spanish acronym) in Mexico, as well as numerous other publications, use these cut-off points as a reference.

IDF article: https://doi.org/10.1111/j.1464-5491.2006.01858.x

ENSANUT example: https://doi.org/10.21149/14809

 

Line 130 -  “Individuals whose dietary data demonstrated inconsistencies were excluded from the analysis” should be in 2.2. Participants;

Response:

We found this relevant suggestion, and the suggested changes have been made.

Line 130 moved to line 99:

“Individuals whose dietary data demonstrated inconsistencies were excluded from the analysis.”

 

Line 132 – “Nutrient intake was adjusted for energy using the residual regression method” should have a brief explanation;

Response:

We appreciate this important suggestion, and the suggested changes have been made.

Line 135:

“This method estimates the residuals from a simple regression in which the nutrient of interest is the dependent variable and energy intake is the independent variable. The residuals from this model are then added to the mean, yielding the nutrient intake adjusted for total energy consumption [41].”

 

 

Line 176 – “tertile 1 (most anti-inflammatory diet) and tertile 3 (most pro-inflammatory diet)” is confusing; is there a tertile 2? How was denominated?  

Response:

Tertiles are statistical measures that divide a dataset, when ordered from lowest to highest, into three equal-sized groups, each containing roughly one-third of the data points. They help understand data distribution by identifying the lower, middle, and upper thirds of data. In our case, tertile 1 comprises the bottom 33% of DII scores and is therefore labeled the most anti-inflammatory diet, tertile 2 the middle 33% of DII scores, and tertile 3 the top 33% of DII scores, therefore labeled the most pro-inflammatory diet.  

Comparing the extreme groups (tertile 1 vs 3) and excluding the middle group (tertile 2) is a common practice in research. This increases the statistical power of the analysis, making it easier to detect a significant effect or relationship if one truly exists, because the larger the difference between the groups, the clearer the observed effect will be.

As described on line 145, negative scores indicate an anti-inflammatory diet, whereas positive scores indicate a pro-inflammatory diet. Shivappa and colleagues (2014) use similar language to describe the “maximal pro-inflammatory diet” when they talk about the highest score, and the “maximal anti-inflammatory diet” about the lowest score. Ren and colleagues (2018) also use the same language. However, we could consider using “diets” instead of “diet” to improve comprehension, as we are referring to a group of diets.

On the other hand, tertile 2 could be considered a “no effect” tertile, or, in our case, an “intermediary anti-inflammatory diet,” because it is oriented toward negative values. However, since we didn’t analyze it, we have decided not to mention it to avoid confusion.

Shivappa et. al., 2014. https://doi.org/10.1017/S1368980013002115

Ren et. al., 2018: https://doi.org/10.3390/nu10070831

We have made a modification in line 180 for better understanding:

“To compare IAS with anti-inflammatory diets to those with pro-inflammatory diets, participants were grouped into tertiles based on their DII score: tertile 1 (most anti-inflammatory diets) and tertile 3 (most pro-inflammatory diets).”

 

Discussion

Line 296 - Where is written “showing that there is a risk of increased obesity markers with each additional point in the DII score, regardless of the individual's position on the DII scale” please explain better “regardless of the individual's position on the DII scale”

Response:

The sentence “regardless of the individual's position on the DII scale” means that the coefficient in a model with a continuous independent variable reflects the change in the dependent variable (in this case kg/m² of the BMI or cm of the WC) associated with a one-unit increase in the independent variable (in this case DII score), and that this effect is the same regardless of the individual’s position on the scale. For example, the sentence in lines 296-298 indicates that each 1-point increase in the DII was associated with a 1.15 cm increase in WC, regardless of whether the 1-point increase means moving from a -4 DII score to a -3 DII score or from a 2 DII score to a 3 DII score, the associated effect would still be a 1.15 cm increase in WC in both scenarios.

