Menstrual vulnerability to acute systemic stress: evidence from 884 women after COVID-19 infection

This retrospective cross-sectional study used an electronic questionnaire to assess menstrual characteristics before and after COVID-19 infection. Women of reproductive age who reported at least one spontaneous menstrual cycle after laboratory-confirmed COVID-19 infection, defined by either RT-PCR or self-administered rapid diagnostic testing, between January 1 and March 1, 2023 were eligible. Participants were recruited through a WeChat-based online platform. Exclusion criteria included: (1) hormonal therapy within the preceding three months; (2) pregnancy or lactation; (3) endocrine or structural gynecological disorders known to substantially affect menstruation. Of 1,033 responses collected, 149 were excluded, leaving 884 participants for analysis (Fig.  1 ). Fig. 1 Subject enrolment/screening log Subject enrolment/screening log This study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The protocol was approved by the Institutional Review Board of Peking University People’s Hospital (2023PHB034-001). Written informed consent was exempted in accordance with the urgent situation and the rules of the Ethics Committee. Participants completed the questionnaire using a mobile phone or tablet via a WeChat QR Code. The survey adhered to the platform’s terms of service. The questionnaire collected demographic data (e.g., age, age at menarche), menstrual history, COVID-19 vaccination status, date of last menstrual period, and method of COVID-19 diagnosis (either RT-PCR or self-administered rapid diagnostic tests). Participants were also asked to report menstrual changes following their COVID-19 infection, including cycle length, flow volume, duration of menstruation, dysmenorrhea, and intermenstrual bleeding. Because infection status was determined based on self-reported RT-PCR or rapid diagnostic test results, some degree of misclassification bias cannot be excluded. The questionnaire was developed by the investigators based on the study objectives and FIGO-based menstrual definitions. It was not formally pilot-tested or adapted from a previously validated instrument, which should be taken into account when interpreting the findings. Menstrual parameters and definitions of menstrual disturbances were predefined according to the International Federation of Gynecology and Obstetrics (FIGO) recommendations and were aligned with the questionnaire structure. A normal cycle length was defined as 24–38 days, duration ≤ 8 days, and subjectively normal bleeding volume. Deviations from these criteria were categorized as menstrual disturbances. Changes in flow were based on participants’ perceived increase or decrease compared with their usual pattern. Pre-existing menstrual irregularity was defined strictly according to FIGO criteria for normal menstrual cycle characteristics. Participants reporting a baseline cycle length shorter than 24 days or longer than 38 days prior to SARS-CoV-2 infection were classified as having pre-existing menstrual irregularity. Post-infection menstrual changes were assessed based on self-reported alterations in cycle length, volume, or duration. Data were analyzed using SPSS version 27.0. Continuous variables are presented as medians (IQR). Paired comparisons were analyzed using the Wilcoxon signed-rank test. Repeated measures across multiple time points were evaluated using the Friedman test, followed by Bonferroni-adjusted pairwise comparisons to control family-wise error. Categorical variables were analyzed using the chi-square test or Fisher’s exact test, as appropriate. Multivariable logistic regression analysis was performed to identify factors associated with menstrual changes. Covariates included in the multivariable model were selected a priori based on clinical relevance and previous literature. Multicollinearity was assessed using variance inflation factors, and no significant collinearity was detected. Missing data were minimal (< 5% for all variables) and were handled using complete-case analysis. A two-sided P value < 0.05 was considered statistically significant.

