Abnormal uterine bleeding during the COVID-19 pandemic among reproductive-age women with and without HIV.

This study analyzed cross-sectional baseline data from women enrolled in the Study of Treatment and Reproductive Outcomes (STAR), a multi-site prospective cohort study of women of reproductive age (18–45 years) living with and without HIV [ 27 ]. As of 2026, enrollment in STAR is ongoing. This analysis includes data from women enrolled from 2021 to 2023 across the southern US, including in Atlanta, Georgia; Washington DC; Miami, Florida; Jackson, Mississippi; Birmingham, Alabama; and Chapel Hill, North Carolina. The 2021–2023 timeframe was selected to include women enrolled after the COVID-19 questions were added to the survey in 2021 through the initial STAR data delivery in 2023, because characteristics of the COVID-19 pandemic changed in late 2023 and in 2024, when non-mRNA COVID-19 vaccines became available, and the US government intermittently suspended the free at-home COVID-19 test program. Throughout the study period, vaccines were widely available, and COVID-19 infection was widespread. For STAR, inclusion criteria were being 18–45 years of age, being a cisgender woman, being willing and able to provide informed consent and complete study procedures in Spanish or English, and living near one of the STAR sites. Sites recruited participants from consent-to-contact databases, the local clinic population, community organizations, and community events. WWoH were required to report at least one sexual behavior or substance use behavior indicating elevated HIV risk, or to have a sexual partner living with HIV or at high risk for acquiring HIV, in the past five years. We confirmed HIV status using 4th generation rapid antigen/antibody testing or through documentation of positive HIV status. Details of the study protocol have previously been described [ 27 ]. We excluded women who had a health condition or were taking a medication known to interfere with menses, including pregnancy or breastfeeding in the past year, diagnosis of polycystic ovary syndrome, current use of hormonal contraceptives or an intrauterine device, history of uterine or ovarian cancer, hysterectomy or oophorectomy, self-report of reaching menopause, currently taking exogenous sex hormones, or missing data on irregular menses (menstrual questions were not asked if last menstrual period was over one year ago). Eligible women from the Miami and Atlanta sites had stored plasma tested for antibodies. These sites were selected because they were the two sites with the largest enrollment and because they represent diverse populations in terms of race and ethnicity. The study was conducted in accordance with the Declaration of Helsinki and was approved by the University of Miami single IRB. All participants completed written informed consent prior to enrollment. After completing an eligibility screener and informed consent, participants completed an interviewer-administered self-report survey. Blood was collected via venous puncture and processed into plasma and cryopreserved for future analyses. Plasma samples were shipped on dry ice from Atlanta to the Miami Center for AIDS Research laboratory where they underwent antibody testing. We measured demographic variables important to characterize the sample, including race, ethnicity, age, and BMI. Participants were asked to what extent they had difficulties “paying for basic needs.” Participants were asked three binary questions about experiencing AUB: [1] “in the last year, have you skipped any periods when you were not pregnant or breastfeeding?” [2], ”in the past one year, has your period come late or early by more than a week?”, and [3] “in the past year, have you noticed any spotting or bleeding between periods?” Participants who reported any late/early periods were also asked the number of periods that were late/early in the past year. A summary item was calculated for “any” versus “no” AUB in the past year. Participants were asked if they had “ever tested positive for COVID-19” and if they had ever received a vaccination for COVID-19. We defined “self-reported SARS-CoV-2” as any self-report of a positive COVID-19 test at any location and “self-reported vaccination” as any prior dose of a COVID-19 vaccine. A subset of 86 women (46% of the current sample), including all eligible women with baseline plasma samples at the Miami and Atlanta sites, had plasma tested for receptor binding domain (RBD) and nucleocapsid antibodies. Having detectable RBD antibodies could indicate prior SARS-CoV-2 infection and/or prior COVID-19 vaccination, so