Effect of COVID-19 vaccination on menstrual periods in a prospectively recruited cohort

As the UK COVID-19 vaccination programme was rolled out to younger participants, the MHRA’s surveillance scheme, Yellow Card, increasingly received reports from people who had noticed a change to their menstrual cycle following vac- cination: at 29 March 2022, 39,494 individuals had made such reports to Yellow Card [1]. It is important to note that most people who report such a change following vaccination find that their period returns to normal the following cycle [2] and extensive investigation has found no evidence that COVID-19 vaccination adversely impacts female fertility [3- 10]. Nonetheless, people are concerned by these reports. In- vestigating the potential link between COVID-19 vaccination and menstrual changes is important for maintaining public trust in the vaccination programme and, if a link is found, to allow people to plan for potential changes to their cycles [11]. In response to similar reports to the USA’s vaccine safety surveillance scheme, V AERS, the National Institutes of Health allocated $1.67m (£1.2m; C1.4m) for studies into a possible connection between COVID-19 vaccination and menstrual changes [12] and the first of these studies has re- cently reported [13]. The authors used prospectively logged data from a menstrual cycle tracking app: 3959 Americans logged at least six consecutive cycles; 2403 of them were vaccinated and 1556 acted as a control group. In individuals receiving only one dose of the vaccine per cycle, the first dose of vaccine had no effect on timing of the subsequent period, while the second dose was associated with a delay of 0.45 days. Individuals who received both doses of the vaccine in a single cycle experienced a 2.32 day delay to their next period. In all groups, cycle lengths returned to normal by two cycles after vaccination. A Norwegian study used mobile-phone questionnaires to ret- rospectively solicit reports of menstrual changes from 5688 women aged 18-30 years who had been recruited prospec- tively to the Norwegian Young Adult Cohort [14]. This study found high levels of variation, even in unvaccinated cycles, with 37.8% of participants reporting at least one difference from normal in their pre-vaccination cycles. However, the study was still able to identify heavier than normal bleeding as a change associated with vaccination, with a relative risk of 1.9 after the first dose of the vaccine and 1.8 after the sec- ond dose. Here, we describe the effects of COVID-19 vaccination on menstrual timing and flow in a cohort of 79 individuals who regularly experience either menstrual periods or withdrawal bleeding as a result of breaks in taking hormonal contracep- tion. These people were recruited before receiving either their first or second dose of the COVID-19 vaccine, and kept a daily record of their vaginal bleeding, and of any vaccine side effects they experienced.

Cohort. 253 people who were over 18, have regular periods or withdrawal bleeds and who were planning to receive ei- ther their first or their second dose of the COVID-19 vaccine were recruited by advertising on social media and in newslet- ters with a largely female readership in the UK. Of these, 43 withdrew for reasons including pregnancy, having entered the menopause, having received a diagnosis of a gynaecological condition that they felt was affecting their periods, or not hav- ing the time to journal every day (Figure 1). Of the remaining participants, 87 returned their journals and 79 of these jour- nals logged at least three consecutive cycles, including one in which a dose of the vaccine was given. Those who did not return their journals were contacted up to twice to find out why they had not completed their journals: 31 participants responded and 92 were lost to follow up. Journals recorded the participant’s age, reproductive his- tory, use of hormonal contraception during the study period, breastfeeding during the study period and whether they have ever been diagnosed with a menstrual or gynaecological con- dition (Table 1). Participants then completed a daily journal Woon et al. | bioRχiv | March 30, 2022 | 1–5 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 30, 2022. ; https://doi.org/10.1101/2022.03.30.22273165doi: medRxiv preprint NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice. Figure 1. Cohort flowchart. in which they reported their bleeding as “heavier than usual for this day of my cycle”, “normal”, “lighter than usual for this day of my cycle”, “spotting (normal for me)”, “spotting (not normal for me)” or “no bleeding”. For each cycle, par- ticipants also reported whether their period had come on the day they had expected it, and if not how many days early or late it was. Participants noted the day on which they received a dose of the vaccine, which brand they received, whether it was the first or second dose, and for seven days afterwards recorded whether they experienced any of the following: sore arm, fever, fatigue, headache, body aches. The study was approved by the Research Governance and Integrity Team at Imperial College London, study number 21IC6988. Statistical analysis. The timing of each cycle was desig- nated as “0” when the period or withdrawal bleed started on the expected day, with negative values indicating days early and positive values days late. For each cycle, a “flow score” was calculated by assigning “no bleeding” as 0, “spotting (normal for me)” and “spotting (not normal for me)” as 1, “lighter than usual” as 3, “normal” as 5 and “heavier than usual” as 7, and then totalling the responses for the first seven days of the cycle. Where participants logged more than one cycle in a particular category (for example, pre-vaccine, in- terdose or post-vaccine cycles) the mean of the timing and flow scores was taken. For each dose of the vaccine, a “side effect score” was cal- culated by the total number of days that the participant re- sponded “yes” to each of the side effect questions. Changes in cycling timing and flow score were assessed us- ing a mixed-effects model, with the Geisser-Greenhouse cor- Table 1. Participant characteristics. rection for sphericity, and Tukey’s test for multiple com- parisons. Where significant changes were found, sensitivity analyses were carried out by assuming that every participant who had not returned their journal had experienced scores for every cycle with the same distribution as the pre-vaccine cy- cle scores recorded by those who did return their journals: the distribution was produced by randomly selecting a score from the pre-vaccine cycle data distribution for each missing value. Associations between side effect, timing and flow scores were assessed using Spearman’s correlation with the Holm- Bonferroni correction for multiple testing. Analysis was carried out and graphs generated using Graph- pad Prism 9.0.0.

