Dietary Behavior of Spanish Schoolchildren in Relation to the Polygenic Risk of Obesity

Round 1

Reviewer 1 Report

The manuscript should be accepted for the publication and requires minor-revision according to the comments to the authors.

The article written on English language has 16 pages, 2 tables, 2 figures, and 43 references and consists of sections: Abstract, Introduction, Materials and Methods, Results, Discussion, Conclusion, and Appendix. The article presents the results of a research conducted among schoolchildren in City of Madrid and contributes to the knowledge of the interactions between genotype and eating behavior in schoolchildren and adolescents.

 The Abstract reflects accurately the contents of the manuscript.

The Introduction presents the current knowledge about the association between single nucleotide genetic polymorphism, the obese phenotype and eating behavior, and states the problem and background adequately.

In the Methodology section, in subsection Participants, sampling method should be described thoroughly.   Instruments, Genetic analysis, Anthropometric study and Statistical procedures are described adequately.

In the Results section, the data in text and tables are sufficient to answer the study questions with enough details and related to the main point. The data on sample characteristic and the data of sample coverage should be added.

The discussion logically derives from the obtained results of statistical processing, which were put into context and compared with the results of similar surveys in appropriate manner. The weaknesses of the study as well as suggestions for further research are described adequately. The strength of the study should be added.

The conclusions logically derive from the obtained results of statistical processing.

The number and choice of references is satisfactory.

Author Response

Thank you very much for your comments. We have made minor changes suggested by other authors. We greatly appreciate your time and dedication.

Reviewer 2 Report

This paper analyses the anthropometric descriptors, eating behaviour and the genotype profile of Spanish children in order to establish an association to the risk of obesity.

The cohort and the associations are performed correct and the manuscript, especially its discussion is written well. However, some of the presentation, text, tables and results section need to be adjusted and rearranged, in order to be accepted for publication.

The introduction is lacking a clear focus (with many small mistakes), as well as the methods and results (rewrite abstract, rearrange results: introduce the description of the population better, and thus start with a new subheader, 3.1 Anthropometric and genetic description of the sample; 3.2 Genetic risk score for obesity and eating behavior, and re-number the Tables accordingly). A table now as supplementary data, should be repositioned in the results.

 

Please address the following issues in your re-submmitted paper:

Abstract

Even though the indication of significant associations (OR interval and p-value) is correct, the authors should reduce this kind of notification in the abstract, as including the named SNPs consequently makes the abstract non-readable. It is suggested to include that full description in the beginning of the discussion, and rewrite the abstract in a more reader-friendly way.

 

Introduction

Line 46/49 should read: ‘With this approach, research groups such as Jääskelainen et al. [4], Llewellyn et al. [5], Steinbeck et al. [6], and Monnereau et al. [7], have associated eating behavior with polygenic risk, quantified from a range between 8 and 32 SNPs.’

Line 52 perhaps this is more clear: ‘It has also been shown that an infant’s self-regulation of feeding arises…’

Line 59/61 should read ‘A pioneering study from the United Kingdom by Wardle et al. [11] studied the association…. control in children.’

Suggestion to re-structure Lines 50/68 according to:

Other investigations have related obesity with lower responsiveness to internal satiety signals and greater receptivity to olfactory or gustatory stimuli, which would lead to a higher caloric intake [8]. It has also been shown that an infant’s self-regulation of feeding arises from genetic-environmental interactions responsible for behaviors such as fast eating and delayed satiety [8,9]. Hardle et al. [10] investigated the role of eating behavior in genetic susceptibility to obesity and their thorough literature analysis over the decade 2010-2020, concluded that emotional and uncontrolled eating behaviors lead to being overweight.

A pioneering study from the United Kingdom by Wardle et al. [11] investigated the association of the SNP rs9939609 genotype of the FTO gene with indicators of adiposity and satiety control in children, and proves the usefulness of psycho-metric scales (Satiety Responsiveness and Enjoyment of Food and the Child Eating Behaviour Questionnaire) as an objective measure of eating behavior. Which is also emphasized by Hardle’s literature review [10] in which the authors highlight the value of psycho-metric tests that objectively measure attitudes toward food.