A total of 884 women with COVID-19 infections were included in the final analysis (Table  1 ). The mean age of the participants was 35.56 years (range: 18–58). The average age at menarche was 13.32 years (range: 9–19). Most participants (83.5%) did not have any history of AUB-related diseases. Conditions such as leiomyoma (11.3%), endometrial polyps (4.3%), and adenomyosis (3.3%) were reported in a smaller proportion of participants. Only 2.1% of participants had more than one comorbidity. Table 1 Demographics, baseline characteristics of females with COVID-19 Participants( n  = 884) Age, years Mean ± SD 35.56 ± 8.03 Range 18–58  50 19 (2.1%) Age of Menarche, years Mean ± SD 13.32 ± 1.51 Range 9–19  14 167 (18.9%) AUB-related disease Endometrial polyps 38 (4.3%) Leiomyoma 100 (11.3%) Adenomyosis 29 (3.3%) Malignant or atypical hyperplasia of endometrium 2 (0.2%) Abnormal coagulation 4 (0.5%) None 738 (83.5%) More than one comorbidity 19 (2.1%) Data are n (%) unless specified otherwise COVID-19 Coronavirus disease 2019, SD Standard Deviation Demographics, baseline characteristics of females with COVID-19 Data are n (%) unless specified otherwise COVID-19 Coronavirus disease 2019, SD Standard Deviation COVID-19 was diagnosed through rapid diagnostic tests (RDTs) in 69.6% of participants and through RT-PCR in 30.4%. A large proportion of participants (79.8%) had received at least three doses of the COVID-19 vaccine, while 5.5% were unvaccinated. The most commonly reported symptoms during the acute phase of infection included fever (94.2%), shortness of breath (76.5%), and insomnia (75.9%). A minority of participants (2.7%) were diagnosed with pneumonia based on CT imaging. Other reported symptoms included sore throat (40.5%) and asthenia (27.0%) (Table  2 ). Table 2 Clinical characteristics of females with COVID-19 infection Participants( n  = 884) Confirmed diagnosis RT-PCR 269 (30.4%) RDTs 615 (69.6%) Vaccination status: 1 dose 14 (1.6%) 2 doses 115 (13.0%) 3 doses (booster 1 dose) 634 (71.7%) 4 doses (booster 2 doses) 72 (8.1%) Not vaccinated 49 (5.5%) Signs and symptoms of COVID-19 Fever 833 (94.2%)  <38.5℃ 313 (35.4%)  ≥38.5℃ 520 (58.8%) Cough 171 (19.3%) Sore throat 358 (40.5%) Shortness of breath 676 (76.5%) Insomnia 671 (75.9%) Asthenia 239 (27.0%) Pneumonia on CT 24 (2.7%) Other symptoms 126 (14.3%) Data are n (%) unless specified otherwise COVID-19 Coronavirus disease 2019, RT-PCR Reverse transcription polymerase chain reaction, RDTs Rapid diagnostic tests Clinical characteristics of females with COVID-19 infection Data are n (%) unless specified otherwise COVID-19 Coronavirus disease 2019, RT-PCR Reverse transcription polymerase chain reaction, RDTs Rapid diagnostic tests Prior to infection, 94.6% of participants reported a menstrual cycle within the normal range of 24–38 days. The majority (86.5%) had a menstrual duration of 3–8 days, with 14.9% experiencing heavy menstrual bleeding. The most commonly reported menstrual symptom was dysmenorrhea, affecting 54.1% of participants, with 38.7% reporting mild pain, 14.3% moderate pain, and 1.1% severe pain. Additionally, 14.1% of participants experienced intermenstrual bleeding before COVID-19 infection (Table  3 ). Table 3 Menstrual characteristics before COVID-19 infection Participants( n  = 884) Cycle length 38d 38 (4.3%) Duration of menstrual periods 8d 99 (11.2%) Menstrual flow a Light 53 (6.0%) Normal 699 (79.1%) Heavy 132 (14.9%) Dysmenorrhea b None 406 (45.9%) Mild 342 (38.7%) Moderate 126 (14.3%) Severe 10 (1.1%) Intermenstrual bleeding Yes 125 (14.1%) No 759 (85.9%) Data are n (%) unless specified otherwise a Definition: Hypomenorrhea: No need to use sanitary napkins during menstruation. Normal: Using no more than 20 daily sanitary napkins during menstruation. Heavy menstrual bleeding: Using more than 20 daily sanitary napkins during menstruation b Definition: Mild: The pain did not affect daily study and life, and anodyne was not required. Moderate: The pain required anodyne and could be relieved. Severe: Anodyne could not relieve pain Menstrual characteristics before COVID-19 infection Data are n (%) unless specified otherwise a Definition: Hypomenorrhea: No need to use sanitary napkins during menstruation. Normal: Using no more than 20 daily sanitary napkins during menstruation. Heavy menstrual bleeding: Using more than 20 daily sanitary napkins during menstruation b Definition: Mild: The pain did not affect daily study and life, and anodyne was not required. Moderate: The pain required anodyne and could be relieved. Severe: Anodyne could not relieve pain Following COVID-19 infection, 74.9% of participants