We read with interest the commentary by Sertoglu et al in response to our article in which we detail the lack of correlation between 25-hydroxy vitamin D (25-OH D) levels and in vitro fertilization outcomes when embryonic aneuploidy and quality are normalized. Indeed, when using either categorical ranges for 25-OH D (<20, 20-29.9, and ≥30 ng/mL) or serum 25-OH D as a continuous variable in a receiver operating characteristic analysis, a meaningful difference was not found.

First, we agree that a study of 25-OH D’s effect on health and disease is challenging as it has been shown to be affected by many factors, including age, race, season of sampling, and body mass index. We further agree with the authors that these factors must be considered and accounted for when analyzing 25-OH D’s effects, which is why all of these factors were controlled for in the final multivariate logistic regression included in the article. Even after controlling for these factors, no difference emerged.

Second, as Sertoglu et al point out, the most appropriate analytic method for serum 25-OH D is controversial. To standardize our measurements, all samples were run in the same reference laboratory on a single analytic immunoassay system. The serum samples, while collected over a year, were batched and run together with the same assay. The intraassay and interassay variation coefficients were included in the article and were within standard published ranges. The use of an alternative measurement technique, such as liquid chromatography-tandem mass spectrometry, has the benefit of providing more precise measurements and allow for an analysis of bioavailable hormone, which may indeed strengthen findings. However, these methods are difficult to generalize to routine clinical use as their availability and expense limit widespread availability. Furthermore, when doing a comparative study, provided the same standardized analysis tool is used throughout the population, the conclusion based on the relative comparisons remain valid.

In conclusion, to the authors point, the time of sampling and/or seasonality of the measurements have already been taken into consideration for the readers benefit and are included in the original multivariate analysis. When considering what type of technology to utilize for 25-OH D measurements, a number of factors, including generalizability of results for clinical use, must be considered. Of utmost importance when doing a population comparative study is the uniform application of the chosen and validated measurement technique.