Constipation Symptoms in Multiple System Atrophy Using Rome Criteria and Their Impact on Personalized Medicine

Round 1

Reviewer 1 Report

Dear Sirs.

I read with attention the manuscript entitled "Diagnosis and distinguishing characteristics of constipation in multiple system atrophy using the Rome criteria" and I must confess that I found it rather confusing. The authors aim at trying to differentiate MSA from PD based on the symptoms of constipation that may (or may not) occur in either one of these diseases. However, only patients with MSA are included in the present manuscript and the data is compared with the findings of a previous paper from their own group, which severely limits the scientific validity of their conclusions. In the final paragraph, the authors gracefully acknowledge the limitations of the study. The problem is that these limitations (small sample size, absence of control group or PD group, short-term, and so on) severely impair the scientific quality of the study. It is believed that it would be better if the study would be re-done avoiding the acknowledged limitations and ressubmitted.

Author Response

Response to Reviewer 1:

Thanks for the comments from Reviewer 1. As reviewer 1 pointed out, this study is for MSA patients only and has limitations. However, although this is a small number of cases, this study is the first attempt to evaluate constipation in MSA with Rome III, and statistically significant results were obtained, and the significance of the results was considered. The title and abstract have been changed as follows in response to comments.

 

Title

Constipation symptoms in multiple system atrophy using Rome criteria and their impact for personalized medicine

 

Abstract

Constipation is one of the most common non-motor symptoms in multiple system atrophy (MSA); however, it has not been evaluated according to the standard diagnostic criteria for constipation in patients with MSA. We evaluated the characteristics of constipation in patients with MSA by using Rome criteria (Rome III), which has been validated and the widely used for gastrointestinal disorders.

Reviewer 2 Report

Misleading statistical reporting

All statistics, including frequencies in table 2 and correlations in table 3 must be reported with confidence intervals (CIs). Without confidence intervals, the authors' results are misleading. Please use Aghresti's method because of the small sample size. Specifically,

 

    * L.128-129:  The lower frequency of constipation in MSA patients compared to the previously reported could be due to the small sample size. Strictly speaking, the reported frequency of constipation,  56.9% has a 95% confidence interval of  [30.0%, 81.7%] (computed according to the Aghresti's method and using the sample size of 51). This interval includes 80%, so the authors should not claim a smaller reported frequency, especially that the previously reported number lacked the CIs and was not based on the Rome III diagnostic criteria.

 

   * L.136-137:  Comparison of two numbers with confidence intervals, 56.9% [30.0%, 81.7%]  and 27% [7.3%, 53.6%], shows the CIs overlap, and the conclusion is incorrect.

 

   * L.138-139:  The CIs of 72.5% [45.9%, 925%] and 81.4% [56.5, 97.1%] overlap completely, so the implication that one number is smaller is misleading.

 

The IBS was not discovered in this study (L.162-165), most likely because of the small sample size.

 

Unclear medical utility

There are no calculations of the likelihood of having MSA given certain constipation signs. So, it is not clear how this study may help distinguish between PD and MSA. If someone presents with symptoms that can be attributed to both PD and MSA, and the doctor tests them for constipation using Rome III, then what? I don't see how this helps discriminate between PD and MSA. I think the paper should not make such a claim.

 

Unclear research purpose

It seems that the paper is concerned with validating Rome III in MSA patients, so what is the gold standard for the constipation diagnosis?

