During the COVID-19 pandemic, millions of women worldwide were vaccinated against the SARS-CoV-2 virus. Not long after, women administered with the first and second doses reported changes in their menstrual cycle1,4,5. The influence of vaccines on women’s menstrual bleedings was previously established, as in the vaccines against Hepatitis B and Papilloma viruses2,9,13,14 where activation of the immune system, inflammation, and disturbed regulation of the uterus regeneration were suggested as the cause4,9,15.

Studies that followed ovarian function in vaccinated women undergoing IVF treatments indicate that the vaccine doesn’t affect ovarian function or oocyte quality (oocyte maturation or fertilization rates following mRNA-based vaccines)32, even when anti-COVID-19 IgG antibodies were detected in the follicular fluids (FF)33. Moreover, ovarian reserve post vaccination, assessed by serum AMH levels34,35 as well as antral follicle count and hormonal serum leves36, were not changed. None the less, irregular periods are important well beyond fertility; they cause discomfort and emotional stress, which are often being ignored. In addition, they are associated with the risk of cardiovascular morbidity, chronic diseases, and premature mortality. The current study shows that changes in the menstrual cycle occur after the 3rd dose (~6 months after the second dose) and not only after the first and second vaccinations, as in most of the reports1,10,11,12.

Regulated by the HPO axis23, the menstrual cycle is comprised of three phases: follicular, ovulatory, and luteal. The follicular phase - from the beginning of the menstrual bleeding to ovulation - governed by FSH promoting follicular growth, and the luteal phase - from ovulation to the next menstrual bleeding - regulated by a surge of LH promoting corpus luteum formation. The normal length of ovulatory cycles (between 21 and 35 days), is usually derived from varying lengths of the follicular phase26. In contrast, the luteal phase is relatively constant in its duration (~14 days)26.

GCs, the somatic endocrine cells of the follicle, participate in the strict HPO feedback loop24,25 in the following manner: FSH stimulates GCs via FSH receptor (FSHR) to secrete endocrine and paracrine regulators, as estrogen (produced by Aromatase within the GCs), AMH and Inhibins. In turn, these hormones regulate FSH, by either directly reducing its synthesis and secretion from the hypophysis (estrogen and inhibins24,25) or indirectly reducing follicle’s sensitivity to FSH (AMH37,38). As GCs play a role in the regulation of the HPO axis, and following the Acutias “Final report”, indicating the LNP vehicle accumulates in ovaries, we aimed to explore if the vaccine can affect directly the expression of ovarian regulators, that in turn may explain post-vaccination changes in menstrual cycles.

To assess the direct effect of the vaccine on the ovary, we pooled hpGCs isolated from several women, thereby mitigating biases due to individual physiological differences and treatment protocols. We examined two vaccine concentrations to determine their effects: (i) 0.05 μg/ml (‘injected dose’) to evaluate cell toxicity and (ii) 0.05 pg/ml (‘end-organ dose’), representing the accumulated concentration in women’s ovaries, based on the Final Report to the FDA20.

We first aimed to find whether the vaccine is toxic to the cells, using a 1000-times higher dose than in the end-organ. We found that the hpGCs’ viability was not compromised after 24 or 48 h of exposure to either the injected or the end-organ doses. Thus, we concluded that the findings gathered in this work were not a result of impairment to the cells vitality but rather related to changes in their activity.

To support our hypothesis, we examined the mRNA level of GCs’ activity-related genes. A 24-h stimulation with the vaccine injected dose caused a decrease in the mRNA levels of aromatase and FSHR, reflecting their reduced activity and receptivity to FSH. Together with the robust increase in the mRNA level of the pro-inflammatory chemokine IL-8, which acts to resolve the inflammatory stimulus and promote healing39, these outcomes most likely represent an acute stress phase experienced by the cells. Such cellular response was expected following a direct exposure to the injected dose of the vaccine. Interestingly, the end-organ dose did not cause a notable change in aromatase, FSHR or IL-8 expression but led to an increase in InhibinB mRNA level that did not reach a statistical significance and was not yet reflected in the secreted protein level.