Based on this recommendation, we have incorporated the following information into the statistical section of the methodology. The text is found from the lines 192 to 216:

“The β coefficient is interpreted in two ways, depending on how the independent variable is expressed. When the independent variable is categorical (Tertile 3 vs. Tertile 1 of the DII), the β coefficient represents the mean difference in the units of the dependent variables (kg/m2 for BMI, or cm for WC) between the persons that consumed the most pro-inflammatory diet (those in tertile 3) versus those with the most anti-inflammatory diet (Tertile 1). On the other hand, when the independent variable is expressed as continuous (such as the DII score), the β coefficient reflects the change in kg/m² of the BMI or cm of the WC associated with a one-unit increase in the DII score.

Similarly, the odds ratio (OR) is also interpreted in two ways, depending on how the independent variable is expressed. When the independent variable is categorical and is a risk related variable (example: Tertile 3 vs. Tertile 1 of the DII), the OR represents how many times great the likelihood is of having overall obesity or abdominal obesity for the group of people who consume a pro-inflammatory diet (those in Tertile 3) in comparison to those with the most anti-inflammatory diet (Tertile 1). On the other hand, when the independent variable is expressed as continuous (such as the DII score), the OR represents how many times greater the likelihood is of having overall obesity or abdominal obesity associated with a one-unit increase in the DII score. In both cases, the one-unit increase applies to any place on the continuous scale of the DII variable.

Assessing the DII both as a categorical and a continuous variable provides complementary information. As a DII categorical, it allows us to compare the lowest versus the highest DII categories; whereas, as a DII continuous, it allows us to evaluate the effect associated with each one-unit increase in the DII score.”

Additionally, we edited the sentence in the discussion section, line 345. Now it reads:

“… showing that there is a risk of increased obesity markers with each additional unit of increase in the DII score, regardless of the individual's position on the DII scale. For example, each 1-point increase in the DII was associated with a 1.15 cm increase in WC, regardless of whether the 1-point increase means moving from a -4 DII score to a -3 DII score or from a 2 DII score to a 3 DII score, the associated effect would still be a 1.15 cm increase in WC in both scenarios.”

 

The questions of the questionaire must be included as suplementary material. Moreover, any relation between the questions, diatary patterns and its trnaslation to nutrientes is explained or easily available for the reader.

Response:

The food frequency questionnaires (FFQs) were cited in lines 126 and 127. To access information regarding the FFQs, the reader can consult our previously published paper (citation number 39), which describes the design and validation of the Yaqui FFQ (doi:10.15174/au.2019.2248). The Seri FFQ is currently undergoing publication.

 

Thank you again for the opportunity to revise our manuscript. We hope that the changes made are satisfactory, and that our revised version meets the Requirements for publication in MDPI Nutrients.

 

 

Sincerely,

Julián Esparza-Romero

Corresponding Author

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

In the manuscript nutrients-4066478, the authors aimed to determine the association between dietary inflammatory index (DII) and obesity risk. This study has strength in conducting a cross-sectional study with a large number of participants; however, minor revision would be helpful.

1. In table 1-2, please show the characteristics and parameters of all the tertiles so that the readers compare T1, T2, and T3.
2. In table 2, please conduct ANOVA for all the parameters and present the p-values.
3. It would be better to split table 3-4. Please show the association between DII (as a continuous variable) and BMI or WC in one table by presenting beta coefficient and 95% CIs, and make another table for the association between DII (as a categorical variable) and the risk of obesity or abdominal obesity by presenting odds ratios and 95% CIs. For calculating odds ratios, logistic regression should be conducted.
4. The number of items in FFQ could affect the conclusion. Please mention that the difference in the number of items in FFQ between Seri and Yaqui as a limitation of the current study.
5. Please change the p-value 0.0000 to <0.0001.

Author Response

Dear Reviewer 2,

 

We sincerely thank you and the other reviewers for the time and effort dedicated to evaluating our manuscript titled "Dietary Inflammatory Index of Northern Mexican Indigenous Adults and Its Association with Obesity: Cross-Sectional Study."

We appreciate the constructive feedback and valuable suggestions, which have helped us improve the quality and clarity of our work. Below, we provide a point-by-point response to your comments. We hope the revisions satisfactorily address the concerns raised.