Recommendations for Authors
Is the research design appropriate?
Re: The sample size is too small

Are the methods adequately described?
Re: It would be nice to understand the rationale behind the chosen statistical methods

Are the results clearly presented?
Re: All results lack confidence intervals making them misleading and impossible to interpret

Are the conclusions supported by the results?
Re: No, the conclusions are not supported by the results b/c the authors disregard uncertainty (they don’t report confidence intervals)

Author Response

Reviewer 2, comment:

All statistics, including frequencies in table 2 and correlations in table 3 must be reported with confidence intervals (CIs). Without confidence intervals, the authors' results are misleading. Please use Aghresti's method because of the small sample size. Specifically,

 

    * L.128-129:  The lower frequency of constipation in MSA patients compared to the previously reported could be due to the small sample size. Strictly speaking, the reported frequency of constipation, 56.9% has a 95% confidence interval of [30.0%, 81.7%] (computed according to the Aghresti's method and using the sample size of 51). This interval includes 80%, so the authors should not claim a smaller reported frequency, especially that the previously reported number lacked the CIs and was not based on the Rome III diagnostic criteria.

 

   * L.136-137:  Comparison of two numbers with confidence intervals, 56.9% [30.0%, 81.7%] and 27% [7.3%, 53.6%], shows the CIs overlap, and the conclusion is incorrect.

 

   * L.138-139:  The CIs of 72.5% [45.9%, 925%] and 81.4% [56.5, 97.1%] overlap completely, so the implication that one number is smaller is misleading.

 

The IBS was not discovered in this study (L.162-165), most likely because of the small sample size.

 

Response to Reviewer 2:

• We would like to thank Reviewer 2’s comments. The suggestion is important in scientific thinking in this manuscript. In the discussion, the expression regarding the comparison of the frequency of constipation in MSA was corrected as follows according to the principle of science.

 

Discussion

Our study tended to have a lower frequency of constipation using Rome criteria, compared to previous reports. It may be due to a more stringent diagnosis of constipation under Rome criteria [4].

 

• We have added the 95% CI to Table 3 as you suggested.

 

Table 3. Correlation analysis of the number of constipation symptoms and baseline data.

	r	P value	95% CI
Age (years)	0.035	.809	-0.252–0.321
Disease duration (years)	0.205	.149	-0.066–0.477
MMSE	-0.124	.387	-0.430–0.182
LED (mg/day)	0.370	.008	0.138–0.602
LED, levodopa equivalent dose; MMSE, Mini-Mental State Examination; CI, confidence interval.

 

Reviewer 2, comment:

Unclear medical utility

 

There are no calculations of the likelihood of having MSA given certain constipation signs. So, it is not clear how this study may help distinguish between PD and MSA. If someone presents with symptoms that can be attributed to both PD and MSA, and the doctor tests them for constipation using Rome III, then what? I don't see how this helps discriminate between PD and MSA. I think the paper should not make such a claim.

 

Response to Reviewer 2:

We would like to thank Reviewer 2’s comments. As you pointed out, the symptoms of constipation alone cannot distinguish between MSA and PD. MSA-P can be difficult to distinguish from PD, especially in the early stages of the disease. Therefore, we considered it from the perspective of constipation so that clinicians could distinguish them as much as possible. Certainly indistinguishable, but certain trends have become apparent. We added the frequency of MSA-P and MSA-C constipation and the need for accurate research to distinguish between MSA-P and PD in the future.

 

Results

The MSA-P group had more constipation than the MSA-C group (P < .05)

 

Table 5. Comparison of constipation symptoms among groups of MSA-C and MSA-P.

	MSA-C (n =22)	MSA-P (n =29)	P value
Constipation (%)	40.9	69.0	.045
Straining　(%)	50.0	72.4	.101
Lumpy or hard stools　(%)	36.4	55.2	.183
Sensation of incomplete evacuation　(%)	31.8	65.5	.017
Sensation of anorectal obstruction　(%)	50.0	82.8	.013
Manual maneuvers　(%)	9.1	17.2	.402
Fewer defecations　(%)	18.2	44.8	.046

 

Discussion

This study also showed that MSA-P had a higher frequency of constipation than MSA-C, which was consistent with previous studies [39]. A detailed analysis of gastrointestinal symptoms may help differentiate PD from MSA, and in the future studies using Roman IV in patients with larger scale PD and MSA are desirable.

