Serum 25-Hydroxyvitamin D and Cancer Risk: A Systematic Review of Mendelian Randomization Studies

Round 1

Reviewer 1 Report

I am grateful for the opportunity to review this article. First of all, I would like to congratulate the authors for the enormity of work put into the preparation of this publication. The work presents an innovative and more precise approach to the long-explored topic of the role of vitamin D in the occurrence of cancer and related mortality. In my opinion, the only limitation of this study is the commonness of the topic of vitamin D and the overestimation of its therapeutic possibilities. Nevertheless, the use of an innovative method, which is Mendelian randomization, makes this article interesting and worth to publish.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The review by Lawler et al. systematically summarized the evidence on vitamin D and cancer incidence and mortality from Mendelian randomization (MR) studies. The manuscript is well-written. I have only a few suggestions to help improve the clarity of the paper.

1.     As the authors mentioned, a valid MR study is based on no violation of the three core assumptions of instrumental variable analysis. Although we cannot completely confirm the other assumptions except for the relevance assumption, a thorough evaluation of the three assumptions for each original MR study is preferred.

2.     It is not appropriate to present the range of effect size for each outcome (Line 18-20 and Section 3.2) because they are not comparable due to different scales of the exposure used, e.g., per SD (absolute or log-transformed), per unit (absolute or log-transformed), per 25 nmol/L. Instead, the number of papers with inverse/positive association vs. the total number of papers on each outcome is recommended.

3.     Both one-sample and two-sample MR studies were included in this systematic review. IVW method is commonly used in two-sample MR analyses using summary-level data. However, the Wald-type ratio or two-stage MR is usually adopted in one-sample MR analyses using individual-level data. Thus, the statement “only the primary IVW MR estimates for each manuscript are presented here” (Line 193-194) could be problematic.

 

4.     The number of papers in the flowchart (Figure 2) is incorrect. A total of 100 papers was excluded after reviewing of title and/or abstract. However, the sum of the number of papers excluded for each reason is not equal to 100 (i.e., 47+7+5+22=81).

Author Response

Please see the attachment.

Author Response File: Author Response.docx