reported one or more changes in their menstrual cycle (Table  4 ). The most common changes were alterations in cycle length (47.6%), menstrual flow (41.7%), and menstrual duration (29.5%). Among the 662 participants with menstrual changes, 21.4% reported a shortened cycle, and 26.3% reported delayed menstruation. Some women experienced changes in menstrual flow: 25.5% reported a reduction, and 16.3% noted an increase. Furthermore, 8.4% of participants developed intermenstrual bleeding after infection, while 6.6% of participants who had previously experienced intermenstrual bleeding no longer reported it after infection. Table 4 Changes of cycle length, duration of menstrual periods, menstrual flow after COVID-19 infection Participants ( n  = 884) Cycle length Interval shortened by 2 weeks 59 (6.7%) Interval shortened by 1 week 130 (14.7%) No change 463 (52.4%) Interval prolonged by 1 week 143 (16.2%) Interval prolonged by 2 weeks 44 (5.0%) Interval prolonged more than 3 weeks 45 (5.1%) Duration of menstrual periods Shortened by 3 days or more 76 (8.6%) No change 623 (70.5%) Prolonged by 3 days or more 112 (12.7%) Prolonged spotting 73 (8.3%) Menstrual flow Decrease 225 (25.5%) No change 515 (58.3%) Increase 144 (16.3%) Data are n (%) unless specified otherwise Changes of cycle length, duration of menstrual periods, menstrual flow after COVID-19 infection Data are n (%) unless specified otherwise Dysmenorrhea also showed complex changes. Among participants who had dysmenorrhea prior to infection, 21.9% reported a decrease in severity, and 3.1% experienced increased pain. A total of 29.1% of participants who had no dysmenorrhea before infection developed it afterward. Notably, women with previously irregular menstrual cycles experienced a significantly higher rate of menstrual changes compared to those with regular cycles (73.0% vs. 62.3%, p  < 0.001). After adjustment for potential confounders, pre-existing menstrual irregularity remained independently associated with reporting post-infection menstrual changes (aOR 1.51, 95% CI 1.12–2.03, p  < 0.01). Other demographic and clinical variables showed no statistically significant associations (Table  5 ). Table 5 Univariable and multivariable logistic regression analyses of factors associated with reporting menstrual changes after COVID-19 infection Variable n (%) with changes Unadjusted OR (95% CI) p value Adjusted OR (95% CI) p value Age (years)  ≦ 40 630 (71.3%) 1.00 (reference) – 1.00 (reference) –  > 40 254 (28.7%) 1.19(0.88–1.62) 0.26 0.78(0.56–1.08) 0.13 Age of Menarche(years)  ≦ 14 717 (81.1%) 1.00 (reference) – 1.00 (reference) –  > 14 167 (18.9%) 1.27(0.88–1.83) 0.20 1.28(0.89–1.85) 0.19 AUB-related disease  No 738 (83.5%) 1.00 (reference) – 1.00 (reference) –  Yes 146 (16.5%) 0.94(0.65–1.36) 0.74 0.95(0.64–1.41) 0.80 Pre-existing menstrual irregularity  No 383 (43.3%) 1.00 (reference) – 1.00 (reference) –  Yes 501 (56.7%) 1.42(1.07–1.88) 0.02 1.51(1.12–2.03) < 0.01 Vaccination status  No 49 (5.5%) 1.00 (reference) – 1.00 (reference) –  Yes 835 (94.5%) 0.68(0.36–1.31) 0.25 0.67(0.35–1.29) 0.23 Fever  No 51 (5.8%) 1.00 (reference) – 1.00 (reference) –  Yes 833 (94.2%) 1.15(0.64–2.07) 0.63 1.25(0.69–2.27) 0.46 AUB Abnormal uterine bleeding Data are presented as odds ratios (ORs) with 95% confidence intervals (CIs) Adjusted ORs were obtained from multivariable logistic regression including age, age of menarche, AUB-related disease, pre-existing menstrual irregularity, vaccination status, and fever status Reference categories are indicated in each section Univariable and multivariable logistic regression analyses of factors associated with reporting menstrual changes after COVID-19 infection AUB Abnormal uterine bleeding Data are presented as odds ratios (ORs) with 95% confidence intervals (CIs) Adjusted ORs were obtained from multivariable logistic regression including age, age of menarche, AUB-related disease, pre-existing menstrual irregularity, vaccination status, and fever status Reference categories are indicated in each section

