Body Fat Mass and Risk of Cerebrovascular Lesions: The PRESENT (Prevention of Stroke and Dementia) Project
Round 1
Reviewer 1 Report
Thank you for nice paper and research.
My remarks:
Paper is dedicated to interesting hypothesis whether high absolute fat mass increases the risk of stroke independently, and the measurement of FM may be more reliable indicator of stroke risk compared to BMI, according to your introduction. This idea is quite clear, however, I wonder why you have not presented BMI of study subjects. I would suggest to add BMI (as traditional risk indicator) into table 1 (Basic demographic data) and try to compare the risks indicated by FM (novel approach) and BMI (older approach).
You have not done a comprehensive survey but extracted a sample of subjects from large cohort. Therefore it is very important to describe the process of extraction (using telephone) precisely (lines 75-83). For example, was it possible to detect by phone the history of dementia reliably (line 77)? Were milder forms of cognitive impairment (MCI, for example) excluded or not? I would suggest clearer description.
Figures 1 and 2 in Results are not informative enough, and I would suggest to present numbers in tables.
Minor grammar and style improvement is needed, avoiding redundancies (in introduction and discussions, line 26 in Summary).
Comments for author File: 
Comments.pdf
Author Response
Abote BMI : Our basic hypothesis is that fat mass itselt is the risk of stroke.According to this concept, the absolute amount os more important than index. The use of BMI, indicator of relative fat mass, was not used because it may cause more confusing results.
About methods: We have added method more detain in the text.
About MCI: We have added definition of dementia in the main text.
About figure: We replaced figure with table as recommend.
The english editing company (www.editage.com) reviewed our text. And before final submission, we will recheck english again.
Author Response File: 
Author Response.docx
Reviewer 2 Report
rewrite the manuscript as advised
Comments for author File: Comments.pdf
Author Response
Thank you for your kindly review. We revised the paper as your recommendation.
Author Response File: Author Response.docx
Reviewer 3 Report
In this study, the authors aim at demonstrating high FM may be an independent risk factor for ischemic stroke among stroke and dementia-free adults, especially in women. It was a real pleasure to read such an article. The introduction of the article is concise and clear. The study is very well performed, the experimental part is very complete. I am not an expert in bioelectrical impedance analysis, but I would really suggest to publish this article without any modifications.
Author Response
Thank you for kindly review
Reviewer 4 Report
The authors used a relatively large cohort of volunteers without previous detected stroke or dementia to analyse the relation between body fat mass (measured by BIA) and risk of silent brain infarction and/or white matter change. After adjusting for known stroke risk factors, the group of high fat mass women showed increased risk of SI/WMC.
The study is clearly designed and well conducted, but some points could be improved to increase manuscript quality and to better visualize data.
1- Authors should consider including two more tables containing the basic demographic data and the incidence of SI and the incidence of WMC according to tertiles. One table containing all the population and another separating men and women.
2- A separate analysis for SI and for WMC could be performed.
3- In previous PRESENT publication, skeletal muscle mass (SMM) is considered. The patients included here are the same from which SMM was obtained? Is it possible to calculate sarcopenia of these patients and to check relation with SI and WMC?
4- Some references are missing or located in sites that could lead to misinterpretations.
Lines 34-37 Please, add some reference/s
Lines 48-49 “After adjusting for other cardiovascular risk factors, BMI was not shown to be a high-risk factor for stroke (5-6)”
Ref.5: “Overweight and obesity are associated with progressively increasing risk of ischemic stroke, at least in part, independently from age, lifestyle, and other cardiovascular risk factors”… “The most important finding was the demonstration of a graded positive relationship of overweight and obesity with the incidence of ischemic stroke. Based on the pooled estimates of risk, overweight and obese individuals had, respectively, 22% and 64% greater probability of an ischemic stroke compared with normal-weight subjects. The relationship was not significantly different in men and women, nor did it differ in relation to the average blood pressure level of the populations examined or to the length of follow-up.”
