By Katherine Rivlin, MD, MSc
Associate Professor, Obstetrics and Gynecology, The Ohio State University Medical Center, Columbus
In a retrospective cohort analysis of prospectively tracked menstrual cycle data from the smart phone application “Natural Cycles,” the COVID-19 vaccine is associated with a less than one day change in menstrual cycle length, and no change in menses length.
Edelman A, Boniface ER, Benhar E, et al. Association between menstrual cycle length and coronavirus disease 2019 (COVID-19) vaccination: A U.S. cohort. Obstet Gynecol 2022; Jan 5. doi: 10.1097/AOG.0000000000004695. [Online ahead of print].
Post-COVID-19 vaccination changes to the menstrual cycle have been reported, particularly through social media, but little evidence exists.1,2 Fewer than 200 individuals have self-reported menstrual disturbances through the Vaccine Adverse Event Reporting System. Unfortunately, vaccine clinical trials did not track menstrual cycle outcomes.
People with normal menstrual cycles may experience sporadic changes to their cycle, such as skipping a menses or a temporary change in length. In most circumstances, this is normal. Yet, in the context of the COVID-19 vaccination, menstrual cycle variability may cause concern to patients. In the absence of data, fears that the vaccine affects the menstrual cycle could lead to vaccine hesitancy, and physicians have had little data to guide their counseling.
This retrospective analysis used prospectively collected menstrual cycle data from the Food and Drug Administration (FDA)-cleared digital fertility-awareness application “Natural Cycles” to assess associations between changes in the menstrual cycle or menses length and the COVID-19 vaccine.
Natural Cycles users voluntarily track menstrual cycle data for the purpose of pregnancy prevention or planning. Users provide consent to the use of their de-identified data for research purposes. Cycle data collected for this study ranged from October 2020 to September 2021, with first vaccine doses occurring between December 2020 and July 2021. The study included U.S. residents aged 18-45 years, with normal cycle lengths (24-38 days) and at least three cycles following pregnancy, or at least three cycles following the discontinuation a hormonal contraceptive.
The Natural Cycles application prompts users to enter a date of vaccination or to confirm if they are unvaccinated. The study included the three U.S. FDA-approved COVID-19 vaccine types, Pfizer, Moderna, and Johnson & Johnson. The study excluded those users without vaccine information or confirmation.
Each participant contributed data from six consecutive menstrual cycles and sociodemographic and clinical variables. Although the application requires that users provide their age, providing additional information is voluntary. For the vaccination group, the three cycles preceding vaccination, the vaccine cycle, and the two cycles following vaccination were included. For the unvaccinated group, six consecutive cycles were included.
The primary outcome was the within-individual change in cycle length from the three cycles preceding vaccination to the initial vaccination cycle, measured in days. In the vaccinated group, cycle four was the initial vaccine cycle. The second vaccine dose cycle varied depending on when an individual’s second vaccination occurred — in cycle four, five, or six. In the unvaccinated group, cycle four was the artificial first vaccine cycle, and cycle five was the artificial second dose cycle; the first three cycles were the artificial pre-vaccine cycles.
Secondary outcomes included the same within-individual change in cycle length for the second vaccine dose cycle, corresponding changes in menses length for the first and second vaccine dose cycles, and the proportion of participants who experienced clinically significant changes in cycle length (eight days or more). Secondary outcomes also included a subanalysis of people who received both vaccines in one cycle compared to those vaccinated over two cycles, and a comparison of cycle six to the three pre-vaccine cycles to assess whether cycle changes persisted over time.
Of 10,179 eligible Natural Cycles users, 3,959 met inclusion, for a total of 23,754 menstrual cycles. Most exclusions resulted from an insufficient number of menstrual cycles tracked. The final analytic sample included 2,403 vaccinated individuals and 1,556 unvaccinated individuals. The vaccinated group was slightly older, more likely to be nulliparous, to have a college education, to identify as white, and to live in the Northeast or West, compared to the unvaccinated group.
The vaccinated group had less than a one-day unadjusted increase in menstrual cycle length during the initial vaccine cycle compared to their three pre-vaccine cycles (0.7-day increase; confidence interval [CI], 0.47-0.94). The unvaccinated group had no change in cycle four compared to their first three cycles. The proportion of participants who experienced a greater than eight-day cycle change did not differ by vaccine status (4.3% unvaccinated, 5.2% vaccinated, P = 0.181).