In this large cross-sectional survey of 884 reproductive-aged women, nearly three-quarters reported at least one change in menstrual characteristics following COVID-19 infection. The most frequently reported alterations involved cycle timing, bleeding volume, and duration. Women with pre-existing irregular cycles were significantly more likely to report post-infection disturbances after adjustment for demographic and clinical factors, suggesting the presence of a potentially vulnerable subgroup. Taken together, these findings support the concept of “menstrual vulnerability,” whereby some women may be more likely to perceive or report menstrual changes in the context of acute systemic stress. Pre-existing menstrual irregularity remained independently associated with a higher likelihood of reporting menstrual changes following COVID-19 infection (aOR 1.51, 95% CI 1.12–2.03, p  < 0.01). This suggests that baseline menstrual characteristics may influence individual susceptibility to menstrual disturbances during systemic illness. This association may also reflect greater symptom awareness or reporting sensitivity among women with pre-existing irregular cycles, and these possibilities cannot be fully disentangled in the present study. These mechanisms are not mutually exclusive and may coexist in shaping the observed association. Although causality cannot be inferred from this cross-sectional design, pre-existing menstrual patterns may serve as a clinical marker for identifying women more likely to report menstrual disturbances during acute physiological stress. Clinically, women with prior menstrual irregularities may benefit from closer monitoring and counseling when experiencing infectious or systemic stressors. Our findings align with growing evidence that menstrual function is sensitive to systemic physiological and psychological stressors [ 6 , 7 ]. Studies during the COVID-19 pandemic have reported changes in cycle regularity, bleeding patterns, and menstrual symptoms following infection or during periods of increased stress [ 8 – 11 ]. Compared with prior studies, often limited by smaller sample sizes or selected populations, our study provides real-world data from a relatively large Asian cohort. The large-scale infectious wave in China provided a unique population-level context to observe menstrual responses to acute systemic illness. Menstrual changes in our cohort were heterogeneous and bidirectional. Some participants reported delayed or reduced menstruation, whereas others reported increased flow or improvement of dysmenorrhea. This variability suggests the hypothesis of functional and potentially reversible neuroendocrine modulation rather than uniform reproductive impairment. Although direct viral effects on reproductive tissues have been proposed, current evidence suggests limited susceptibility of female reproductive organs to SARS-CoV-2 infection [ 12 – 15 ]. Acute infection is associated with inflammatory activation, fever, sleep disruption, metabolic stress, and psychological distress, all of which may transiently affect HPO axis regulation and ovulatory function [ 16 – 19 ]. In our cohort, insomnia was highly prevalent, and sleep disruption has been associated with alterations in hypothalamic–pituitary–ovarian axis regulation and reproductive hormone dynamics, suggesting a possible link between sleep quality and menstrual function [ 20 ]. Fever severity may also reflect systemic inflammatory burden, which has been linked to altered ovarian follicular dynamics and may transiently influence ovulatory function [ 21 ]. Although these variables were not independently modeled as mediators in this study, they represent plausible contributors within the stress-response framework. However, as hormonal and ovulatory markers were not assessed, these mechanisms remain speculative and require confirmation in future studies incorporating endocrine measurements. From a clinical perspective, menstrual changes may cause anxiety and lead to unnecessary medical consultations or investigations. Awareness that menstrual variability is commonly reported following acute systemic illness may help clinicians provide reassurance and appropriate guidance. Importantly, many reported changes were likely mild or transient and may not represent clinically significant menstrual disorders requiring medical intervention. Women with pre-existing menstrual irregularities may benefit from closer follow-up and individualized counseling. Clinicians should therefore interpret short-term menstrual variability after acute illness with caution, balancing reassurance for likely transient changes against appropriate evaluation when symptoms are persistent, severe, or accompanied by other gynecologic warning signs. These findings also support the broader concept that menstrual health is an important component of overall women’s health. Monitoring menstrual patterns may serve as a sensitive indicator of physiological resilience during systemic illness or public health emergencies. Incorporating menstrual health into routine counseling and post-infection care may improve patient-centered care. Cultural norms surrounding menstruation and differences in healthcare-seeking