COVID-19 vaccination is associated with a delay to the next period in spontaneously cycling individuals. Considering only those who were spontaneously cycling, we found that the period after both the first and the second dose of the vaccine came significantly later than expected: com- pared to the day on which the participant expected their period, pre-vaccine periods occurred on average 0.17 days 2 | bioRχiv Woon et al. | COVID-19 vaccination and menstruation . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 30, 2022. ; https://doi.org/10.1101/2022.03.30.22273165doi: medRxiv preprint Figure 2. Violin plots showing the distribution by which periods or withdrawal bleeds began in pre-vaccine cycles, the cycle fol- lowing dose 1 of the COVID-19 vaccine, interdose cycles, the cycle following dose 2 of the COVID-19 vaccine, and subse- quent cycles, where “0” denotes the period or withdrawal bleed beginning on the expected day, negative numbers denote days early and positive numbers days late. Data for spontaneously cycling (A) and participants on hormonal contraception (B) are shown. * p’ < 0.05, ** P < 0.01. Figure 3. Violin plots showing the distribution of flow scores for periods or withdrawal bleeds in pre-vaccine cycles, the cycle fol- lowing dose 1 of the COVID-19 vaccine, interdose cycles, the cycle following dose 2 of the COVID-19 vaccine, and subse- quent cycles. Data for sponta- neously cycling (A) and partic- ipants on hormonal contracep- tion (B) are shown. early, whereas the period following dose one occurred a mean of 2.3 days late (p’ = 0.0045) and the period following dose two occurred a mean of 1.3 days late (p’ = 0.041) (Figure 2a). Periods in interdose cycles and post-vaccination cycles occurred a mean of 0.3 days late and 0.47 days early, respec- tively, values which were not significantly different from the pre-vaccination average. Therefore, in spontaneously cycling individuals, vaccination was associated with a delay to the subsequent period, but this quickly reversed. Although this finding was in line with that from a large and well-designed study carried out in the USA which also de- tected delays to the period following a dose of the vaccine [13], the delays we found were somewhat larger. We there- fore considered the possibility that those participants who had returned their journals were more likely to have no- ticed any change than those who did not, inflating the post- vaccination changes. To address this possibility, we carried out two sensitivity analyses. We generated distributions con- sistent with no change occurring and used them to impute the data that we would have expected to see had the participants who responded that they had stopped logging their periods because they had not noticed a change returned their jour- nals: the differences between pre-vaccination cycle timing and post-dose 1 and -dose 2 timing remained significant at p’ = 0.0013 and p’ = 0.011, respectively. Even assuming that none of the participants who failed to return their journals had noticed any change, significant delays to the post-dose 1 period (p’ = 0.0035) and the post-dose 2 period (p’ = 0.045) remained. Considering only those participants who were taking hor- monal contraception, and considering all types of hormonal contraception together due to the small numbers of these par- ticipants, we found no effect of vaccination on the timing of the subsequent withdrawal bleed (Figure 2b). COVID-19 vaccination is not associated with any change to menstrual flow.We found no significant change to self-reported menstrual flow in the period or withdrawal bleed following vaccination, either in spontaneously cycling participants (Figure 3a), or in those taking hormonal contra- ception (Figure 3b). Commonly-reported side effects are not associated with menstrual changes. One hypothesis that has been put forward to explain the existence of menstrual changes follow- ing vaccination is that immune activation may transiently in- terfere with the hypothalamic-pituitary-adrenal axis [15,16]. or with immune cells in the lining of the uterus, which con- trol the breakdown and regeneration of this tissue [17]. In ei- ther of these cases, we might consider that menstrual changes are not dissimilar to other short-term and reversible side ef- fects of vaccination. Therefore, we might expect menstrual changes to be more likely to occur in those who experienced common side effects. To explore this hypothesis, we examined correlations be- tween self-reported side effect score and either the timing or the flow score of the subsequent period: no significant corre- lations were found (Table 2). Therefore, in this small cohort we were unable to find evidence that menstrual changes cor- relate with other common side effects of vaccination. Woon et al. | COVID-19 vaccination and menstruation bioRχiv | 3 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 30, 2022. ; https://doi.org/10.1101/2022.03.30.22273165doi: medRxiv preprint Table 2. Spearman’s r, p and p’ values are shown for the comparisons indicated.