women with detectable RBD antibodies were tested for antibodies to nucleocapsid, which are indicative of prior SARS-CoV-2 infection and do not cross-react with vaccination. We defined “prior SARS-CoV-2 infection” as having detectable antibodies to nucleocapsid and/or self-report of a positive COVID-19 test in any setting. All women who self-reported vaccination also had detectable RBD antibodies, but there were six women who had detectable RBD antibodies despite denying vaccination and not having any other evidence for SARS-CoV-2 infection. It is possible these individuals misreported not receiving prior vaccination or had more distant COVID-19 infection where nucleocapsid antibodies waned before RBD antibodies, without any recent exposure to SARV-CoV-2; however, it is not possible to identify with certainty if these participants were previously vaccinated or infected. We consequently labeled these cases as “RBD+ Undefined” in the primary analysis and performed a sensitivity analysis to assess if conclusions change when either considering these participants to be vaccinated for COVID-19 or considering them to have been previously infected with SARS-CoV-2. For the antibody assays, 96 well plates were coated with 1 µg/mL SARS-CoV-2 antigen overnight at 4 °C, which was followed by blocking with 3% skim milk in Phosphate Buffered Saline containing 0.05% Tween-20 for two hours at room temperature. Heat inactivated plasma was added to the plates at 1:100 dilution and incubated at room temperature, which was followed by washing and the addition of anti-human IgG or IgM peroxidase secondary antibodies. After incubation, plates were developed with 3,3’,5,5’-Tetramethylbenzidine substrate followed by stopping reaction with 2 M sulfuric acid. The plate was read on an ELISA plate reader at 450 nm, and optical densities were background subtracted. A positive control standard, which was created by pooling plasma from eight convalescing COVID-19 patients, was run on each plate to calculate titers as relative units for all samples using non-linear regression interpolations to quantify the amount of IgG and IgM antibodies against nucleocapsid and RBD present in each specimen. The cut off for antibody detectability was three standard deviations above the mean titer for COVID-negative controls, which were samples from HIV uninfected individuals collected in 2018 under other IRB approved protocols. We placed participants into the following combined COVID-19 history categories based on combined antibody testing and self-report variables: [1] no infection nor vaccination [2], vaccination only [3], infection only [4], both infection and vaccination, and [5] RBD+ undefined. Perceived stress in the past month was measured using the Perceived Stress Scale, which measures perceived stress as a continuous scale (0 to 40), where higher scores indicate greater perceived stress [ 28 ]. Scores 14–26 are considered moderate stress, and scores 27–40 are considered high stress [ 28 ]. We calculated descriptive statistics for all key variables. Starting with the larger survey sample ( n = 188), we stratified by HIV status and examined the association between AUB in the past year (any versus none) and self-reported SARS-CoV-2 infection, self-reported COVID-19 vaccination, age, race, ethnicity, and BMI using chi-square tests for categorical variables and independent t-tests for continuous variables. We also examined the association between specific types of AUB with self-reported SARS-CoV-2 infection and COVID-19 vaccination. We subsequently restricted our analyses to the sub-sample with antibody test results available and examined the proportion of WWH and WWoH with past-year AUB according to whether they had detectable antibodies to nucleocapsid (indicative of recent SARS-CoV-2 infection) or any evidence of prior SARS-CoV-2 infection (self-report and/or via nucleocapsid antibody). The prevalence of AUB across the combined COVID-19 history groups was described. Missing data were minimal for all variables except for BMI, which had a higher level of missing data (28%). Missing data were handled via pairwise deletion in the primary analysis, while missing BMI values were also addressed using multiple imputation to impute 20 datasets in a sensitivity analysis. The sample size was pre-determined by the larger cohort study. We focused significance testing on the 2 × 2 analyses with adequate power to detect small-to-moderate effects, while considering the comparison of AUB across the combined COVID-19 history variable to be preliminary and descriptive. All analyses were conducted in R [ 29 ].

