Body Mapping as Risk Factors for Non-Communicable Diseases in Ghana: Evidence from Ghana’s 2023 Nationwide Steps Survey

Reviewer 1 Comment

Response

Reference

An interesting epidemiological study in which, based on a sample of 5,775 participants (the population of Ghana was studied), the relationship of a number of medical and demographic (age, gender, marital status, etc.) and anthropometric characteristics with the risk of developing a number of the most common non-communicable diseases was investigated. The study found that anthropometric indicators such as waist circumference, hip circumference, BMI, and waist-height ratio are closely related to the risk of developing major non-communicable diseases such as hypertension, diabetes, dyslipidemia, and cardiovascular diseases. There are the following suggestions for this work:

Thank you for the excellent summary of our work and finding this work as interesting.

-

1. The rationale for relevance in the abstract of the work is very large, this material should be significantly reduced to 2-3 sentences.

Thank you for this guidance, we have accordingly summarized the relevance statement  to 3 sentences.

Page 1

2. In the abstract before the materials and methods section, the purpose of the work should be clearly stated.

Thank you for this suggestion. We made efforts to clearly state the purpose as suggested.

Page 1

3. In the introduction of the work, data should be provided on the forecast of the increase in the prevalence of the most common noncommunicable diseases studied in the work among the world's population for the coming decades.

Thank you for this suggestions which we believe strongly that addressing this greatly improved the introduction section of the manuscript.

Page 2, lines 70 – 76`

4. In the introduction, the authors cite literature data showing the importance of obesity as a risk factor for a number of diseases, including among the Ghanaian population (lines 92-109). At the same time, it is necessary to clearly identify which unresolved issues remain on the topic of this study so that the novelty of this work would be clearly understood.

Thank you for highlighting this. We have addressed that and made mention of the gaps

Page three, lines110-119

5. In the materials and methods, the authors indicate that they used WHO approaches to classify BMI and other anthropometric parameters. References to relevant literary sources should be provided.

Two references have been provided as directed. Thank you

Page 6, line 216  -219

6. In Tables 3 and 4, the authors provide data on the odds ratio and their confidence intervals adjusted for confounders. The notes to these tables should include a list of confounders that the authors took into account in their statistical analysis

The list of confounding variables were added as suggested

Page 12, lines 340-350

Reviewer 2 Comment

Response

Reference

1. Firstly, the abstract is clearly too long and full of unnecessary general information.

Thank you for pointing this out to us. We have made significant steps to reduce the number of words of the abstract, by summarizing and removing seemingly unnecessary general information. Word count reduced from 389 to 323

Page 1, line 23 -52

2. The sample size is fine and appropriately designed. But if you say that pregnant women are excluded from the anthropometric measurements, then pregnant women should also be completely excluded from this study. How many of the respondents were actually pregnant?

Question M8 of the WHO steps instrument ask:

For women: Are you pregnant?

Yes 1  If Yes, go to M 16

No   2

From the data, 167 (4.9%) of women reported being pregnant and were excluded from anthropometric measurement as well as from this analysis

Line 163 – 165 was added which reads: “From the data, 167 (4.9%) of women reported being pregnant and were excluded from anthropometric measurement as well as from this analysis”

3. It should be noted that the biochemical parameters were determined from capillary blood.

“biochemical parameters were determined from capillary blood” was added to lines 195 -196

Page 5, Lines 195 -196

4. Blood pressure was measured with automatic measuring devices (Omron) – which type?

In general, it is necessary to specify the type/model for all devices used

Thank you for reminding us of this omission. We have updated including  the type. For the BP, information was updated as “Omron body composition monitor, Model HBF-701, Karada Scan”

Page 5, Line 167-168

5. The WHO recommendations for BMI are not given to 3 decimal places https://www.who.int/europe/news-room/fact-sheets/item/a-healthy-lifestyle---who-recommendations  The same applies to waist and hip circumference: chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/

https://iris.who.int/bitstream/10665/44583/1/9789241501491_eng.pdf

If you have other data, please provide the reference.

Thank you for this useful comment. Decimal places in table 1 were changed to reflect your recommendation of one-decimal place.

Table 1, line 247

6. It is not necessary to reproduce all the results from the table in text form in the Results chapter.

Some of the results can be moved to an appendix, especially as they are reproduced in the main text

Thank you. Some sections of the result, text that were deleted

Page 248 -332

7. Although the topic is important, the results presented are familiar and offer no new methodological or theoretical insights.

The discussion is sparse and only some of the results are discussed. The discussion lacks in-depth analysis in the context of literature and practice. The conclusions are general and already known.

Thank you, the results section have been expanded with some in-depth analysis included to address this comment.

Pages 13 - 15

8. There are also no practical recommendations or suggestions for further research.

Recommendation for practice and research has been included.

Line 432 - 437

9. English editing is required.

The manuscript was subjected to English language editing

Reviewer 2 Comment

Response

Reference

1. The instructions for authors state: A single paragraph of about 200 words maximum
Therefore, I thank the authors for their effort, but shortening it to 323 words is better, but by no means satisfactory.

