Assessing the role of ultrasound in uterine bleeding among patients on gender-affirming testosterone: a retrospective cohort study.

This was a retrospective cohort study examining ultrasound findings in transgender and gender-diverse individuals on testosterone who did and did not experience uterine bleeding within 3 months of ultrasound. The study was conducted at Boston Children’s Hospital and was approved by the Institutional Review Board (IRB-P00049922) in accordance with the Declaration of Helsinki, with exempt status granted, waiving the requirement for informed consent. All data were de-identified prior to analysis. The study population included transgender and gender-diverse individuals on testosterone who underwent pelvic ultrasound imaging between 2019 and 2024. Participants were identified through electronic medical record (EMR) review using a prescription code for testosterone therapy. Individuals who were cisgender males, not on testosterone, or did not have a uterus were excluded. Data abstraction was conducted by a team of researchers with expertise in reproductive health, hormone therapy, menstrual suppression, and gender-affirming care. Demographic and clinical data collected from the EMR included age at testosterone initiation, age at ultrasound, latest body mass index (BMI), latest testosterone and estrogen levels, and testosterone dose and route of administration. Additional reproductive history variables included age at menarche, history of menstrual concerns prior to testosterone therapy, and presence of gynecologic comorbidities. Bleeding history was documented, including whether uterine bleeding occurred within three months of ultrasound, whether the patient had achieved amenorrhea for at least six months after starting testosterone, and, if bleeding was present, the volume of bleeding reported. Ultrasound findings were reviewed for the indication for imaging, uterine size, endometrial stripe thickness, ovarian size, and any radiology comments regarding the endometrium, uterine body, or ovaries. Demographics and ultrasound findings of patients with and without uterine bleeding within three months of ultrasound were compared to assess potential differences. Statistical analyses were performed using Stata 18.0 (StataCorp LLC; College Station, TX). They included descriptive statistics, including counts, percentages, means and standard deviations. Chi square analysis was used to test for differences between groups for binary or categorical variables, and one-way analysis of variance (ANOVA) was used to test for differences between groups for continuous variables. P-values < 0.05 were considered statistically significant.

The final sample included a total of 144 individuals on gender-affirming testosterone at the time of ultrasound evaluation. The study compared the endometrial stripe thickness between individuals who experienced uterine bleeding within 3 months of ultrasound ( n  = 63) and those who did not experience bleeding ( n  = 81). As shown in Fig.  1 , there was no significant difference in endometrial stripe measurements between the two groups ( p  = 0.901). There were several incidental findings – two nabothian cysts, one possible 2 cm subserosal fibroid, and one unspecified subserosal lesion in a person with small, simple vaginal cysts that had been previously diagnosed. All but the last were in individuals who did not experience bleeding within 3 months of ultrasound. Table 1 Shows the demographic information for the study population. The median age at testosterone initiation was somewhat older for individuals without uterine bleeding (17.0 [15.0–18.0] vs. 15.0 [14.0–17.0], p  = 0.003). The median age at time of ultrasound evaluation was also slightly older for those without bleeding (20.0 [18.0–22.0] vs. 18.0 [16.0–22.0], p  < 0.001) 20.0. On average, individuals in both groups had been on testosterone therapy for approximately three years. There was no significant difference in BMI between the two groups Fig. 1 Endometrial stripe size by whether a person had experienced bleeding within three months of ultrasound. (No difference by ANOVA p  = 0.901) Endometrial stripe size by whether a person had experienced bleeding within three months of ultrasound. (No difference by ANOVA p  = 0.901) The mean testosterone level taken most recently before ultrasound evaluation was significantly higher for those with bleeding (535.1, 95% CI [453.0 617.2 ] vs. 417.1 95% CI [353.3 