A total of 519 women were enrolled in STAR at the time of this analysis, with 375 enrolled after the COVID-19 variables were added, and 188 eligible for the current analysis after excluding women with conditions that interfere with menses. The primary reasons for being ineligible were recent pregnancy or breastfeeding ( n  = 90) or currently taking hormonal contraceptives ( n  = 67). A sub-set of 86 eligible women had antibody tests available. No variables differed between the sample with assays available and the sample of women without assays ( ns ) other than site, since all samples were selected from Atlanta and Miami. Table  1 presents demographics for the entire sample ( N  = 188) stratified by HIV status. Age ranged from 18 to 46 years, with a mean of 36 ( SD  = 7) years and 128 (68%) women being under 40 years-of-age. The sample was primarily non-Latina Black ( n  = 138, 73%), followed by non-Latina white ( n  = 18, 10%), and Latina white ( n  = 10, 5%). The sample was diverse in terms of socioeconomic status, with 38 (20%) reporting paying for basic needs was very hard. Perceived stress in the past month ranged from 0 to 40, with a mean of 15 ( SD  = 7) falling into the category of moderate perceived stress. When stratified by HIV status, WWH were four years older on average (37 vs. 33, p  < 0.001), and 37% of WWH and 22% of WwoH were over 40 years-of-age ( p  = 0.04). No other demographic variables differed by HIV status. Table 1 Demographics and key variables by HIV status Survey Variables WWH WWoH Overall ( n  = 115) ( n  = 73) ( N  = 188) Age , Mean (SD) 37 (6) 33 (7) 36 (7)  Missing 0 (0) 1 (1) 1 (1) Over 40 Years of Age , n (%) 43 (37) 16 (22) 59 (31)  Missing 0 (0) 1 (1) 1 (1) Race and Ethnicity  Latina Black 2 (2) 3 (4) 5 (3)  Latina Multi-Racial or Other 3 (3) 3 (4) 6 (4)  Latina White 6 (5) 4 (6) 10 (5)  Non-Latina American Indian/Alaskan Native 1 (1) 0 (0) 1 (1)  Non-Latina Asian 0 (0) 1 (1) 1 (1)  Non-Latina Black 86 (75) 52 (71) 138 (73)  Non-Latina Multi-Racial or Other 5 (5) 4 (5) 9 (5)  Non-Latina White 12 (10) 6 (8) 18 (10) Perceived Stress Scale , Mean (SD) 15 (7) 15 (8) 15 (7) Body Mass Index , Mean (SD) 34 (11) 33 (11) 33 (11)  Missing, n (%) 40 (35) 12 (16) 52 (28) Had Abnormal Uterine Bleeding in Past Year , n (%) 53 (46) 39 (53) 92 (49) Skipped Any Periods , n (%) 23 (20) 19 (26) 42 (22)  Missing, n (%) 8 (7) 5 (7) 13 (7) Had Late/Early Periods , n (%) 38 (33) 29 (40) 67 (36)  Missing, n (%) 0 (0) 1 (1) 1 (1) Number of Late/Early Periods , Mean (SD) 4 (3) 3 (3) 3 (3) Spotting/Bleeding Between Periods , n (%) 19 (17) 14 (19) 33 (18)  Missing, n (%) 0 (0) 1 (1) 1 (1) Tested Positive for SARS-CoV-2 via Self-Report , n (%) 37 (32) 27 (37) 64 (34) Vaccinated for COVID-19 via Self-Report , n (%) 69 (60) 40 (55) 109 (58) Antibody Sub-Sample Variables WWH ( n   = 44) WWoH ( n   = 42) Overall ( N   = 86) Detectable Nucleocapsid Antibodies (SARS-CoV-2 Infection) , n (%) 22 (50) 26 (62) 48 (56) SARS-CoV-2 via Self-Report or Antibodies , n (%) 28 (64) 31 (74) 59 (69) Vaccinated for COVID-19 via Self-Report , n (%) 24 (55) 23 (55) 57 (55) COVID-19 History via Self-Report or Antibodies , n (%)  Infection + Vaccination 15 (34) 17 (40) 32 (37)  Infection Only 13 (30) 14 (33) 27 (31)  Vaccination Only 9 (20) 6 (14) 15 (17)  No Vaccination or Infection 3 (7) 3 (7) 6 (7)  RBD Antibody Positive, but Undefined (No Other Evidence for Prior Vaccine or Infection) 4 (9) 2 (5) 6 (7) Note. WWoH = Women without HIV. WWH = Women with HIV. RBD = Receptor Binding Domain. Values were rounded to the nearest whole number Demographics and key variables by HIV status WWH ( n   = 44) WWoH ( n   = 42) Overall ( N   = 86) Note. WWoH = Women without HIV. WWH = Women with HIV. RBD = Receptor Binding Domain. Values were rounded to the nearest whole number Overall, 92 (49%) women had AUB in the past year, which included 67 (36%) with late or early periods, 42 (22%) with skipped periods, and 33 (18%) with spotting/bleeding between periods. Among women with any