Response to: There is not enough evidence to support the claim that Covid‐19 vaccination BNT162b2 temporarily impairs semen concentration and total motile count

1.

Dear Editor,

We would like to thank you for the option to correspond to Rhodes et al. ¹ For the best of our knowledge, our study ² is the longest published post‐covid‐19 vaccination follow‐up. Three different statistical analyses resulted with identical outcome ‐ selective sperm concentration and total motile count reduction 3 months post‐vaccination.

Generalized estimated equation (GEE) model was used for repeated measures analysis. The working correlation structure was first‐order autoregressive relationship. The robust covariance matrix estimator was used to calculate standard errors. As accepted for several decades ³ and mentioned by Rhodes et al., corrections for multiplicity are not mandatory and therefore were not performed. To confirm GEE results, our approach was to use additional different analyses including median comparisons between T1–T3 to T0 for first and means sample per donor. Analysis of first sample only does not necessary align with donor's mean in each time frame. For example, in case of impaired semen sample, a possible action by sperm bank laboratory would be to postpone further samples from that donor for several weeks. Such hypothetical bias may have resulted with different outcome between first versus mean comparison. Consequently, these two analyses manifest different approaches with similar findings which further support GEE outcome.

We decided to define specific time frames (T1 = 15–45 days after vaccine; T2 = 75–125 days; T3 = 145+). Most current literature is focused short term (correlating with T1 for comparison). However, we were most interested in three‐ and six‐months post‐vaccination time frames. T2 is based on physiological duration of spermatogenesis to evaluate the impact of the first days‐weeks after vaccination completion, while T3 was expected to express possible continuous vaccination impact. This time frames approach is widely accepted.

Since the study included anonymous sperm donors, participants’ anonymity is highly crucial. According to our protocols, raw data, which include ages and dates of vaccination doses, can be shared only under very strict and limited conditions. However, processed data were shared with other researchers upon their request to perform meta‐analysis on that topic, and we wait for their results.

Medical and scientific obligations require unbiased examination of vaccinations’ impact using scientific – statistical tools as any other study and to describe the outcome regardless of social impact. Social media and scientific reports are separate systems, and our responsibility is to maintain scientific paradigm and principles. The current study included diverse statistical approaches, which all resulted with temporary sperm concentration and total motile count (TMC) decline, which were objectively reported. Additionaly, these analysis contribute reassuring data regarding covid‐19 vaccinations’ safety over time. As mentioned in the limitation section, the retrospective design and inclusion of sperm donors necessitate further reserach. Using smoothing splines model and statistical analyses suggested by Rhodes et al. are as acceptable as the methodologies used in the current research, and may be used in future studies.