We continued to follow the hpGCs after a 48-h exposure to the vaccine, as was measured in the “Final report”. At this time point, the stress-related changes detected after the 24-h exposure to the injected dose reverted to their baseline state, pointing to a recovery of the cells’ activity and responsiveness. However, exposure to the end-organ dose resulted in a prominent elevation in InhibinB mRNA level. The level of the secreted protein displayed a similar trend, further supporting our results. Even though secretion of InhibinB into the culture medium was elevated compared to control in 3 independent experiments, it was not statistically significant. This may be due to the differences in InhibinB concentration (as measured by ELISA) of the 3 independent controls, caused by different attachment and survival rates of the hpGCs during the culture period. These differences are reflected in the high STDEV of the control group, which in turn affected the P value. Additionally, we might have detected a more prominent elevation in the secreted protein had we allowed a longer culture period, as proteins are expressed after mRNAs. Nonetheless, these results support the observed elevated-mRNA levels and imply a rise of InhibinB levels in post-vaccinated women’s sera. The latter may cause a disruption of the hormonal axis rooted at the base of the reported changes in menstrual cycle, implying an altered InhibinB level ~1 month after vaccination.

We further followed other participants in the HPO axis.

InhibinA, another member of the Inhibin family, is involved in regulating FSH and LH secretion. As it is expressed mainly by the corpus luteum22 we were not surprised to find no effect of the vaccine on its expression in hpGCs.

AMH, produced exclusively by GCs, mainly of small growing follicles up to small antral follicles37,38,40, contributes to the menstrual cycle by affecting other follicles in a paracrine manner. It downregulates the FSHR level in pre-antral follicles, and inhibit the activation of primordial follicles from the ovarian pool29. Even though we used hpGCs that were isolated from preovulatory follicles, we did detect AMH expression. This might be due to the fact that in IVF, oocytes are isolated from small antral follicles as well, or due to a low basal expression of AMH by large antral follicles41. We observed a significant reduction in AMH transcripts in both vaccine doses, 48 h after administration in-vitro. It implies to a similar reduction in AMH expression in smaller follicles in the ovary following vaccination, that may result in a larger population of hormonally-active follicles, that can cause an even higher serum level of InhibinB and disruption to the cycle. However, as showed, it does not represent a clinically diminished primordial reservoir42, but a local and transient effect.

Both InhibinB and AMH are secreted from the GCs of growing follicles. Since most of the growing follicles undergo atresia (follicle apoptosis) and only a few continue to develop toward ovulation, atresia of the vaccine-affected follicles could clarify why the reported menstrual changes are transient. The varying responses to the vaccine among different individuals, might also be influenced by the menstrual cycle phase (follicular or luteal) at the time of vaccination.

The levels of FSH and InhibinB change along the menstrual cycle, but their ratio remains relatively constant, independent of the day of cycle43; this ratio is expected to be similar between consecutive menstrual cycles of the same woman. In line with our in-vitro experiments, our in-vivo analysis of the FSH/InhibinB ratio in women before and ~1 month after the 3rd dose of COVID-19 vaccine showed that the post vaccination FSH/InhibinB ratio was changes by 2–3 folds. This change reinforces out hypothesis that vaccination caused an immediate elevation in InhibinB expression, that led to changes in the menstrual cycles’ length and bleeding, as well as to an altered FSH/InhibinB ratio a month later.

As the anti-COVID-19 vaccine is the first commercially available mRNA-based vaccine, and since there is no available vehicle to serve as “control”, we cannot discard the possibility that the changes we characterized in the hpGCs were induced by the vaccine envelope and not specifically by COVID-19 mRNA sequence. Today, when there are more mRNA-based vaccines in the pipeline44 this issue is highly relevant.

To summarize, this study reveals a unique, independent mechanism for vaccine related menstrual changes, concomitant with the vaccine-inflicted immune response. Our work suggests that at exposure, the COVID-19 vaccine can affect GCs directly, though not by reducing their viability. Exposure to the end-organ concentration of the vaccine exerted changes in the transcripts of two ovarian-regulatory key factors: a prominent upregulation of InhibinB and a downregulation of AMH. These changes can strongly affect FSH serum levels in vaccinated women; lead to disrupted follicular growth (i.e., too many follicles growth at the “wrong” time of the cycle) and activity (i.e., estrogen production); and ultimately affect the uterus cyclicity that is clinically displayed by changes in the menstrual bleeding pattern. Serum analysis of vaccinated women who reported menstrual changes, showed a transformed FSH/InhibinB level, supporting our results.