late/early periods, the mean number of late/early periods was 3.3, with a range from 1 to 12. AUB did not differ between WWoH (53%) and WWH (46%). In the overall sample, 109 (58%) reported being vaccinated, which did not differ between WWoH (55%) and WWH (60%). In the sub-sample with antibody testing, 47 (55%) reported being vaccinated, all of whom had detectable RBD antibodies. In the overall sample, 64 (34%) self-reported testing positive for SARS-CoV-2, which did not differ between WWoH (37%) or WWH (32%). In the sub-sample with antibody testing, 26 (30%) self-reported testing positive for SARS-CoV-2, but 48 (56%) had detectable nucleocapsid antibodies, indicating that about half of individuals with antibody evidence for prior SARS-CoV-2 infection had not received a prior positive SARS-CoV-2 test. All but one participant self-reporting prior COVID-19 had detectable RBD antibodies. When combining self-report and nucleocapsid antibody testing, 59 (69%) had evidence for prior SARS-CoV-2 infection; of these, 11 participants reported testing positive for COVID-19 but did not have detectable nucleocapsid antibodies. In the sample with antibody testing, prior SARS-CoV-2 infection did not significantly differ between WWoH (74%) and WWH (64%). When combining self-report and antibody testing, 32 (37%) had both infection and vaccination, 27 (31%) had infection only, 15 (17%) had vaccination only, 6 (7%) had neither vaccination nor infection, and 6 (7%) were RBD+ undefined (i.e., had detectable RBD antibodies with no other evidence for infection or vaccination). Among participants with self-reported vaccination, about half had prior COVID-19 infection: 24 (51%) vaccinated participants had detectable nucleocapsid antibodies, 18 (38%) vaccinated participants self-reported testing positive for SARS-CoV-2, and 32 (68%) vaccinated participants had evidence of prior SARS-CoV-2 infection when combining self-report and antibody testing. Among WWoH in the overall sample, AUB in the past year was associated with younger age (31.6 vs. 35.3, p  = 0.04) and higher perceived stress in the past month (16.9 vs. 13.1, p  = 0.04) (Table  2 ). AUB did not vary by race, ethnicity, or BMI among WWoH. No demographic variables were associated with AUB among WWH. To ensure that results were not biased by missingness in BMI, we used multiple imputation to impute 20 datasets with BMI; findings of this analysis were qualitatively similar, with no associations found between BMI and AUB. Table 2 Self-Reported COVID-19 history, demographics, and stress among women with and without HIV associated with abnormal uterine bleeding in the past year WWH WWoH No AUB ( n  = 62) Yes AUB ( n  = 53) P -Value No AUB ( n  = 34) Yes AUB ( n  = 39) P -Value Age , Mean (SD) 38 (6) 36 (6) 0.167 35 (7) 32 (7) 0.036* Perceived Stress , Mean (SD) 14 (7) 15 (7) 0.486 13 (8) 17 (8) 0.040* Body Mass Index , Mean (SD) 32 (11) 36 (11) 0.186 33 (11) 33 (11) 0.805 Latina/Hispanic , n (%) 6 (10) 5 (9) 1.00 5 (15) 5 (13) 1.00 Race , n (%) 0.805 0.347  American Indian/Alaskan Native 1 (2) 0 (0) 0 (0) 0 (0)  Asian 0 (0) 0 (0) 1 (3) 0 (0)  Black 48 (77) 40 (75) 28 (82) 27 (69)  Multi-Racial 3 (5) 1 (2) 1 (3) 3 (8)  Other 2 (3) 2 (4) 0 (0) 3 (8)  White 8 (13) 10 (19) 4 (12) 6 (15) Note. WWoH = Women without HIV. AUB = Abnormal Uterine Bleeding. *P-value significant at p < 0.05. Values were rounded to the nearest whole number. P-values are from chi-square tests, Fisher’s exact tests, or independent t-tests. Body Mass Index was missing for 28% of the overall sample Self-Reported COVID-19 history, demographics, and stress among women with and without HIV associated with abnormal uterine bleeding in the past year Note. WWoH = Women without HIV. AUB = Abnormal Uterine Bleeding. *P-value significant at p < 0.05. Values were rounded to the nearest whole number. P-values are from chi-square tests, Fisher’s exact tests, or independent t-tests. Body Mass Index was missing for 28% of the overall sample Rates of AUB across COVID-19 variables