To facilitate tracking of changes, the manuscript's new additions, based on the reviewers’ suggestions, are highlighted in yellow. In the response letter, line numbers refer to the latest version of the manuscript.

 

1. In table 1-2, please show the characteristics and parameters of all the tertiles so that the readers compare T1, T2, and T3.

Response:

Based on the rationale we present for Table 2, we believe it is appropriate to retain the current structure of Table 1. Doing so allows us to maintain coherence and parallelism between the two tables, ensuring that the information is presented in a consistent and comparable manner. We respectfully consider that maintaining this alignment strengthens the clarity of the results.

 

2. In table 2, please conduct ANOVA for all the parameters and present the p-values.

Response:

There are two response variables (intake and score) for each tertile; adding tertile 2 would result in the addition of two more columns (one for intake and another for score), creating an overly complex table that would be difficult for the reader to understand. However, the effect of the complete variable on obesity, which includes all three DII tertiles, is studied in the main analysis using multiple regression (Tables 3 and 4). In this analysis, the differences or risks of obesity between tertiles 2 and 3 are compared with tertile 1 (the reference group).

Including tertile 2 in Tables 2 would not add meaningful information to the manuscript, as this group does not differ substantially from tertiles 1 and 3, and its inclusion would reduce the clarity and readability of the tables. Comparing the extreme groups (tertile 1 vs 3) and excluding the middle group (tertile 2) is a common practice in research. This increases the statistical power of the analysis, making it easier to detect a significant effect or relationship if one truly exists, because the larger the difference between the groups, the clearer the observed effect will be.

 

3. It would be better to split table 3-4. Please show the association between DII (as a continuous variable) and BMI or WC in one table by presenting beta coefficient and 95% CIs, and make another table for the association between DII (as a categorical variable) and the risk of obesity or abdominal obesity by presenting odds ratios and 95% CIs. For calculating odds ratios, logistic regression should be conducted.

Response:

We believe this modification may not be appropriate for the present report, and we would like to clarify the rationale. Tables 3 and 4 are differentiated by the type of regression analysis performed: linear regression (reporting β coefficients) and logistic regression (reporting odds ratios, ORs). Within each table, the structure is consistent, presenting four models: Models 1 and 2 correspond to BMI in Table 3 (or Obesity in Table 4), and Models 3 and 4 correspond to WC in Table 3 (or Abdominal Obesity in Table 4). Additionally, Models 1 and 3 assess the DII as a continuous variable, whereas Models 2 and 4 assess it as a categorical variable. Although these distinctions could theoretically be presented in additional tables, we consider the current format to be the most coherent and efficient way to present the information.

 

4. The number of items in FFQ could affect the conclusion. Please mention that the difference in the number of items in FFQ between Seri and Yaqui as a limitation of the current study.

Response:

We believe this does not represent a limitation on the analysis. Both questionnaires were designed to assess food intake over the last year and used the same methodology. Differences in the number of items are expected, given that the number of items represents their food dietary pattern. Each questionnaire provided the average dietary intake of individuals belonging to the respective Indigenous group (Yaqui or Seri). Therefore, we are confident that the difference in the number of items reflects a realistic situation between the two populations.

 

5. Please change the p-value 0.0000 to <0.0001.

Response:

We appreciate this important observation, and the suggested changes have been made in Table 1.

 

 

Thank you again for the opportunity to revise our manuscript. We hope that the

changes made are satisfactory, and that our revised version meets the

Requirements for publication in MDPI Nutrients.