behavior may also influence symptom perception and reporting, and should be considered when interpreting the findings in this population. Several limitations should be acknowledged. First, this was a retrospective study based on self-reported menstrual characteristics, and recall bias cannot be excluded. Women who perceived menstrual changes may have been more likely to recall or emphasize cycle variations, potentially leading to overestimation of menstrual changes, whereas women with stable cycles may have underreported subtle fluctuations. In addition, subjective reporting of menstrual flow and cycle variation may have influenced effect estimates. Menstrual flow was estimated based on the number of sanitary pads used, which may not precisely correlate with objective blood loss and may be influenced by individual or cultural differences in pad usage habits. Standardized instructions regarding pad type or absorbency were not provided. Moreover, no objective menstrual tracking tools, such as menstrual diary records, app-based tracking, or clinical documentation, were available. Second, the absence of a contemporaneous non-infected control group limits the ability to attribute menstrual changes specifically to SARS-CoV-2 infection. External factors related to the pandemic environment, including psychological stress, behavioral changes, and environmental influences, may also have contributed. Third, although multivariable analyses were performed, residual confounding and potential misclassification of infection status cannot be completely excluded. Finally, recruitment through a WeChat-based online platform may have introduced selection bias toward younger, digitally engaged, or more health-conscious individuals, and the relatively young age profile of the cohort may limit generalizability to older populations or those with lower levels of digital engagement.

In conclusion, our findings suggest that menstrual changes may occur following SARS-CoV-2 infection in a subset of women. These changes most commonly involve cycle length, menstrual volume, and cycle regularity, and are likely multifactorial in nature. Although most reported alterations appear to be mild and potentially transient, awareness of possible menstrual variability following acute systemic illness may help clinicians provide appropriate counseling and reassurance. Further prospective studies with appropriate control groups are warranted to better characterize the temporal patterns, clinical significance, and underlying mechanisms of infection-associated menstrual changes.

Menstrual function reflects the dynamic interaction of the hypothalamic–pituitary–ovarian (HPO) axis, immune regulation [ 1 ], metabolic status, and psychosocial factors [ 2 ], and is widely considered an important indicator of women’s overall health. Even transient systemic stressors, including acute illness, inflammation, sleep disturbance, or emotional stress, may disrupt ovulatory function and lead to temporary alterations in cycle length, bleeding patterns, or menstrual symptoms. Previous studies have shown that acute systemic conditions and functional hypothalamic disturbances may result in reversible menstrual irregularities [ 3 ]. Since the emergence of coronavirus disease 2019 (COVID-19), increasing attention has been directed toward its potential reproductive and menstrual implications. Several surveys and small cohort studies have described menstrual changes during or after infection; however, existing evidence is heterogeneous and often limited by small sample sizes, Western-dominant populations, or lack of standardized assessments [ 4 , 5 ]. Following the rapid relaxation of epidemic control policies in China in late 2022, a large proportion of women experienced COVID-19 infection within a short period, providing a unique opportunity to evaluate menstrual responses to an acute systemic health event in a real-world setting. Asian women remain underrepresented in this field of research. Therefore, we conducted a large cross-sectional survey to characterize self-reported menstrual changes following COVID-19 infection and to explore whether women with pre-existing menstrual irregularities might represent a more vulnerable subgroup. Importantly, the study aimed to describe associations rather than establish causal relationships. In this study, the term “menstrual vulnerability” refers to an increased likelihood of reporting menstrual changes following acute systemic stress, which may reflect a combination of physiological susceptibility, symptom perception, and reporting behavior.

Supplementary Material 1 Supplementary Material 1