are shown in Fig.  1 for WWoH. In the overall sample, AUB was significantly more common among WWoH who self-reported testing positive for SARS-CoV-2 compared to WWoH who never tested positive for SARS-CoV-2 (70% vs. 44%, p  = 0.03). In the sample that underwent antibody testing, WWoH with detectable nucleocapsid antibodies had a higher proportion of AUB versus WWoH without detectable antibodies (62% vs. 38%, p  = 0.13), but this did not reach statistical significance. When SARS-CoV-2 infection was assessed via a combination of self-report and/or nucleocapsid antibody testing, AUB was more common among WWoH with prior SARS-CoV-2 infection versus WWoH without prior SARS-CoV-2 infection (61% vs. 27%, p  = 0.05) (Fig.  1 ), and the mean number of late/early periods was higher among WWoH with prior SARS-CoV-2 infection versus WWoH without prior SARS-CoV-2 infection (3 vs. 1, p  = 0.04). Among WWH (Fig.  2 ), there were no associations of SARS-CoV-2 infection, defined via self-report in the overall sample or via nucleocapsid antibodies in the sample with antibody testing, with any AUB or with specific types of AUB. Fig. 1 Proportion of abnormal uterine bleeding among women living without HIV for each self-report and antibody-confirmed COVID-19 variable. Proportion of abnormal uterine bleeding were significantly higher among women who self-reported testing positive for SARS-CoV-2 and who have SARS-CoV-2 infection by either self-report of a prior positive test and/or the presence of detectable antibodies to nucleocapsid protein. * p < 0.05 Proportion of abnormal uterine bleeding among women living without HIV for each self-report and antibody-confirmed COVID-19 variable. Proportion of abnormal uterine bleeding were significantly higher among women who self-reported testing positive for SARS-CoV-2 and who have SARS-CoV-2 infection by either self-report of a prior positive test and/or the presence of detectable antibodies to nucleocapsid protein. * p < 0.05 Fig. 2 Proportion of abnormal uterine bleeding among women living with HIV for each self-report and antibody-confirmed COVID-19 variable. There were no significance differences found for any COVID-19 variables tested Proportion of abnormal uterine bleeding among women living with HIV for each self-report and antibody-confirmed COVID-19 variable. There were no significance differences found for any COVID-19 variables tested Among WWoH in the overall sample, self-reported vaccination was not associated with AUB (60% among vaccinated persons vs. 45% among unvaccinated persons, p  = 0.21). Self-reported vaccination in the overall sample was associated with a greater number of instances of late/early periods among WWoH with at least one late/early period (3.58 among vaccinated persons vs. 1.50 among unvaccinated persons, p  = 0.01). Among WWH in the overall sample, self-reported vaccination was not associated with AUB over the past year (45% among vaccinated persons vs. 48% among unvaccinated persons, p  = 0.76) or with specific types of AUB ( ns ). Sensitivity analyses were conducted to test whether conclusions changed when re-categorizing the six participants who had detectable RBD but did not have additional evidence for prior infection or vaccination as either (a) being previously SARS-CoV-2 infected or (b) as being vaccinated. When these participants were considered as having prior SARS-CoV-2 infection, the number of women with prior SARS-CoV-2 rose to 65 (76%). When these participants were considered vaccinated, 53 (62%) had evidence of vaccination. AUB remained higher in WWoH with SARS-CoV-2 infection (61%) compared to women without (22%), although the finding was no longer statistically significant ( p  = 0.06). AUB remained unrelated to vaccination in WWoH (53% among vaccinated vs. 52% among unvaccinated, p  = 0.95). Among WWH, AUB remained unassociated with SARS-CoV-2 infection (47% with SARS-CoV-2 vs. 50% without, p  = 1.0) and vaccination (39% among vaccinated vs. 63% among unvaccinated, p  = 0.13).