Here we report that, in a prospectively-recruited cohort, post- vaccination periods occurred later than expected after both the first and second dose of the vaccine, and this finding was robust to a sensitivity analysis designed to account for the possibility that participants who noticed a change may be more likely to return their journals. Importantly, periods re- turned to coming at the expected time in the interdose and post-vaccine cycles. These findings are in line with those from a recent study in the USA, which reported increases in cycle length associated with receiving either the second dose of vaccine, or both doses in the same cycle, but again noted that cycles rapidly returned to normal [13]. In this cohort, we were unable to identify any change in self- reported flow associated with COVID-19 vaccination, and this is in contrast to findings from a Norwegian study, which identified an increase in reports of heavier than normal bleed- ing after either dose of the vaccine [14]. This cohort was smaller than the Norwegian one, so it is possible that we were not powered to detect a real change in menstrual flow. On the other hand, our participants recorded data in real time, whereas participants in the Norwegian cohort were asked to recall their experiences, so it is also possible that the differ- ence reported in the Norwegian cohort is at least partially impacted by recall bias. Because our participants were prospectively recruited, this study should be less affected by recruitment bias than survey- based studies [18-20] and because the participants were asked to record their experiences in real-time the data is unlikely to be affected by recall bias. However, there are also a number of important weaknesses. Because the cohort is small, we may not have been able to detect real effects of vaccination on menstrual flow, or real relationships between menstrual and non-menstrual side effects. Some bias is also likely to have been introduced because it is unlikely that the partici- pants who did not return their sheets did so at random, al- though we did attempt to account for this with a sensitivity analysis. Some bias is also likely to be introduced because “expected date of period” and “menstrual flow” are at least partially subjective, so there is the potential for participant’s expectations of how vaccination will impact their periods to impact their reports. In conclusion, in this small, prospectively-recruited cohort, we found that both the first and second doses of the COVID- 19 vaccine were associated with a small delay to the sub- sequent period. This delay was not reported in interdose and post-vaccination cycles, suggesting that any changes are temporary. We were unable to detect changes to men- strual flow. These findings are reassuring, but work in larger prospectively-recruited cohorts is needed to more accurately determine how vaccination affects menstrual parameters. Data availability Cleaned, fully anonymised data, together with our analy- sis files, are available from the Open Science Framework at https://osf.io/upbyg/.

This study did not receive any specific funding. The investigators were funded by the preterm birth research charity Borne.

