Long COVID's Impact on Menstrual Cycles

There have been many reports of menstrual disturbance from women and people who menstruate who are suffering from long-term symptoms of COVID-19. Menstrual bleeding occurs following the decline in ovarian sex hormone production at the end of the menstrual cycle, resulting in the shedding of the upper two-thirds of the endometrium at menstruation. Menstrual symptoms have been standardised and defined by the International Federation of Gynecology and Obstetrics (FIGO) AUB System 11. Menstrual bleeding typically has a frequency of every 24–38 days, duration of no more than 8 days, variation in shortest to longest menstrual cycle of less than or equal to 7–9 days (age dependent) and a flow volume of subjectively normal. Common AUB symptoms are frequent/infrequent, prolonged, irregular and heavy menstrual bleeding (HMB). The symptom of HMB is defined as excessive menstrual blood loss that interferes with a woman’s physical, emotional, social and/or material quality of life2.

Pre-pandemic, AUB was extremely prevalent. Depending on the population studied and definition adopted, global figures vary but were consistently high3,4. One in three women was reported to find their menstrual loss excessive, with this figure rising to one in two as the menopause approaches5,6. Over 800,000 women sought treatment for HMB per year in the UK alone, with many more suffering in silence3,7. These menstrual symptoms can have a profoundly negative impact on quality of life, interfering with physical, social, mental and material wellbeing2,8. HMB is a leading cause of iron deficiency anemia in developed countries9, and, when extreme, can necessitate blood transfusion. HMB also affects work productivity, with results from a US study demonstrating that those with self-reported HMB were less likely to be working10. The annual indirect cost of menstrual bleeding disorders in the US was estimated to be $12 billion11. Therefore, any increase in prevalence of AUB due to COVID-19 has the potential to increase the gender health gap and add to the financial burden for health services and the economy.

Menstrual disturbances reported during the COVID pandemic may be due to COVID-19 vaccination, infection with the SARS-CoV-2 virus or pandemic-related stress and/or lifestyle changes. The contribution of each factor to menstrual disturbance is beginning to be delineated12,13,14 but is hindered by the lack of menstrual data collected during the pandemic using standardized nomenclature1 to facilitate scientific and clinical comparison of menstrual data globally. When examining COVID-19 vaccination and AUB, studies have revealed small changes in menstrual frequency15,16,17,18 but were consistent in their findings that any menstrual disturbance related to vaccination was transient, similar to findings with other vaccines16,19,20,21. Our previous UK-based online survey22 revealed that 18% of women reported a change to their menstrual symptoms after vaccination, but that menstrual symptoms were not significantly different in the COVID vaccination group when compared to those who had not been vaccinated. The impact of acute COVID-19 on menstruation was also examined by a few small studies in China early in the pandemic, revealing an association with menstrual disturbance23,24. These associations were found in subsequent larger, community-based US and UK studies22,25, consistent with infection with SARS-CoV-2 having a larger effect on menstruation than vaccination.

AUB and reproductive health in those with sustained post-COVID-19 infection sequelae (long COVID or long haul COVID) has been significantly understudied26. A recent WHO-led Delphi process reached a consensus that this post-COVID-19 condition occurs in individuals with a history of probable or confirmed SARS-CoV-2 infection, usually 3 months from the onset, with symptoms that last for at least 2 months and cannot be explained by an alternative diagnosis27. A patient-led survey of those experiencing long COVID symptoms revealed nine main symptom clusters: systemic, reproductive/genitourinary/endocrine, cardiovascular, musculoskeletal, immunologic/autoimmune, head, eyes, ear, nose and throat, pulmonary, gastrointestinal and dermatologic28. Evidence suggests that long COVID affects twice as many women as men and, of those under 50, women were five times less likely to report feeling recovered than men of the same age29. A survey of those with long COVID revealed menstrual issues were reported by 33.8%28. In a Spanish survey, those with suspected/diagnosed long-COVID-19 (n = 748) had an increased risk of self-reported menstrual alterations when compared to those who had never had COVID-19 or those with acute COVID-19 who had recovered30. These initial data suggest long COVID is associated with menstrual disturbance, but the type of menstrual disturbance was not reported.

As well as long COVID affecting menstrual symptoms, the ovarian hormone fluctuations across the menstrual cycle have the potential to impact the symptoms of long COVID. A cross-sectional study of long COVID patients found that over one-third of menstruating patients reported an exacerbation of their symptoms the week before or during menses28. The menstrual cycle and ovarian sex hormones may also modulate COVID disease susceptibility and severity26,31, with oestradiol known to boost the host immune response32. Women under 50 years old were reported to have an increased risk of developing long COVID29, consistent with ovarian sex hormones increasing the likelihood of developing long COVID. However, the effects of the menstrual cycle on longer-term symptoms of COVID-19 remain under-researched.