 

Sincerely,

Julián Esparza-Romero

Corresponding Author

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript titled "Dietary Inflammatory Index of Northern Mexican Indigenous Adults and its Association with Obesity: A Cross-Sectional Study" addresses a critical and often overlooked area of nutritional epidemiology. The study provides significant insights into the nutritional transition of the Seris and Yaquis communities and explores how the inflammatory potential of their diet correlates with obesity markers. The overall quality of the research is high, and the findings are well-documented. However, to further enhance the impact and clarity of the paper, the following points should be considered:

Strengths and Core Contributions: The authors have done an excellent job of contextualizing the research within the unique socioeconomic and cultural framework of northern Mexico. The use of the Dietary Inflammatory Index (DII) is particularly appropriate here, as it captures the cumulative effect of diet on inflammation, which is a key driver of metabolic disorders. The statistical rigor, including the adjustments for age, physical activity, and modernization levels, adds substantial weight to the validity of the results.

Specific Points for Improvement

1. Data Visualization and Tables: While the tables are comprehensive, Table 2, which details the 30 DII food parameters, is quite dense. It would be beneficial for the reader if the authors could highlight or provide a brief summary in the text of the specific parameters that showed the most significant variation between the tertiles. This would allow for a more immediate understanding of which dietary components are driving the pro-inflammatory scores in these populations.

2. Discussion on Traditional vs. Modern Diets: The discussion section provides a solid analysis of the findings. However, a deeper exploration of the traditional food systems of the Seris and Yaquis could further enrich the narrative. Explicitly discussing which traditional anti-inflammatory foods are being displaced by ultra-processed items would provide a more nuanced understanding of the nutritional transition.

3. Policy and Public Health Recommendations: The authors conclude that anti-inflammatory dietary patterns should be promoted. It would be valuable to expand this into more specific, culturally tailored recommendations. For instance, how can traditional knowledge be integrated into modern health interventions to improve the DII scores within these specific indigenous contexts?

4. Methodology Clarification: Although the methods are well-described, a brief mention of the seasonal availability of certain foods and how this might have influenced the Food Frequency Questionnaire (FFQ) results would add another layer of methodological transparency.

Conclusion​: This is a well-executed study that makes a meaningful contribution to the field of indigenous health and nutrition. The English language is of high quality, and the logic is sound. Addressing the minor points regarding data presentation and expanding slightly on the cultural context of the diet will further strengthen this valuable piece of research

Author Response

Dear Reviewer 3,

 

We sincerely thank you and the other reviewers for the time and effort dedicated to evaluating our manuscript titled "Dietary Inflammatory Index of Northern Mexican Indigenous Adults and Its Association with Obesity: Cross-Sectional Study."

We appreciate the constructive feedback and valuable suggestions, which have helped us improve the quality and clarity of our work. Below, we provide a point-by-point response to your comments. We hope the revisions satisfactorily address the concerns raised.

To facilitate tracking of changes, the manuscript's new additions, based on the reviewers’ suggestions, are highlighted in yellow. In the response letter, line numbers refer to the latest version of the manuscript.

 

1. Data Visualization and Tables: While the tables are comprehensive, Table 2, which details the 30 DII food parameters, is quite dense. It would be beneficial for the reader if the authors could highlight or provide a brief summary in the text of the specific parameters that showed the most significant variation between the tertiles. This would allow for a more immediate understanding of which dietary components are driving the pro-inflammatory scores in these populations.

Response:

The parameters presented in lines 265 to 268 (β-carotene, vitamin D, vitamin C, magnesium, vitamin A, vitamin B6, onion, zinc, omega-3 fatty acids, fiber, vitamin E, selenium, chili pepper, and niacin) are those that most differentiate tertiles 1 and 3, with higher consumption among individuals in tertile 1. Similarly, a lower intake of these parameters results in higher DII scores, which leads to a pro-inflammatory diet in these populations. To improve the comprehension of this idea, the corresponding lines now read:

“This anti-inflammatory dietary contribution was characterized by higher intake of β-carotene, vitamin D, vitamin C, magnesium, vitamin A, vitamin B6, onion, zinc, omega-3 fatty acids, fiber, vitamin E, selenium, chili pepper, and niacin in subjects of tertile 1 compared with tertile 3 (p < 0.05). Accordingly, lower intake of these parameters brings higher DII scores, which leads to a pro-inflammatory diet in these populations.”