1. Medicine and Healthcare Products Regula- tory Agency. Coronavirus vaccine—weekly sum- mary of Yellow Card reporting. 25 March 2022. 2021.https://www.gov.uk/government/publications/coronavirus- covid-19-vaccine-adverse-reactions/coronavirus-vaccine- summary-of-yellow-card-reporting 2. Medicines and Health- care Products Regulatory Agency. COVID-19 vaccines: updates for August 2021. https://www.gov.uk/drug-safety- update/covid-19-vaccines-updates-for-august-2021 3. Male V . Are covid-19 vaccines safe in pregnancy? Nat Rev Immunol 2021;21:200-1. doi:10.1038/s41577-021-00525-y 4. Hillson K, Clemens SC, Madhi SA, V oysey M, Pollard AJ, Minassian AM, Oxford COVID Vaccine Trial Group. Fertility rates and birth outcomes after ChAdOx1 nCoV- 19 (AZD1222) vaccination. Lancet 2021;398:1683-4. doi: 10.1016/S0140-6736(21)02282-0. 5. Morris RS. SARS-CoV-2 spike protein seropositivity from vaccination or infection does not cause sterility. F S Rep 2021; 2:253-255. doi: doi:10.1016/j.xfre.2021.05.010. 6. Orvieto R, Noach-Hirsh M, Segev-Zahav A, Haas J, Nahum R, Aizer A. Does mRNA SARS-CoV-2 vaccine in- fluence patients’ performance during IVF-ET cycle?Reprod Biol Endocrinol2021;19:69. doi:10.1186/s12958-021- 00757-6 7. Bentov Y , Beharier O, Moav-Zafrir A, et al. Ovar- ian follicular function is not altered by SARS-Cov- 2 infection or BNT162b2 mRNA Covid-19 vaccina- tion. medRxiv 2021:2021.04.09.21255195. [Preprint.] doi:10.1101/2021.04.09.21255195 8. Safrai M, Rottenstreich A, Herzberg S, Imbar T, Reubinoff B, Ben-Meir A. Stopping the misinformation: BNT162b2 COVID-19 vaccine has no negative effect on women’s fertility. medRxiv 2021:2021.05.30.21258079 [Preprint.] doi:10.1101/2021.05.30.21258079 9. Aharon C; Lederman M; Ghofranian A et al. In Vitro Fertilization and Early Preg- nancy Outcomes After Coronavirus Disease 2019 (COVID- 19) Vaccination, Obstetrics Gynecology 2022 doi: 10 Wesselink AK, Hatch EE, Rothman KJ et al. A Prospec- tive Cohort Study of COVID-19 Vaccination, SARS-CoV- 4 | bioRχiv Woon et al. | COVID-19 vaccination and menstruation . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 30, 2022. ; https://doi.org/10.1101/2022.03.30.22273165doi: medRxiv preprint 2 Infection, and Fertility. Am J Epidemiol 2022;, doi: 10.1097/AOG.0000000000004713 11. Male V . Menstrual changes after covid-19 vaccination. BMJ 2021;374:n2211 doi: 10.1136/bmj.n2211 12. Eunice Kennedy Shriver National Institute of Child Health and Human Development. Item of interest: NIH funds studies to assess potential ef- fects of COVID-19 vaccination on menstruation. 2021.https://www.nichd.nih.gov/newsroom/news/083021- COVID-19-vaccination-menstruation 13. Edelman A, Boniface ER, Benhar E, et al. Association between menstrual cycle length and coronavirus disease 2019 (covid-19) vaccination: a US cohort. Obstet Gynecol 2022. doi:10.1097/AOG.0000000000004695 14. Tragostad L. Increased occurrence of menstrual distur- bances in 18- to 30-year-old women after covid-19 vaccina- tion. [Preprint.] 2022. doi:10.2139/ssrn.3998180 15. Turnbull A V , Rivier CL. Regulation of the hypothalamic- pituitary-adrenal axis by cytokines: actions and mechanisms of action. Physiol Rev 1999;79:1–71. doi: 10.1152/phys- rev.1999.79.1.1 16. Karagiannis A, Harsoulis F. Gonadal dysfunction in systemic diseases. Eur J Endocrinol 2005; 152:501-13. doi:10.1530/eje.1.01886 pmid:15817904 17. Monin L, Whettlock EM, Male V . Immune responses in the human female reproductive tract. Immunology 2020;160:106-15. doi:10.1111/imm.13136 18. Alvergne A, Kountourides G, Argentieri MA, et al. COVID-19 vaccination and menstrual cycle changes: A United Kingdom (UK) retrospective case-control study. MedRXiv 2021: 2021.11.23.21266709 [Preprint.] 19. Male V . Effect of COVID-19 vaccination on menstrual periods in a retrospectively recruited cohort. MedRXiv 2021: 2021.11.15.21266317 [Preprint] doi: 10.1101/2021.11.15.21266317 20. Lee KLM, Junkins EJ, Luo C, Fatima UA, Cox ML, Clancy KBH. Investigating trends in those who experi- ence menstrual bleeding changes after SARS-CoV-2 vacci- nation. MedRXiv 2021 2021.10.11.21264863 [Preprint] doi: 10.1101/2021.10.11.21264863 Woon et al. | COVID-19 vaccination and menstruation bioRχiv | 5 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 30, 2022. ; https://doi.org/10.1101/2022.03.30.22273165doi: medRxiv preprint