Thank you for helping us conform to the journal specification.

The Abstract is now reduced to 200 words

Page 1, lines 24- 42

2. If the calculated sample size required for this research to be significant is 5775 respondents, how could you only take 5207, which is 10% less than the required number, it is known that at least that much is needed, and it should be taken into account that you have excluded a further 167 pregnant respondents, so the number is even smaller.

Also, the number you gave of 4.9% of women would correspond to a total number of 3408 women in the population sample, which does not fit with the results presented in your paper.

Thank you for this comment. The sample size calculation adjusted for a response rate of 80%. However, this study had a response rate of 94.04% (5412/5775). It is obvious from your comment that, we did not explain this section very well in the manuscript for our readership to understand. We have decided to include in the sample size calculation section the full details of the sample size calculation as captured in the proposal of the study. Please find the insertion below

To determine the sample size necessary to obtain the desired population estimate for the key NCDs risk factors survey,the WHO recommends using the formula: n = z2Dp(1-p)/e2, where:

n = Sample size

D = Design effect

p = Estimated proportion or prevalence within the target population at the time of the first survey

z = Level of confidence in the sample mean or prevalence as an estimate of the population mean

or prevalence

e = Margin of error

In this study, p represents the nationwide prevalence of any NCD-related risk factors. As this was the first nationally representative survey to provide prevalence data on key risk factors for NCDs in Ghana, WHO STEPS survey protocol recommends assuming that 50% of participants will have at least one risk factor (such as tobacco use, alcohol use, unhealthy diet, or physical inactivity) for the purposes of sample size estimation. In cases where p is unknown, it is often assumed to be 50%.

To calculate the minimum sample size (n), the following assumptions were made:

z = 1.96 (corresponding to a 95% confidence level)

D = 1.5 (design effect)

p = 0.5 (or 50%)

e = 0.05 (or 5% margin of error)

Using these parameters, the minimum sample size obtained is 576. This sample size is then adjusted

for the number of strata, which in this case is 8 (comprising four age groups: 18-29, 30-44, 45-59, 60-69, and two sex categories: male and female). Additionally, the sample size is adjusted for an expected response rate of 80% (0.8).

The final adjusted sample size for the survey was 5,775, based on the distribution to EAs. Since one

participant is selected per household, this number is equivalent to the number of households to be

selected. With a fixed number of 15 households per EA, a total of 385 EAs was required to achieve the desired sample size.

Also, Yes, 167/ 3412 (4.8944% =4.9%) pregnant women were excluded from all anthropometric measurements.

Lines 115 -159

And table 2

3. Furthermore, the model of the blood pressure monitor used to measure blood pressure has still not been specified. The "Omron body composition monitor, model HBF-701, Karada Scan", which you mention in the answers as a BP model, but not in the text of the manuscript itself, does not measure blood pressure as far as I know.

Thank you for pointing this out. It has been corrected.

The Omron body composition monitor, model HBF-701, Karada Scan, was used for weight measurement and for Blood pressure measurement, Omron automatic Blood pressure monitor, Model: M6 AC (HEM-7322.E) was used

Lines 155 - 170

4. As for the decimal places, thank you for the correction, but other items in the text should also have been checked, such as the year, which is given with two decimal places, and the body weight and height, which, according to the measuring instruments mentioned, cannot possibly be given with the accuracy of the number of decimal places given.

Thank you for teaching us this that we often overlook. We have revised the decimal points to conform with the measuring instrument’s level of accuracy.

Lines 233- 268)

5. Within the results there are 5 tables with the same results (some of which burden the main manuscript), which could have been included in the appendix. In this sense, no changes have been made.

Thank you, table 3 and table 5 were deleted from the main manuscript and added as supplementary tables 1 and 2 respectively

Supplementary tables

Reviewer 2 Comment

Response

Reference

The manuscript has improved in some parts compared to the earlier draft. The authors have responded to several of the reviewer's comments and have, among other things, expanded the description of the methodology and formulated parts of the results and discussion more clearly. The overall presentation is clearer.

All thanks to you

In the list of references related to the discussion, the references are missing, so you can't see them,

Thank you for bringing our attention to this. We have worked on the references

Lines 367 - 578

and it seems to me that the comparison with data from regionally or economically close countries is missing.

We have included comparison with data from regionally or economically close countries

“Comparable evidence from sub-Saharan African settings reinforces the limitations of relying solely on anthropometric measures to predict NCD risk. In Tanzania, a community-based study in Dar es Salaam demonstrated that individuals with elevated waist circumference had a 34% higher likelihood of hypertension, underscoring the measure’s utility as a pragmatic, cost-effective screening tool [38]. Similarly, a four-year cohort study in Malawi found that waist circumference, particularly when combined with fasting plasma glucose, significantly improved the prediction of progression from impaired fasting glucose to type 2 diabetes, yielding an area under the receiver operating characteristic curve (AUC) of 0.79 [39]. These findings affirm that while waist circumference enhances prediction modestly at the population level, it does not achieve high discriminative accuracy for individual-level screening.”

Lines 433 -444