Reply to "Double-counting of effect sizes and inappropriate exclusion of studies in "The influence of vitamin D supplementation on IGF-1 levels in humans: A systematic review and meta_analysis"

Abstract

Dear Editor, We appreciate the comments provided to us in the letter from Amiri et al. We would like to offer some clarifications in response to the comments.

Firstly, the authors claimed that we included the control group more than once from the same publication including different arms. The authors also identified mistakes made during data screening and expressed concerns that the mistakes would lead to errors within our analysis (Kord-Varkaneh et al., 2020). We sincerely thank the authors for pointing out these mistakes. In response, based on The Cochrane Handbook for Systematic Reviews, we have now reanalyzed the results by combining the intervention groups to create a single pair-wise comparison and updated our meta-analysis as displayed below. We initially excluded Sorva, Antti et al. from our study because based on our inclusion criteria we only included studies that had more than 10 participants in each group (Sorva et al., 1994), although we did not originally include this criterion in our publication (Kord-Varkaneh et al., 2020). However, after including Norenstedt, S. et al. and Sinha-Hikim, I. et al. (Sinha-Hikim et al., 2015; Norenstedt et al., 2013) our overall effect size changed from (WMD: 4 ng/ml, 95 % CI: -4 to 11, p = 0.35) to (WMD: 1.01 ng/ml, 95 % CI: --6.78 to 8.81, p = 0.799) without a change in overall significance (Fig. 1). In addtion, subgroup analyses showed that vitamin D dosage ≤1000 IU/day (WMD: 11.36 ng/ml, 95 % CI: 1.57–21.16, I2 = 79 %) significantly increased IGF-1 than vitamin D dosage <1000 IU/day (WMD: -2.8 ng/ml, 95 % CI: -7.02 to 1.6, I2 = 51 %). Moreover, an intervention duration of ≤12 weeks (WMD: 9.92 ng/ml, 95 % CI: 3–16.83, I2 = 77 %) significantly increased IGF-1 compared to a duration of <12 weeks (WMD: -5.43 ng/ml, 95 % CI: -10.25 to -0.61, I2 = 0.0 %). In addition, baseline serum vitamin D (<20 ng/mL vas ≥20 ng/mL) we did not observe significantly.