Ref.6: “compared with participants of normal weight (BMI 18.5–24.9), relative hazard (95% confidence interval) of incident stroke was 0.86 (0.80 – 0.93) for participants who were underweight (BMI 18.5), 1.43 (1.36 –1.52) for those who were overweight (BMI 25–29.9), and 1.72 (1.55–1.91) for those who were obese (BMI 30). The corresponding relative hazards were 0.76 (0.66 – 0.86), 1.60 (1.48 –1.72), and 1.89 (1.66 –2.16) for ischemic stroke and 1.00 (0.89 – 1.13), 1.18 (1.06 –1.31), and 1.54 (1.27–1.87) for hemorrhagic stroke. For stroke mortality, the corresponding relative hazards were 0.94 (0.86 –1.03), 1.15 (1.05–1.25), and 1.47 (1.26 –1.72). Linear trends were significant for all outcomes (p 0.0001)”…“These results suggest that elevated BMI increases the risk of both ischemic and hemorrhagic stroke incidence, and stroke mortality in Chinese adults. Mixed men and women.”
On basis of these results and conclusions, I would not say, “BMI was not shown to be a high-risk factor for stroke”.
Lines 71-71 & 82-83 Repeated lines, please reorganize paragraph
Lines 149-152 “higher values of total FM conferred an increased risk of developing cerebral infarctions. However, only women showed statistical significance even after adjustment. These findings were consistent with those in previous reports”.
Ref.5: They found no differences between men and women.
Ref.6: Mixed men and women no differences between sexes reported.
Ref.7: Only male cohort, impossible to give support to your findings about incidence in women.
Ref.8: Only male cohort, impossible to give support to your findings about incidence in women.
Ref.9: Only male cohort, impossible to give support to your findings about incidence in women.
Authors should discuss better this result. To find articles comparing men vs women and stroke depending on BMI or tFM. Then discuss similarities or differences with their findings.
Author Response
1. About demographic date: Thank you for your good comment . However, we think that demographic data dose not take high-portion of entire paper. We think separating the table will further cloud the paper. More over according to another reviewer's request, two more table are added. It is difficult to add more tables.
2. Thank you for good advice. We used CT scan instead of MRI scan because of budget problem. Due to limitation of CT resolution, we could not separate SI and WMC.
3. We think that if we can adjustment the presence of SI/WMC by sarcopenia the better results will be obtained. For use sarcopenia as confounding variable, We need definition of sarcopenia. There is no definition of sarcopenia by BIA method. The study is too small to analysis result using muscle mass fat mass as continuous variable. It is limitation of this study.
4-1 We add reference at line 37
4-2 What we are try to say is not that "BMI is not risk factor" but BMI is not high risk factor" .As mentioned in Ref 5, RR of is obesity is high but not very high [1.26 (95% CI 1.07~1.40) ]. Comparison with figure 1(before adjustment with major stroke risk factor) and table 2( after adjustment of major risk factor) in reference 5, we can see that RR was reduced after adjustment.
4-3 We delete repeated line.
4-4 We deleted this sentence .
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
it is still not clear why author discuss about BMI when author wants to identify association between FM and strokes. Introduction should be rewritten according to author's hypotheses.
it is still not clear how many people are recruited
it is still not clear whether this was a cross sectional study or observational study
i also suspect that methodology may have written by someone or have taken from previously published papers since the flow and grammar was quite good compared to introduction
I also cant understand the description for statistics for tables For an example Table 2 What do you mean by Total 1 Need proper descriptions for each variable in the table and type of statistics that used, significant value I didn’t go through the discussion since I am not happy with this.
Author Response
1 Thank you for your kindly review. What we claim is that BMI is not a good indicator. We think it is more theoretically correct to classify body mass into muscle and fat. Most people with high BMI have lot of fat, so BMI and stroke are thought to be related. previously, it was difficult to measure fat mass separately, but now we can measure it with bio-impedance method such and InBody machine. So we think it would be better to measure body fat directly and use it as a health care marker than BMI. I'm sorry, but it is hard to rewrite the introduction because the other three reviewers agreed on our introduction.
2. Sorry for confusion. We have described a method more detail.
3. We think our study is more closer to cross-sectional study.
4. The gramma will be checked again by professional English editing agency.
5.Total T1( No Total1) , Total means men+women. We have modified table. the definition of the variables used in the table are described in the text. We have described a method more detail.
Author Response File: Author Response.docx