Defining the mechanisms that underpin any AUB associated with long COVID will inform clinical management. The underlying cause(s) of AUB may be classified as structural or non-structural, as outlined in the FIGO AUB System 21. Structural causes are those that can usually be detected during routine examination or investigations (e.g., imaging, histopathology) and include Polyps, Adenomyosis, Leiomyomas (fibroids) or Malignancy (PALM). Non-structural causes are not detected on imaging and include Coagulopathies, Ovulatory disorders, primary Endometrial disorders, Iatrogenic causes and those that are Not otherwise classified (COEIN). The rapid onset of menstrual disturbance described with SARS-CoV-2 infection favours a non-structural cause such as an ovulatory or endometrial disorder. Development and shedding of the endometrium at menstruation is controlled by the ovarian hormones. Perturbations of the cyclical ovarian sex hormone production (e.g., ovulatory disorders) can lead to changes in menstrual regularity and volume. Those with AUB due to endometrial disorders have previously been shown to have excessive endometrial inflammation at the time of menstruation33,34. Whether similar aberrations are present in those with long COVID remains to be determined. Defining the underlying mechanisms of AUB associated with long COVID will facilitate the precise treatment of menstrual disturbance. Similarly, uncovering the mechanisms that result in increased severity or number of long COVID symptoms across the menstrual cycle may reveal new treatment options for females suffering from long COVID.

Hence, we hypothesised (i) that long COVID is associated with increased reports of AUB, (ii) women with long COVID experience increased number and severity of their long COVID symptoms prior to and during menstruation, (iii) that those with long COVID have altered ovarian sex hormones production or response and/or excessive peripheral or endometrial inflammation. We tested these hypotheses using three approaches: (i) a large online UK COVID and reproductive health survey, (ii) a longitudinal study of long COVID symptoms across the menstrual cycle, and (iii) collection and analysis of carefully categorised biological samples of serum and endometrium at three phases of the menstrual cycle from those with and without long COVID.

In a UK survey, menstrual volume, duration and intermenstrual bleeding were increased in those with long COVID when compared to controls

Sample characteristics

Out of the 26710 individuals who completed our online survey, “The COVID-19 Pandemic and Women’s Reproductive Health”, we excluded participants who did not have a menstrual bleed in the 12 months preceding the survey, those who were post-menopausal or peri-menopausal, breastfeeding or pregnant, those who did not live in the UK, those enrolled in a clinical trial, and those with unknown vaccine, COVID or long COVID status. Since Long COVID was defined as infection ≥30 days ago, we excluded participants with very recent acute COVID infections (in the last 30 days) to ensure temporal comparability and prevent overrepresentation of newer variants in the acute COVID group (See Suppl. Fig. 1 for variant distribution across groups). The final sample size was 12187, of which 9423 (77%) had never been diagnosed with COVID (no COVID group), 1716 (14%) of participants reported previous acute COVID and 1048 (9%) had long COVID (Table 1). 4814 (40%) reported having been vaccinated, with either one (n = 4009) or two doses (n = 832). The median age was 36 years old (inter-quartile range (IQR): 29–43) for those with long COVID, while it was 31 years old (inter-quartile range (IQR): 25–39) for those with previous acute COVID and 32 years old (inter-quartile range (IQR): 25–40) for those who did not report COVID. Of note, 92% of participants were white, 64% were nulliparous, and 47% had a university or college degree. 57% reported one or more abnormal menstrual symptoms (e.g., irregular cycles, heavy bleeding, frequent or infrequent cycle length, prolonged periods) at baseline and 20% had a formal diagnosis of a condition known to affect reproductive function (e.g., endometriosis, polycystic ovary syndrome, HIV/AIDS, underactive/overactive thyroid, uterine polyps, uterine fibroids, eating disorders, interstitial cystitis).

Risk for “abnormal” menstrual characteristics

To investigate the independent effects of long COVID on abnormal menstrual parameters as defined by the FIGO criteria for abnormal uterine bleeding, we compared menstrual cycle parameters across 3 groups (Table 1)1: No COVID2, previous acute COVID and3 long COVID. The relationships between cycle parameters and the history of COVID-19 disease were adjusted for relevant menstrual cycle parameters before the pandemic, age, BMI, exogenous hormone use, and presence of diagnosed reproductive condition at baseline.