In this study of reproductive-age women during the COVID-19 pandemic, self-reported AUB in the past year - defined as periods that were late/early by more than one week, spotting/bleeding in between periods, and missed periods - were reported by nearly half of women, regardless of HIV status. This AUB prevalence is higher than has been previously reported before the pandemic (e.g., 22–25%), which is especially notable because we excluded women who had many known causes of AUB [ 2 , 5 ], although we cannot discount the possibility that women had alternative causes of AUB, such as changes to contraception or weight in the past year. In this study, AUB was more common among WWoH who self-reported previously testing positive for COVID-19. Although this finding could be driven by increases in SARS-CoV-2 testing rather than COVID-19 infection itself, results were supported by nucleocapsid antibody testing for confirmed SARS-CoV-2 infection. When we conducted sensitivity analyses with an alternate definition of SARS-CoV-2 infection, which included participants with ambiguous antibody testing (RBD positive and nucleocapsid negative, but no self-report of vaccination) that could result from more distant SARS-CoV-2 infection, results followed the same pattern. Prior research suggests that individuals with greater COVID-19 symptoms are more likely to experience AUB after infection, which could further explain an association between getting tested for COVID-19 (i.e., because of symptoms) and having AUB [ 11 ]. There were no associations between COVID-19 vaccination and overall AUB, although the number of late/early periods was slightly higher among WWoH who reported vaccination and who had at least one late/early period compared to WWoH who were not vaccinated. When considering antibody-confirmed SARS-CoV-2 infection, the lowest rates of AUB were among WWoH with no SARS-CoV-2 infection, regardless of prior COVID-19 vaccination. Prior studies demonstrated an association between vaccination and short-term changes to menstrual patterns symptoms, and there is biological plausibility for vaccines to influence menstrual irregularity [ 14 – 16 ], so it is possible that a small association between vaccination and AUB could also be found in our study given a larger sample size, or if we had measured early/late periods with briefer delays (i.e., shorter than one week); however, any relationship between vaccination and AUB would be weaker than the relationship between SARS-CoV-2 infection and AUB, which has also been found in prior research [ 13 ]. Sensitivity analyses suggested that results were robust to alternate definitions of COVID-19 vaccination based on re-categorizing women with ambiguous antibody testing. However, a lack of significant relationships should be considered a preliminary finding that requires confirmation in larger samples, as our small sample could lead to Type II error. Findings suggest AUB was common even among women without SARS-CoV-2 infection or vaccination, and perceived stress in the past month was associated with past-year AUB among WWoH. Although our cross-sectional observational study was not designed to determine if there is a causal relationship between stress and AUB, relationships between stress and AUB have also been found in prior studies [ 6 , 21 , 23 ]. While our conclusions are preliminary and not definitive, this hypothesis-generating research highlights the potential importance of psychosocial factors in understanding the mechanisms through which the COVID-19 pandemic could be associated with AUB. Given prior research that suggests COVID-19 vaccination may be protective against AUB after SARS-CoV-2 infection [ 19 ], and given the high likelihood of SARS-CoV-2 infection overall, avoiding the COVID-19 vaccine due to concerns about AUB is unlikely to be an effective strategy for reducing AUB risk. The relationship between SARS-CoV-2 infection and AUB was only found among WWoH, and not among WWH. Infection with HIV was not associated with AUB in our study, and there were no factors associated with AUB among WWH. WWH in our sample were slightly older on average, and age has been previously associated with AUB [ 2 ]. WWH often have higher rates of co-morbid conditions, including diabetes and obesity, which are known to be associated with higher risk of AUB [ 2 , 30 ], although BMI in the current sample was not associated with AUB. Since rates of AUB were higher among WWH than expected in the general population, there is a need for research to understand distinct causes of AUB among WWH, including research with larger samples to reduce the risk of Type II error due to small sample size. The current study was descriptive and not designed to provide mechanistic insights about the causes of AUB. This study is limited in that we cannot confirm the temporal order between AUB onset, psychosocial factors, and SARS-CoV-2 infection, since all measures were retrospective and about half of individuals with antibody evidence of prior SARS-CoV-2 infection did not report having a positive COVID-19 test prior to participating in the study. Some measures differed in recall period (e.g., perceived