Undoubtedly, the development of eating habits is a process with a complex etiology, including psychological aspects especially the influence of family, social, and affective ones [??]. Some research has highlighted a particular circumstance, reporting that in overweight subjects, there is a positive relationship between an emotional eating pattern and body mass index [12]. Hence, food is not eaten to satisfy physiological hunger but also to fill an inner emptiness or comfort the mood.

Lines 72/74 should read: Therefore, this study aims to analyze the eating behavior of a sample of Madrid schoolchildren concerning their genetic pre-disposition to obesity through conduction of the CEBQ "Children's Eating Behaviour Questionnaire" in combination with genotyping assessment from a panel of 32 SNPs.

 

Methods:

Lines 77/79 should read: ‘Data were collected between 2019-2021 during a cross-sectional descriptive study, in which 258 schoolchildren aged 6-16 years (67.83% male; 32.17% female) from different schools in the Community of Madrid, Spain, participated.’

Lines 86/87 should read: ‘Parents or guardians completed the CEBQ (Child Eating Behaviour Questionnaire) to assess eating behavior.’

Lines 89/90 should read: ‘The CEBQ [14] is a validated questionnaire that enables to know the participant’s responses to satiety, taste for food, speed of intake, and emotional food consumption, among other parameters.’

Lines 101/103 should read: ‘The present study used a Spanish-language translated and adapted version of the CEBQ that was previously validated among Spanish and Chilean populations [15].’

Rearrange lines 104/118 and adjust references, according to:

Merge lines 105 and 115/118:

A total of 32 SNPs were analyzed: all risk alleles of the selected SNPs have been associated in previous research with childhood and adolescent obese phenotype or eating behavior [18-20]. The SNPs selected are reported in Table A1 (Supplementary file: Table A1).

The extraction protocol used was that of BioTools B&M S.A. laboratories for saliva [16], and the "Speedtools Tissue DNA Extraction" kit was used. Subsequently, genotyping was carried out in the laboratories of the University of Santiago de Compostela Node of the National Genotyping Center (CeGen), which is part of the Network Platform of Bio-molecular and Bioinformatics Resources of the Carlos III Institute of Health. The iPLEX® Gold technology for the MassArray platform of Agena Bioscience Inc [17] was used. Once genotyping was completed, all caruncles were burned for the data protection of the participants.

Line 122: Should PRG be replaced by GRS? And accordingly throughout the whole manuscript (please check lines 120/127 and line 142 for this).

Line 122 should read: ‘The total genetic risk score (GRS) was obtained from the sum of the score of the 32 SNPs,’

 

Results

Introduce the description of the population better, and thus start with a subheader, and re-number the Tables accordingly. Explain ‘ICT’, should that be ‘WHR’?

 

3.1 Anthropometric and genetic description of the sample.

The anthropometric description of the sample participating children is given in Table 1 (=currentTable A3, as this information should be given up-front and included in the main body of the manuscript. Please note to spell ‘weight’ in Table A3). The genotypic frequencies and risk allele frequency of each selected SNP are presented in Table A2 (please note that ‘European’ should be positioned on a single line), and it can be observed that no allele showed a frequency less than 0.05, which corresponds with the minimal data quality criterion for the association study (Hardy-Weinberg equilibrium).

As shown in Table 1 (will be 2), the proportion of overweight and obesity categorized by BMI is significantly higher among subjects with a higher genetic predisposition to obesity (p=0.040). The WHR and %BF trend also reflects a higher prevalence of overweight or adiposity in genetically predisposed individuals without significance.

 

Table 1: Why are numbers indicated after main descriptors BMI, WHR and %BF? Please, make them superscript without brackets.

Line 168: change to: ‘3.2 Genetic risk score for obesity and eating behavior’

Lines 169/171 should read:

Based on the anthropometric and genotypic description of the sample participant the resulting genetic risk score (GRS) categories were Q1: ≤19 points; Q2: 20-22 points; Q3: 23-25 points; and Q4:>25 points. Table 2 presents the mean scores of the 8 subscales of the CEBQ, stratifying the results according to GRS quartile categories. The trend is increasing in the pro-eating scales and decreasing in the anti-eating scales.

Lines 195/199 should read: ‘To investigate how much the risk allele of each SNP increases the possible obesogenic behavior, the inverse OR values were plotted when constructing Figure 2 corresponding to the anti-intake subscales. Thus, obesogenic behaviors are positioned on the right of Figure 2, and non-obesogenic behaviors on the left.’