Among participants who received a second vaccine dose, the unadjusted mean increase in cycle length during the second cycle was slightly higher, (0.91-day increase; CI, 0.63-1.19). A slightly higher proportion of this group also had a greater than eight-day cycle change (4.6% unvaccinated, 6.5% vaccinated, P = 0.017). This increase in cycle length came largely from a subgroup of 358 users who received both vaccines in the same cycle (cycle four). This group had two days unadjusted mean cycle length increase (2.38-days increase; CI, 1.52-3.24) and 10.6% had a greater than eight-day cycle change, compared to 4.3% in the unvaccinated group (P < 0.001). By cycle six, changes in cycle length between groups no longer differed, including for those who received both vaccines in one cycle. Menses length did not change for the first or second postvaccine cycle for any group.
COMMENTARY
Menstrual cycle length is controlled by the hypothalamic-pituitary-ovarian (H-P-O) axis. The H-P-O axis can be affected by life circumstances, environment, or health stressors.
When stressors occur during the first half of the menstrual cycle (the follicular phase, when a dominant follicle is recruited and matures), cycle length can change.3 The immune response generated by messenger ribonucleic acid (mRNA) vaccines could function as a stressor and temporarily affect the H-P-O axis, depending on the timing of vaccination. Because of the 21-day dosing schedule for the Pfizer mRNA vaccine and 28-day dosing schedule for the Moderna mRNA vaccine, receiving two doses in a single cycle necessitates that the first dose occurs during the early follicular phase for a normal cycling individual. This theory validates the higher rates of cycle change in those undergoing two vaccinations in once cycle.
This study is the first to track prospectively collected menstrual cycle data to better understand how the COVID-19 vaccination may affect the menstrual cycle. Although the changes in cycle length in the vaccinated group were technically statistically significant, the difference in cycle length was less than one day, which is not clinically significant. All changes had resolved by the sixth cycle, even for those receiving two doses in one cycle, and menses length was unaffected by vaccination. This study has a large sample size, and the prospectively collected data reduces recall bias, a clear limitation of prior data and anecdotal reports of cycle changes. It includes a non-vaccinated cohort for comparison, which is especially important given the potential confounding effect of generalized stress arising from the COVID-19 pandemic.
Results from this study may not be generalizable to the larger U.S. population. Natural Cycles users are avoiding all hormonal contraceptives and are more likely to be white, college-educated, and to have a lower body mass index compared to the general population. The study only included subjects who have normal menstrual cycles, which excludes a large subset of the U.S. population with irregular cycles.4 Larger studies with fewer exclusion criteria could better address individual concerns for patients experiencing menstrual cycle changes after vaccination, especially for those who do not fit inclusion in this study.
Normally menstruating individuals can experience small variations in cycle length. The less than one-day change documented in this study does not raise clinical concern and should not prompt additional medical evaluation. The American Society for Reproductive Medicine only recommends additional evaluation for secondary amenorrhea that lasts at least three months. Yet a delay of menses for as little as one week could require the exclusion of pregnancy.5 For an individual who is actively trying for or avoiding pregnancy, the potential cycle changes documented in this study could raise either hopes or concerns unnecessarily.
Although the findings of this study are reassuring both to patients and physicians, they also validate individual experiences of cycle changes following vaccination, especially for patients undergoing vaccination twice in one cycle. A clinician can consider providing anticipatory guidance to patients considering vaccination, particularly for those actively avoiding or trying for pregnancy, for those who receive two doses in one cycle, and for those experiencing vaccine hesitancy related to menstrual cycle changes.
REFERENCES
- Lee KMN, Junkins EJ, Fatima UA, et al. Characterizing menstrual bleeding changes occurring after SARS-CoV-2 vaccination. medRxiv 2021; Oct 21. https://doi.org/10.1101/2021.10.11.21264863. [Preprint].
- Brumfiel G. Why reports of menstrual changes after COVID vaccine are tough to study. NPR. Published Aug. 9, 2021. https://www.npr.org/sections/health-shots/2021/08/09/1024190379/covid-vaccine-period-menstrual-cycle-research
- Williams NI, Berga SL, Cameron JL. Synergism between psychosocial and metabolic stressors: Impact on reproductive function in cynomolgus monkeys. Am J Physiol Endocrinol Metab 2007;293:E270-276.
- Bull JR, Rowland SP, Scherwitzl EB, et al. Real-world menstrual cycle characteristics of more than 600,000 menstrual cycles. NPJ Digit Med 2019;2:83.
- Practice Committee of American Society for Reproductive Medicine. Current evaluation of amenorrhea. Fertil Steril 2008;90:S219-225.