 

Reviewer 2, comment:

Unclear research purpose

 

It seems that the paper is concerned with validating Rome III in MSA patients, so what is the gold standard for the constipation diagnosis?

 

Response to Reviewer 2:

We would like to thank Reviewer 2’s comments. We consider the Rome criteria to be the gold standard for diagnosing constipation, which has already been described in the Introduction as follows.

 

Introduction

Rome criteria are the most used standard criteria for the diagnosis of functional gastrointestinal disorders; Rome III diagnostic criteria were published in 2006 [5, 6].

 

Reviewer 2, comment:

Recommendations for Authors

Is the research design appropriate?

Re: The sample size is too small

 

Response to Reviewer 2:

As you pointed out, the sample size of this study is small. However, although this is a small number of cases, this study is the first attempt to evaluate constipation in MSA with Rome III, and statistically significant results were obtained, and the significance of the results was considered.

 

Reviewer 2, comment:

Are the methods adequately described?

Re: It would be nice to understand the rationale behind the chosen statistical methods

 

Response to Reviewer 2:

Thank you for the reviewer's comment. We add the methods that have not been changed below, but we believe that we have selected the appropriate analysis method based on these contents.

 

Methods

Student’s t-test (for quantitative variables) and Fisher’s exact test (for qualitative variables) were used to compare the demographic differences between the patients with constipation and patients without constipation and IBS. Multiple logistic regression analysis was used to estimate the odds ratio (OR) and 95% confidence interval (CI) to assess the association between the self-awareness of constipation and the symptoms of constipation. Spearman’s rank correlation coefficient was used to evaluate the association between the frequency of constipation symptoms and baseline data.

 

Reviewer 2, comment:

Are the results clearly presented?

Re: All results lack confidence intervals making them misleading and impossible to interpret

 

Are the conclusions supported by the results?

Re: No, the conclusions are not supported by the results b/c the authors disregard uncertainty (they don’t report confidence intervals)

 

Response to Reviewer 2:

We would like to thank Reviewer 2’s comments. We have added the 95% CI to Table 3 as you suggested. The discussion was based on this result.

 

Table 3. Correlation analysis of the number of constipation symptoms and baseline data.

	r	P value	95% CI
Age (years)	0.035	.809	-0.252–0.321
Disease duration (years)	0.205	.149	-0.066–0.477
MMSE	-0.124	.387	-0.430–0.182
LED (mg/day)	0.370	.008	0.138–0.602
LED, levodopa equivalent dose; MMSE, Mini-Mental State Examination; CI, confidence interval.

 

Reviewer 3 Report

Mishima T. et al. evaluated symptoms of constipation and IBS in 51 MSA patients according to the standard diagnostic criteria (Rome III). The authors claim that the differences of intestinal symptoms may help in the early differentiation of PD and MSA. The same research group published in 2017 a similar work on IBS-constipation symptoms in PD patients. The manuscript is clear and well written. A direct comparison between MSA patients, PD patients, and controls would have added value to the work. Indeed, the discussion on PD-MSA comparison is the most interesting and useful part of the manuscript. Several other limitations are acknowledged by the authors in the discussion.

I have no major remarks but few minor:

- The number of MSA-P and MSA-C patients should be clearly stated in the main text. 

- Is there any difference of intestinal symptoms between MSA-C and MSA-P?

- Please clarify the hypothesized mechanism by which dopaminergic treatment may increase the severity of constipation in MSA patients.

- Do MSA-P patients take more dopaminergic therapy? 

Author Response

Reviewer 3, comment:

Mishima T. et al. evaluated symptoms of constipation and IBS in 51 MSA patients according to the standard diagnostic criteria (Rome III). The authors claim that the differences of intestinal symptoms may help in the early differentiation of PD and MSA. The same research group published in 2017 a similar work on IBS-constipation symptoms in PD patients. The manuscript is clear and well written. A direct comparison between MSA patients, PD patients, and controls would have added value to the work. Indeed, the discussion on PD-MSA comparison is the most interesting and useful part of the manuscript. Several other limitations are acknowledged by the authors in the discussion.