Make sure to have Figure 1 and Figure 2 more separately printed in the manuscript. For example by positioning Figure 1 at the end of section 3.1, before starting section 3.2

 

Discussion

Lines 245/249 should read: ‘The results with other subscales are generally in line with expectations that the average scores have been increasing with higher genetic risk scores in the pro-intake scales, and decreasing in the anti-intake scales, following the trend of previous studies, which highlight the more significant evidence concerning the ability to satiety, and the demand to food [10].’

Line 251 should read ‘a more significant response to, or enjoyment of food, more appetite, and less responsive’

Line 255/257 should read ‘The literature concludes that children who enjoy food more, or express a greater desire for food, consume more daily energy on average [31].’

 

Conclusion

Line 308 should read: ‘However, the present study highlights the role in subscales related to emotional eating, either by over’

Please address the following issues in your re-submmitted paper:

Abstract

Even though the indication of significant associations (OR interval and p-value) is correct, the authors should reduce this kind of notification in the abstract, as including the named SNPs consequently makes the abstract non-readable. It is suggested to include that full description in the beginning of the discussion, and rewrite the abstract in a more reader-friendly way.

 

Introduction

Line 46/49 should read: ‘With this approach, research groups such as Jääskelainen et al. [4], Llewellyn et al. [5], Steinbeck et al. [6], and Monnereau et al. [7], have associated eating behavior with polygenic risk, quantified from a range between 8 and 32 SNPs.’

Line 52 perhaps this is more clear: ‘It has also been shown that an infant’s self-regulation of feeding arises…’

Line 59/61 should read ‘A pioneering study from the United Kingdom by Wardle et al. [11] studied the association…. control in children.’

Suggestion to re-structure Lines 50/68 according to:

Other investigations have related obesity with lower responsiveness to internal satiety signals and greater receptivity to olfactory or gustatory stimuli, which would lead to a higher caloric intake [8]. It has also been shown that an infant’s self-regulation of feeding arises from genetic-environmental interactions responsible for behaviors such as fast eating and delayed satiety [8,9]. Hardle et al. [10] investigated the role of eating behavior in genetic susceptibility to obesity and their thorough literature analysis over the decade 2010-2020, concluded that emotional and uncontrolled eating behaviors lead to being overweight.

A pioneering study from the United Kingdom by Wardle et al. [11] investigated the association of the SNP rs9939609 genotype of the FTO gene with indicators of adiposity and satiety control in children, and proves the usefulness of psycho-metric scales (Satiety Responsiveness and Enjoyment of Food and the Child Eating Behaviour Questionnaire) as an objective measure of eating behavior. Which is also emphasized by Hardle’s literature review [10] in which the authors highlight the value of psycho-metric tests that objectively measure attitudes toward food.

Undoubtedly, the development of eating habits is a process with a complex etiology, including psychological aspects especially the influence of family, social, and affective ones [??]. Some research has highlighted a particular circumstance, reporting that in overweight subjects, there is a positive relationship between an emotional eating pattern and body mass index [12]. Hence, food is not eaten to satisfy physiological hunger but also to fill an inner emptiness or comfort the mood.

Lines 72/74 should read: Therefore, this study aims to analyze the eating behavior of a sample of Madrid schoolchildren concerning their genetic pre-disposition to obesity through conduction of the CEBQ "Children's Eating Behaviour Questionnaire" in combination with genotyping assessment from a panel of 32 SNPs.

 

Methods:

Lines 77/79 should read: ‘Data were collected between 2019-2021 during a cross-sectional descriptive study, in which 258 schoolchildren aged 6-16 years (67.83% male; 32.17% female) from different schools in the Community of Madrid, Spain, participated.’

Lines 86/87 should read: ‘Parents or guardians completed the CEBQ (Child Eating Behaviour Questionnaire) to assess eating behavior.’

Lines 89/90 should read: ‘The CEBQ [14] is a validated questionnaire that enables to know the participant’s responses to satiety, taste for food, speed of intake, and emotional food consumption, among other parameters.’

Lines 101/103 should read: ‘The present study used a Spanish-language translated and adapted version of the CEBQ that was previously validated among Spanish and Chilean populations [15].’