 

2. Discussion on Traditional vs. Modern Diets: The discussion section provides a solid analysis of the findings. However, a deeper exploration of the traditional food systems of the Seris and Yaquis could further enrich the narrative. Explicitly discussing which traditional anti-inflammatory foods are being displaced by ultra-processed items would provide a more nuanced understanding of the nutritional transition.

Response:

We identified in this observation a significant opportunity to enhance our discussion. Following your suggestion, we added the following paragraph in line 383:

“Despite existing geographic and cultural differences, the traditional diets of both ethnic groups may exert an anti-inflammatory effect. Regarding the Yaqui group, various authors have pointed out that their diet, practiced for over three centuries, was primarily characterized by the consumption of vegetables, whole grains, legumes, as well as wild fruits and regional plants [5,56]. This diet provides a significant amount of fiber, vitamins, and minerals, as well as a low intake of saturated and trans fats, which could be described as an anti-inflammatory diet. However, from the 1960s to the present, the introduction of processed foods (especially processed meats), the regular consumption of eggs, and the use of dressings in Yaqui communities have had adverse repercussions on the nutritional quality of their traditional food systems [5,56]. On the other hand, historical records of the Seri diet, dating back more than one hundred years, describe it as one based on seasonal consumption of food that was obtained from the practice of a very rudimentary fishing using harpoons, complemented by activities of collecting fruits and seeds of plants from the desert and the sea, as well as hunting animals in the region, mainly characterized by the consumption of sea turtles, fish, aquatic birds, and terrestrial animals, as well as a moderate intake of vegetables such as pitaya, seagrass, and maguey [57,58]. More recent studies characterize the traditional Seri diet as highly consumed of low-fat cereals, wheat and corn tortillas, and legumes, in addition to fish and seafood [7]. This new dietary pattern is rich in monounsaturated and polyunsaturated fatty acids, fiber, vitamins, and minerals. However, foods from Westernized diets, such as beef, chicken, processed meat products, sugar-sweetened beverages, and sweet desserts, have been incorporated into their diets [7]. This diet is characterized by a high content of saturated fats and added sugars, as well as a low contribution of fiber, vitamins, and minerals, which can lead to a pro-inflammatory profile. In this context, the incorporation of meats and ultra-processed foods replaces traditional eating patterns, transforming a potentially anti-inflammatory diet into a pro-inflammatory one, which increases the risk of developing chronic non-communicable diseases, such as obesity.”

 

 

3. Policy and Public Health Recommendations: The authors conclude that anti-inflammatory dietary patterns should be promoted. It would be valuable to expand this into more specific, culturally tailored recommendations. For instance, how can traditional knowledge be integrated into modern health interventions to improve the DII scores within these specific indigenous contexts?

Response:

Following this suggestion, we added this text to the conclusion in line 434:

“Promoting anti-inflammatory compound–rich foods that are already part of this population’s diet and recovering traditional foods from these communities that have anti-inflammatory potential.”

 

 

4. Methodology Clarification: Although the methods are well-described, a brief mention of the seasonal availability of certain foods and how this might have influenced the Food Frequency Questionnaire (FFQ) results would add another layer of methodological transparency.

Response:

We do not believe this could have an influence, since the questionnaires are also designed to include important seasonal foods for the diet. As part of the FFQ design, seasonal foods and traditional dishes that could not be captured through the 24-hour recalls were addressed through focus groups to learn about their relevance and periods of consumption (http://doi.org/10.15174/au.2019.2248). To clarify the text for the reader, it was edited for greater understanding, specifically in lines 126 and 129.

“Dietary intake was assessed using food frequency questionnaires (FFQs) designed to represent the average annual diet for each population. The Seri FFQ included 88 items, and the Yaqui FFQ included 123 [38,39]. Both FFQs included traditional dishes and seasonal food of each population.”

 

Thank you again for the opportunity to revise our manuscript. We hope that the changes made are satisfactory, and that our revised version meets the Requirements for publication in MDPI Nutrients.

 

Sincerely,

Julián Esparza-Romero

Corresponding Author

Author Response File: Author Response.pdf