Disturbances of menstrual cycle after immunization against SARS‐CoV‐2 and their risk factors: Cross‐sectional clinical study

Abnormalities of the menstrual cycle were reported after infection with SARS‐CoV‐2 and vaccination against it, but the available data are very heterogeneous, do not reflect intermenstrual variations or regional differences, and their risk factors are missing.


| INTRODUC TI ON
Since the beginning of the global COVID-19 pandemic, there have been emerging reports of menstrual cycle irregularities in association with COVID-19 infections as well as with the vaccination against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2, coronavirus). 1Previous studies on the possible effects of COVID-19 infection or vaccination on the menstrual cycle analyzed changes in menstrual parameters based on data from digital cycle applications without including demographic data and without consideration of a third vaccination or heterogenous immunization. 2,3Studies evaluating risk factors for menstrual cycle changes are lacking.
Knowledge of the risk factors for the changes mentioned could be helpful in raising trust in vaccines.Additionally, identifying the people receiving vaccination at risk and obtaining information on the association between subjective and objective perceptions of the changes could help in developing individualized strategies for patients.
We focused on perceived changes in the three menstrual cycle parameters-cycle length (CL), menses duration (MD), and bleeding volume (BV)-following vaccination against SARS-CoV-2 or infection with coronavirus, also identifying additional risk factors for cycle disturbance in the first study evaluating data from Germany.

| MATERIAL S AND ME THODS
The population of this cross-sectional study comprised female/diverse employees and students at Ulm University Hospital, as they represent a cohort of women of reproductive age with high vaccination and/or infection rates.A total of 6383 people received an invitation to complete an online questionnaire.Inclusion criteria were age over 18 years and menstruating women who were vaccinated at least once against COVID-19 and/or after a proven infection with the SARS-CoV-2 by nucleic acid testing.

| Data collection
The study took part between March 1 and 31, 2021.The study population was invited twice per email to answer an online questionnaire anonymously after confirming informed consent.To avoid selection bias, as those affected are more likely to report, the study population was motivated to participate by the possibility of winning one of three €50 vouchers in a separate lottery.

| Study design
The specifically designed online questionnaire included 47 ques- In addition, participants were asked about characteristics of their menstrual cycle in terms of CL, MD, and BV before and after the vaccination/infection.To account for physiological intermenstrual variation, these parameters were collected as categorical variables with three categories each.These categories were: under 25 days (short), 25-35 days (normal), over 35 days (long) for CL; under 3 days (short), 3-6 days (normal), over 6 days (long) for MD; and less than four tampons or menstrual pads/day (weak), four to six tampons or menstrual pads/day (normal), more than six tampons or menstrual pads/day (strong) for BV.
Participants were also asked about subjectively perceived changes regarding the menstrual cycle.The first question was if there was a general change (yes/no) after immunization and/or infection.If the answer was yes, the person was then asked how long the menstrual cycle was affected (one cycle, <3 months, 3-6 months, >6 months) and when those effects occurred (after first, second, or third vaccination, or infection).

| Statistical analysis
Descriptive statistics of categorical data are provided in terms of absolute and relative frequencies.Age and BMI as continuous variables are presented as median and range.The marginal homogeneity test was used to compare the changes in the CL, MD, and BV categories before and after immunization/infection.Cohen's kappa coefficient was calculated to assess the agreement between changes in the CL, MD, and BV categories and subjectively perceived changes (yes/ no); the level of agreement was categorized according to Landis and Koch. 4 Binary logistic regressions were performed to evaluate factors associated with the changes (yes vs no) in the three categories of the menstrual cycles.Analyzed factors were age (years), BMI, working night shifts (yes vs no), thyroid dysfunction (yes vs no), PCOS (yes vs no), endometriosis (yes vs no), count of pregnancies (at least one vs none), known previous cycle irregularities (yes vs no), age at menarche (years), infection status (yes vs no), vaccination status (yes vs no), and combination of vaccines (heterogenic vs homogenous).
Odds ratios with 95% confidence intervals were calculated for univariable and multivariable binary logistic regressions (including all factors listed earlier).Statistical analysis was performed with IBM SPSS Statistics software package version 28 (IBM Corp., Armonk, NY, USA) and P-values less than 0.05 were considered statistically significant.

| Ethical approval and registration
The study was approved by the ethical committee of Ulm University (460/21) and registered in the German clinical trials register (DRKS00028135).

| RE SULTS
A total of 1788 (28.0%) people completed the survey.Overall, 62 people were excluded (24 postmenopausal women and 38 women with negative or unknown infection and vaccination status) for further analysis-thus, the final dataset for analysis comprised 1726 people.The characteristics of the study population are provided in Table 1 and their immunization status is given in Table 2.
TA B L E 1 Baseline characteristics of patients.

| Changes in CL, MD, and BV
There was a significant change with regard to CL from before to after immunization/infection (P < 0.001), with 21.1% of participants reporting changes in the CL categories (12.4% reported an increase in CL, 8.6% reported a decrease in CL; see Table 3).Likewise, there was a significant change in BV (P < 0.001), with 17.5% and 4.4% of participants reporting an increase or a decrease, respectively (Table 3).Overall, 13.6% of participants reported changes in the MD categories (7.4% increase, 6.2% decrease); however, the change from before to after immunization/infection regarding categories was not significant (P = 0.123; Table 3).
The distributions of the changes in CL, MD, and BV after immunization are shown in Figure 1.

| Subjectively perceived changes
In the subjective evaluation, 483 respondents (28.0%) stated there had been a disruption in their menstrual cycle after immuniza- menstrual cycle in 165 (9.6%) subjects up to more than 6 months in 103 (6.0%) (see Table 4).