 

I have no major remarks but few minor:

 

- The number of MSA-P and MSA-C patients should be clearly stated in the main text.

 

Response to Reviewer 3:

Thank you very much for your very important remarks. We appended the number of MSA-P and MSA-C as below.

 

Results

MSA-C and MSA-P were detected in 22 patients (43.1%) and 29 patients (56.9%), respectively.

 

Reviewer 3, comment:

- Is there any difference of intestinal symptoms between MSA-C and MSA-P?

 

Response to Reviewer 3:

We would like to thank Reviewer 3’s comments. We compared MSA-P and MSA-C as follows.

 

Results

Characteristics of the patients with MSA-C and MSA-P are shown in Table 4. Comparison of constipation symptoms among groups of MSA-C and MSA-P are listed in Table 5. The MSA-P group had more constipation than the MSA-C group (P < .05) (Table 5), and more LED and laxative use than the MSA-C group (P < .05) (Table 4). With regard to symptoms of constipation, MSA-P had the higher frequency of sensation of incomplete evacuation, sensation of anorectal obstruction, and fewer defecations than MSA-C (P < .05) (Table 5).

 

Table 4. Comparison of demographic and clinical characteristics among groups of MSA-C and MSA-P.

	MSA-C (n =22)	MSA-P (n =29)	P value
Age (years)	63.4 ± 10.8	67.0 ± 9.6	.505
Male (%)	36.3	48.3	.569
Disease duration　(years)	4.6 ± 3.3	4.3 ± 2.8	.466
MMSE	27.3 ± 2.7	26.4 ± 3.2	.956
LED (mg/day)	40.9 ± 140.3	377.1 ± 326.2	.032
Laxative use (%)	18.2	62.0	.002
Data are presented as the mean ± SD. MSA-C, multiple system atrophy, predominant cerebellar ataxia; MSA-P, multiple system atrophy, predominant parkinsonism; IBS, irritable bowel syndrome; LED, levodopa equivalent dose; MMSE, Mini-Mental State Examination.

 

Table 5. Comparison of constipation symptoms among groups of MSA-C and MSA-P.

	MSA-C (n =22)	MSA-P (n =29)	P value
Constipation (%)	40.9	69.0	.045
Straining　(%)	50.0	72.4	.101
Lumpy or hard stools　(%)	36.4	55.2	.183
Sensation of incomplete evacuation　(%)	31.8	65.5	.017
Sensation of anorectal obstruction　(%)	50.0	82.8	.013
Manual maneuvers　(%)	9.1	17.2	.402
Fewer defecations　(%)	18.2	44.8	.046

 

 

Reviewer 3, comment:

- Please clarify the hypothesized mechanism by which dopaminergic treatment may increase the severity of constipation in MSA patients.

 

Response to Reviewer 3:

We would like to thank Reviewer 3’s comments. Since it has been reported that dopaminergic treatment may inhibit the movement of the gastrointestinal tract (Li ZS, et al. J Neurosci. 2006 26, 2798-807), the following modifications have been made as below.

 

Discussion

Because dopaminergic treatment may increase the severity of constipation in patients with MSA due to inhibiting the movement of the intestinal tract [24], careful consideration of dosage is recommended.

 

Reviewer 3, comment:

- Do MSA-P patients take more dopaminergic therapy?

 

Response to Reviewer 3:

Thank you very much for your important remarks. It has been reported that it varies from patient to patient (Ishida C, et al. Intern Med. 2021 60, 367-72), and the text has been revised as follows

 

Although dopaminergic treatment is effective to some degree in some patients with MSA [23], its potency may be limited.

 

Round 2

Reviewer 1 Report

Dear Sirs.

As most of the issues I pointed in my first review were adequately addressed, it is believed that this paper is now ready for publication, with the new title.