Rearrange lines 104/118 and adjust references, according to:

Merge lines 105 and 115/118:

A total of 32 SNPs were analyzed: all risk alleles of the selected SNPs have been associated in previous research with childhood and adolescent obese phenotype or eating behavior [18-20]. The SNPs selected are reported in Table A1 (Supplementary file: Table A1).

The extraction protocol used was that of BioTools B&M S.A. laboratories for saliva [16], and the "Speedtools Tissue DNA Extraction" kit was used. Subsequently, genotyping was carried out in the laboratories of the University of Santiago de Compostela Node of the National Genotyping Center (CeGen), which is part of the Network Platform of Bio-molecular and Bioinformatics Resources of the Carlos III Institute of Health. The iPLEX® Gold technology for the MassArray platform of Agena Bioscience Inc [17] was used. Once genotyping was completed, all caruncles were burned for the data protection of the participants.

Line 122: Should PRG be replaced by GRS? And accordingly throughout the whole manuscript (please check lines 120/127 and line 142 for this).

Line 122 should read: ‘The total genetic risk score (GRS) was obtained from the sum of the score of the 32 SNPs,’

 

Results

Introduce the description of the population better, and thus start with a subheader, and re-number the Tables accordingly. Explain ‘ICT’, should that be ‘WHR’?

 

3.1 Anthropometric and genetic description of the sample.

The anthropometric description of the sample participating children is given in Table 1 (=currentTable A3, as this information should be given up-front and included in the main body of the manuscript. Please note to spell ‘weight’ in Table A3). The genotypic frequencies and risk allele frequency of each selected SNP are presented in Table A2 (please note that ‘European’ should be positioned on a single line), and it can be observed that no allele showed a frequency less than 0.05, which corresponds with the minimal data quality criterion for the association study (Hardy-Weinberg equilibrium).

As shown in Table 1 (will be 2), the proportion of overweight and obesity categorized by BMI is significantly higher among subjects with a higher genetic predisposition to obesity (p=0.040). The WHR and %BF trend also reflects a higher prevalence of overweight or adiposity in genetically predisposed individuals without significance.

 

Table 1: Why are numbers indicated after main descriptors BMI, WHR and %BF? Please, make them superscript without brackets.

Line 168: change to: ‘3.2 Genetic risk score for obesity and eating behavior’

Lines 169/171 should read:

Based on the anthropometric and genotypic description of the sample participant the resulting genetic risk score (GRS) categories were Q1: ≤19 points; Q2: 20-22 points; Q3: 23-25 points; and Q4:>25 points. Table 2 presents the mean scores of the 8 subscales of the CEBQ, stratifying the results according to GRS quartile categories. The trend is increasing in the pro-eating scales and decreasing in the anti-eating scales.

Lines 195/199 should read: ‘To investigate how much the risk allele of each SNP increases the possible obesogenic behavior, the inverse OR values were plotted when constructing Figure 2 corresponding to the anti-intake subscales. Thus, obesogenic behaviors are positioned on the right of Figure 2, and non-obesogenic behaviors on the left.’

Make sure to have Figure 1 and Figure 2 more separately printed in the manuscript. For example by positioning Figure 1 at the end of section 3.1, before starting section 3.2

 

Discussion

Lines 245/249 should read: ‘The results with other subscales are generally in line with expectations that the average scores have been increasing with higher genetic risk scores in the pro-intake scales, and decreasing in the anti-intake scales, following the trend of previous studies, which highlight the more significant evidence concerning the ability to satiety, and the demand to food [10].’

Line 251 should read ‘a more significant response to, or enjoyment of food, more appetite, and less responsive’

Line 255/257 should read ‘The literature concludes that children who enjoy food more, or express a greater desire for food, consume more daily energy on average [31].’

 

Conclusion

Line 308 should read: ‘However, the present study highlights the role in subscales related to emotional eating, either by over’

Author Response

We greatly appreciate the comments made. We have carefully read all your suggestions and have incorporated them point by point into the final document. Thanks to you, our work is much more competent. Thank you again.

Reviewer 3 Report

A paper with a good structure, on a major health issue.

It would be in great interest if you could extend the research on a higher number of participants, in order to achieve some significant results. 

Author Response

Thank you very much for your time and dedication. We hope to replicate the work with a much larger sample.