| Agreement between reported changes in menstrual cycle categories and subjectively perceived overall change
Cohen's kappa was significant for changes in CL categories (0.111, P < 0.001), MD categories (0.058, P < 0.001), and BV (0.047, P < 0.001), but with only slight levels of agreement according to Landis and Koch. 4 No significant associations were found for BMI, working night shifts, endometriosis, age at menarche, infection status, or combination of vaccines.

| Risk factors for change in CL
In the multivariable analysis, only the previous cycle irregularities factor was significantly associated with a CL change (OR 2.35, 95% CI 1.77-3.10,P < 0.001).

| Risk factors for change in bleeding volume
Regarding the change in BV, univariable binary logistic regression analysis revealed significant associations with age (OR 1.03, 95% CI

| DISCUSS ION
To our knowledge, the present study is the first to identify risk factors for perceived changes in menstrual cycle characteristics after immunization against SARS-CoV-2 (vaccination and/or infection) and the first one to describe a cohort from a German population.

F I G U R E 1
Changes in attributes of menstrual cycle after immunization (vaccination or infection) against SARS-CoV-2.

| Key findings
Our data showed that more than one-quarter of women who were vaccinated or infected experienced changes in the menstrual cycle, which were temporary but could last for up to more than 6 months and occurred especially after the second vaccination.The most affected parameter was CL followed by BV.Changes in MD were not statistically significant.
Participants' reports about alterations in the menstrual cycle during the Covid pandemic were the subject of previous reports.menstrual bleeding volume after the first and second vaccinations. 7,8Based on data from a cycle app, Edelman et al. showed that changes in CL were less than 1 day but were more common after the second dose. 2 Phelan et al. concluded that there were no changes in bleeding volume but a wider variability in the CL. 9 Data for the third vaccination and combination of vaccination and infection are lacking.In general, despite a heterogeneous methodology, all studies observed at least signs of irregularities in some characteristics of the menstrual cycle.
The menstrual cycle, succumbing to external and internal factors, is divided into the follicular and luteal phases, the former being considered as variable in length and the latter as stable. 10,11Recent studies have suggested that cycle irregularities after immunization, such as shortened or prolonged menstrual cycle duration, depend on the phase of the cycle in which the vaccine was administered. 12The inflammatory response to the vaccine could interact with the hypothalamus-pituitaryovarian regulating cascade, leading to delayed follicular recruitment and resulting in a prolonged cycle or disturbance of endometrial stability in the luteal phase, leading to preterm initiation of menstrual processes. 13,14As immunization triggers an immune response and temporary disturbance of homeostasis, we suggest that one possible explanation for the more intense bleeding could be higher levels of cytokines.Vaccines induce the production of tumor necrosis factor α (TNFα), interferon γ, and interleukin 2 by T-helper 1 cells (Th1s). 15Th1s are dominant during menstrual bleeding in the normal menstrual cycle. 16Fα has been shown to be higher in women with heavy menstrual bleeding and mediates inflammation in the endometrium, which leads to vascular dilatation, hyperemia, and a delayed ability for endometrial repair. 17Interleukin 2 could alter the function of the platelets. 18These are factors that could have led to the observed changes in bleeding pattern.The most recently published study supports this possibility. 19ry few studies have reported on the risk factors for menstrual cycle irregularities after immunization against COVID-19. 12We were able to identify four risk factors for changes in the menstrual cycle after infection and vaccination, while surprisingly a combination of vaccines (homogeneous or heterogeneous) as well as different routes to immunization (infection and vaccination) had no significant effect on cycle variations.Changes in characteristics of the menstrual cycle depended significantly on age, pregnancy count, BMI, and previously observed irregularities.
Women with a higher BMI are at a risk of menstrual irregularities, attributed to higher testosterone levels and low sex hormones.
tions focused on general information (gender, age, body height, and weight), use of a menstrual cycle app, factors known or suspected to affect menstrual bleeding (age at menarche, working on night shifts, thyroid diseases, polycystic ovarian syndrome [PCOS], endometriosis), number of pregnancies and deliveries, contraceptive methods, and known previous cycle irregularities.Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters.The following information about infection with and immunization against COVID-19 was collected: vaccination status, month(s) and year(s) of vaccination, type(s) of vaccine, and, in the case of previous infection with SARS-CoV-2, month and year of infection.

1
Nguyen et al. found no changes in characteristics of the menstrual cycle caused by the COVID-19 pandemic, contrary to Takmaz et al., who suggested increased stress induced a prevalence of menstrual disturbances among healthcare providers in their survey. 3,5The possible trigger for the changes could be the vaccination against or infection with coronavirus, as changes in the menstrual cycle as a result of vaccination are known from human papillomavirus vaccination reports. 6An Italian survey performed by Lagana et al. and a study by Muhaidat et al. are in agreement with our results suggesting longer menstrual cycles and stronger

1609) Cycle length after vaccination/infection <25 days (short) 25-35 days (normal) <35 days (long)
tion, with the highest rate occurring after the second vaccination (265 [15.4%]).The duration of cycle irregularities varied from one TA B L E 3 Changes in the three menstrual cycle parameters of cycle length, menses duration, and bleeding volume following vaccination against COVID-19 and/or COVID-19 infection.Cycle length (n = Note: P-values refer to the marginal homogeneity test.
Time and duration of subjectively perceived changes in menstrual cycle.Did your cycle change after infection/vaccination?
20,21Furthermore, age is a known factor affecting TA B L E 4