stress in the past month) compared to the measure of any AUB in the past year, precluding the ability to make any assumptions about temporal relationships. For example, it is possible that AUB preceded or followed COVID-19 infection. Our study was conducted during the COVID-19 pandemic, when women may have been paying greater attention to menstrual patterns due to reports of AUB in the media, and women may have experienced other factors that could impact stress and AUB, such as food insecurity or sleep disruption, which could serve as confounding variables in future causal research. Prospective data collection using menstrual tracking apps may be an effective strategy to reduce measurement error in future research. Because this study was started during the COVID-19 pandemic, it is not possible to know whether rates of AUB were lower prior to the pandemic among this sample; however, future cohort analyses will provide the opportunity to track AUB over time. Finally, not all women in our sample had antibody testing, since available resources only allowed us to test samples in Miami and Atlanta. Self-reported SARS-CoV-2 testing underreported the true number of infections, and the smaller antibody sample had weaker power to identify significant differences that may be evident in a larger population; however, the associations of AUB and antibody confirmed COVID-19 variables were consistent with the associations between AUB and self-reported COVID-19 variables. The primary finding of this descriptive study is the high prevalence of self-reported AUB among women during the COVID-19 pandemic. The presence of AUB during the COVID-19 pandemic has not been found to pose any direct threats to reproductive health. However, there may be indirect risks due to the use of menstruation as an indicator for family planning. In the US, 25% of women seeking to become pregnant use fertility-awareness methods informed by the timing of menstruation to enhance the chance of pregnancy [ 31 ]. From 2013 to 2015, 3% of women engaging in pregnancy prevention used fertility awareness methods to prevent pregnancy [ 31 , 32 ]. About 7% of cisgender women living with HIV in the US have had sex without using any HIV prevention strategy in the last year (including condoms or other methods, such maintaining an undetectable viral load), and about 86% of STAR cohort participants had sex without a condom, meaning there is a high risk for pregnancy given the low uptake of hormonal contraceptives [ 24 , 33 ]. Menstruation-based fertility awareness methods have a high failure rate for the purposes of contraception, with 11–14% of women experiencing an unintended pregnancy across a year of typical use [ 34 ]. Even slight deviations in menstrual regularity can increase the risk of unintended pregnancy or delay identification of pregnancy, decreasing options for pregnancy termination and avoidance of teratogens that affect embryo development [ 2 ]. This is especially relevant in the southern US, where Florida, Georgia, and South Carolina currently limit abortions to the first six weeks of pregnancy. Women should be made aware of preliminary data that suggests AUB may have been more common as a consequence of the COVID-19 pandemic, and caution should be encouraged for future novel viral infections. Although these findings are preliminary and require future research with larger longitudinal samples to confirm causality, given the high risk of unintended pregnancy and general benefits of condom use, women should be provided this education and encouraged to use alternative contraceptive methods such as condoms instead of or in addition to fertility-awareness, opt to take regular pregnancy tests if at risk for unintended pregnancy and experiencing AUB, and consider tracking their menstrual periods and speaking with their medical providers should concerns about AUB arise. We found no associations between COVID-19 vaccination and AUB, which makes a strong causal link between vaccination and AUB seem unlikely, although it cannot be definitively ruled out based on our study design and small sample size, which may have been underpowered to identify a small relationship between COVID-19 vaccination and AUB. However, knowledge of our findings may be sufficient to decrease vaccine hesitancy related to concerns about AUB.

AUB occurred in almost half of women during the COVID-19 pandemic (2021–2023) regardless of HIV status, with higher rates among WWoH with prior SARS-CoV-2 infection compared to WWoH without SARS-CoV-2 infection. No association was found between AUB and COVID-19 vaccination, although this result should be considered preliminary due to our small sample size, and null findings should not be interpreted as evidence of no association. Further research is needed to understand mechanisms underlying AUB among women with and without HIV, including stress and infection among WWoH, and other unique factors associated with AUB among WWH. A better understanding of mechanisms contributing to AUB could enhance reproductive health knowledge and empower women to make informed decisions regarding menstrual health, fertility, and pregnancy prevention.