Menstrual Frequency (n = 9843, including 817 Long COVID, 1382 Acute COVID and 7644 controls). Data for menstrual frequency were missing for 1455 participants, and these were excluded from this specific analysis. We also excluded participants who reported “Too irregular to say” for the outcome variable “Cycle length during the pandemic” (n = 889), as we were interested in ascribing frequency. Another 1445 individuals were excluded due to missing data for baseline menstrual frequency before the pandemic. Across all groups of remaining participants, the most probable outcome was to report normal cycles (between 24 and 38 days, 71%), followed by frequent (<24 days, 26%) and infrequent cycles (>38 days, 3%) (Fig. 1A). The relative risk or risk ratio (RR) of reporting abnormal (frequent vs. normal cycles or infrequent vs. normal cycles) does not vary across groups, Table S1).

Menstrual regularity (n = 12,187). Across all groups of participants, the most probable outcome is to report regular cycles at the time of survey (less than 10 days difference between shortest and longest cycles, 79.7%), followed by very irregular (over 20 days difference, 9.8%), and somewhat irregular (between 10–20 days difference, 10.5%, Fig. 1B). The risk of reporting somewhat irregular vs. regular cycles increased by 39% for those with Long COVID, although the difference is not significant at the false discovery rate (RRR = 1.39, 95% CI = [1.03 to 1.87], p value = 0.04, FDR p value = 0.09; Table S2). The relative risks of reporting irregular vs. regular cycles are associated with obesity and the use of exogenous hormones (increased with progestin-based therapies and decreased with combined contraceptives).

Flow volume (n = 12,187). Across all groups of participants, the most probable participant reported outcome was “no changes” (41.0%), followed by “heavier” (24.9%), “heavier, and lighter” (19.2%) and “lighter”’ (14.9%) (Fig. 1C). As compared to control, a history of COVID-19 disease increased the risk of “heavier” vs. “normal” flow by ca. 19% for previous acute COVID (RRR = 1.19, 95% CI = [1.03 to 1.38], FDR p value = 0.05) and by 93% for long COVID (RR = 1.93, 95% CI = [1.59 to 2.35], FDR p value < 0.001). The risk of “lighter” flow vs. “no changes” increased by 36% for long COVID (RRR = 1.36, 95% CI = [1.08 to 1.71], FDR p value = 0.03) but did not vary between the control group and the acute COVID group. Long COVID also increased the risk of “lighter and heavier” flow as compared to “no changes” by 57% (RRR = 1.57, 95% CI = [1.26 to 1.96], FDR p value < 0.001, Table S3).

Menstrual duration (n = 1,938 including 192 Long COVID, 288 Acute COVID and 1458 controls). Data for menstrual duration were missing for 10,249 participants, and these were excluded from this specific analysis. Compared to the control group, the prevalence of menstruation lasting longer than 8 days is increased twofold for the long COVID group (PR = 2.26, 95% CI [1.46 to 3.49], FDR p value < 0.001), a tendency not observed for those with previous acute COVID-19 disease only (PR = 1.22, 95% CI [0.85; 1.77], FDR p value = 0.33, Table S4, Fig. 1D). The prevalence of periods longer than 8 days decreased with age but increased with abnormal menstrual duration at baseline and presence of diagnosed reproductive pathology at baseline, copper IUD use and the use of progestin-based contraceptives.

Intermenstrual bleeding or “spotting” (n = 12,187). As compared to the no COVID group, a history of COVID-19 increased the risk of more intermenstrual bleeding by 59% for long COVID (RRR = 1.59, 95% CI = [1.29 to 1.97], FDR p value < 0.001), but not for previous acute COVID (RRR = 1.15, 95% CI = [0.97 to 1.36], FDR p value = 0.22) (Fig. 1E). There was no association between a history of COVID-19 and less intermenstrual bleeding. Both more and less inter-menstrual bleeding increased with the use of hormonal contraceptives, the copper IUD, abnormal menstrual symptoms at baseline and presence of diagnosed reproductive pathology at baseline. (Table S5).

Missed episodes of menstruation (n = 12,187). As compared to the control group with no covid, the prevalence of reported “missed” and/or “stopped” periods increases by 39% in the long COVID group (PR = 1.39, 95% CI = [1.13 to 1.7], FDR p value = 0.003) but not in the previous acute COVID group (PR = 1.15, 95% CI = [0.97 to 1.37], FDR p value = 0.17, Table S6) (Fig. 1F). The probability of reporting missed or stopped periods decreases with age but increases with obesity, presence of diagnosed reproductive pathology at baseline and the use of exogenous hormones, especially progestin-based contraceptives.

In summary, reported menstrual flow volume, menstrual duration (>8 days), inter-menstrual bleeding and missed episodes of menstruation were significantly increased in those with long COVID versus no COVID. Menstrual frequency and regularity were unchanged. In contrast, in those with previous acute COVID, only menstrual volume was increased, but did not reach statistical significance.

A prospective study of long COVID symptoms revealed increased severity during the peri-menstrual and proliferative phases of the cycle