Key Risk Factors, Sex Differences, and the Influence of High-Intensity Exercise on Colorectal Carcinogenesis: A 10-Year Cohort Study Based on 1,120,377 Individuals from the NHISS Data

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Please describe the study design section more in depth, longitudinal is insufficient. 

How and why were the 21 variables selected (patients vs control group)? Explain.

The last paragraph of 2.1 repeats the aim of the study, which is part of the introduction's last paragraph. They should be merged in the introduction. 

2.2: describe in- and exclusion criteria and how random selection occured.

2.3: which exercise related questionnaires did you use? validation? 

The materials and methods section is lacking sources.

3.1: explain PSM, how does it work?

 

Tables: the legend should be in every page, not only at the end. Similarly for the titles of the columns: they should be repeated on every page to increase the readability. Significant p-values need to be highlighted somehow. In general the tables are numerous and very extensive, which makes it difficult and hard to read.

 

Table 2 and 3 are in the wrong order. It should be chronologically. Table 3's red text are the significant p-values?

 

3.3 a space is lacking in this sentence: we observed that engaging in 231 high-intensity exercise3 days per week was significantly associated 

 

 

 

Figure 2: explain the dotted line in the legend.

 

The discussion needs a first paragraph, summarising the research question(s)/aim(s)/objective(s).

 

What does the 'respectively' stands for?

Globally, mortality rates attributed to smoking, overweight, obesity, and lack of ex-292 ercise were 21% and 2%, respectively 

 

ACSM guidelines of 2019 are not the most up to date. The KNGF guidelines for exercise in cancer were published in 2022 (Dutch) and 2023 (English).

 

The discussion needs subtitles to structure it better.

 

The second 'twice a week' is a mistake, I assume? This is confusing...

 

Regarding frequency, the number of patients with 316 cancer vs. noncancer exercising includes twice a week (227±8.10 vs. 208±7.42), twice a 317 week (147±5.25 vs. 200±7.14), four times per week (74±2.64 vs. 89±3.18), five times per week 318 (82±2.93 vs. 74±2.64), six times per week (46±1.64 vs. 31±1.11), and daily (94±3.35 to 319 78±2.78). Therefore, 3~4 times per week seems most appropriate. 

 

What about the obesity paradox in oncology? Please elaborate on this, because now the statement is too narrow.

Maintaining an optimal weight is crucial for overall health, characterized by achiev-348 ing an ideal body composition that includes high skeletal muscle and low-fat mass. 

Author Response

Dear Reviewer,

I appreciate very much for developing the manuscript.

Please see the attached Word file for my answers.

Thank you very much. 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This is an interesting study, but I do have some concerns/edits/comments:

1. On line 69 the authors state that this was a longitudinal study. Was it a longitudinal study or a case-control study, by the way the authors selected cases and then the controls? Not clear, needs to be revised.

2. On line 95 the authors wrote "In South Korea, all citizens are required to undergo medical checkups ..." How often are these required medical checkups and stating which age? Needs some clarification.

3. On line 115 the authors wrote "The variable for alcohol consumption was defined as “number of drinks per week”". What type of drinks? High alcohol level drinks are very different from low alcohol level drinks. Why did the authors not classify heavy drinkers, binge drinkers, etc?

4. On lines 118-119 the authors wrote "Additional variables included height (cm), weight (kg), waist circumference (cm), body mass index (kg/m2), ..." Since the BMI is highly correlated with height and weight, they should not be included in the same model. 

5. How did the authors check to make sure that highly correlated variables were not included in the same model? Need more details in here.

6. Since this was a national level study, did the authors employ a logistic regression or a weighted logistic regression model? An unweighted logistic model may significantly bias the final findings. Did the authors consider employing a conditional logistic regression model?

7. On lines 140-141 the authors wrote "The optimal cutoff points for these significant factors were determined using receiver operating characteristic (ROC) curve analysis". This is very confusing. Are the authors stating that they didn't use validated measures? Not clear, this needs more details.

8. Did the authors check to determine that after the mentioned exclusion criteria in Figure 1, the final study sample was still representative of the national population?

9. The results in Table 3 are confusing. Let's look at the smoking status for example. The authors are looking first the the simple logistic regression and then at the multivariable logistic regression using smoking as the exposure of interest and then colorectal cancer as the outcome of interest. Which criteria did the authors use to select the independent variables in the multivariable models? Was the directed acyclic graph (DAG) theory employed when selecting the minimum sufficient set of confounders for a minimally biased model? Same question applies for all the other exposure variables included in Table 3.

Author Response

Dear reviewer,

Thank you for your endeavor for developing the manuscript.

Please see the attached file that answer one by one of your precious comments.

Thank you very much.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for tackling all comments and concerns. 

Author Response

Dear Reviewer,

 

I appreciate for your precious comments for bettering my manuscript.

 

Thank you.

 

Reviewer 2 Report

Comments and Suggestions for Authors

The authors should not confuse longitudinal/prospective cohort studies with retrospective cohort studies. There is no such thing as a retrospective longitudinal cohort studies. There are 3 different types of cohort studies: longitudinal/prospective, retrospective, and mixed. Which one is this? Propensity score matching is a very controversial method used for analyses as it can result in biased findings. 

Author Response

Dear Reviewer,

Thank you for your precious advice.

Please see the attached file that I replied as you commented.

Thank you.

Author Response File: Author Response.pdf