480.9]). Of the 63 patients who reported bleeding, 13 (20%) reported a testosterone dose change within 6 months, almost half of those 6 (46%) within one month of ultrasound. Testosterone levels were not significantly different between bleeders who did and did not report a dose change. The majority of individuals in both groups used intramuscular testosterone injections rather than other routes of administration: 60 (74.1%) of those without bleeding and 47 (74.6%) of those with bleeding. There was no difference in distribution of administration route between those who did and did not experience bleeding. Estrogen levels did not differ across bleeding groups for the 83 patients with recorded data. Several patients were using other hormonal medications in addition to testosterone therapy. There were significantly more individuals using progesterone only contraceptives at the time of ultrasound in those with uterine bleeding (41.3 vs. 18.5%; risk ratio (RR): 2.2 95% CI [1.3 3.2]). Those without uterine bleeding were significantly less likely to have a history of menstrual concerns prior to the start of gender-affirming testosterone therapy (31.2 vs. 48.1%; RR: 1.5 95% CI [1.001 2.09]). The most notable difference in terms of prior menstrual concerns was whether the patient had a history of heavy menses prior to testosterone start. There were significantly more patients that had a history of heavy menses prior to testosterone start in the group that had bleeding at time of ultrasound(19.1% vs. 6.2%; RR: 3.1 95% CI [1.2 6.7]) For those with uterine bleeding at the time of ultrasound, the majority experienced spotting or bleeding with wiping ( n  = 27, 42.9%), followed by monthly bleeding or bleeding like a period ( n  = 23, 36.5%), heavy bleeding ( n  = 7, 11.1%), and bleeding only with missed or late testosterone doses ( n  = 6, 9.5%). Ultrasound indications for those without uterine bleeding were mostly hysterectomy preparation ( n  = 66, 81.5%). Among individuals with uterine bleeding, the most common indication for ultrasound was the bleeding itself ( n  = 27, 42.9%), followed by pelvic pain ( n  = 20, 31.7%), and hysterectomy preparation ( n  = 15, 23.8%). Additional descriptive information regarding the ultrasounds can be found in Table  2 . Table 1 Demographics Bleeding within 3 months of ultrasound p -value No Yes ( N  = 81) Median [IQR] ( N  = 63) Median [IQR] Age at T start 17.0[15.0–18.0]) 15.0 [14.0–17.0] 0.003 Time on T (years) 3.4 (1.7) 2.9 (2.0) 0.07 Age at US 20.0 [18.0–22.0] 18.0 [16.0–22.0] < 0.001 Age at menarche 11.7 (1.6) 11.2 (2.1) 0.12 Latest documented BMI 27.7 [23.3–33.2] 25.7 [22.5–29.8] 0.086 Mean [95% CI] Mean [95% CI] Latest T Level 417.1 95% CI [353.3 480.9] 535.1 95% CI [453.0 617.2] 0.023 N (%) N (%) T application/prescription type  Transdermal 13 (16.0) 11 (17.5) 0.97  Injection -Intramuscular 60 (74.1) 47 (74.6)  Injection -Subdermal 2 (2.5) 2 (3.2)   Pellets 6 (7.4) 3 (4.8) Estrogen level  < 20 8 (27.6) 4 (15.4) 0.066  20–40 17 (58.6) 11 (42.3)  41–60 4 (13.8) 7 (26.9)  61–80 0 (0) 0 (0)  79.9+ 0 (0.0) 4 (15.4) Other hormonal medication  GnRH Agonist 5 (6.2) 4 (6.3) 1.00  Hormonal IUD 2 (2.5) 5 (7.9) 0.24  Combined Oral Contraceptives 1 (1.2) 0 (0.0) 1.00   Progesterone Only contraceptives 15 (18.5) 26 (41.3) 0.003 History of menstrual concerns prior to T 25 (31.2) 26 (48.1) 0.069  Dysmenorrhea 19 (23.5) 17 (27.0) 0.70  Irregular menses 6 (7.4) 10 (15.9) 0.12  Oligomenorrhea 2 (2.5) 0 (0.0) 0.50   Heavy menses 5 (6.2) 12 (19.0) 0.021 D Rows in bold have differences significant at p  < 0.05 Demographics D Rows in bold have differences significant at p   6 months on T -- 17 (43.6%) -- Bleeding classification  Only with missed or late doses 0 (0.0) 6 (9.5) 0.006  Spotting with wiping 0 (0.0) 27 (42.9) < 0.001  Heavy 0 (0.0) 7 (11.1) 0.003  Monthly/like a period 0 (0.0) 23 (36.5) < 0.001 Indication for ultrasound  Bleeding 0 (0.0) 27 (42.9) < 0.001  Pelvic pain 11 (13.6) 20 (31.7) 0.013  Hyst prep 66 (81.5) 15 (23.8) < 0.001  Other 7 (8.6) 11 (17.5) 0.13 Ultrasound Measurements  Length of uterus 7.1 (1.4) 7.0 (1.7) 0.9  Width of uterus 3.1 (0.9) 3.3 (1.1) 0.5  Endometrial stripe size 24.8 (7.8) 25.0 (8.5) 0.9  R ovary length 19.9 (7.5) 20.7 (9.4) 0.56  R ovary width 1.8 (0.5) 1.8 (0.8) 0.99  R ovary height 13.7 (6.9) 14.7 (7.1) 0.42  L ovary length 2.9 (0.9) 3.1 (1.0) 0.22  L ovary width 1.8 (0.6) 1.7 (0.6) 0.53  L ovary height 1.9 (0.7) 2.1 (0.9) 0.13 Data are presented as mean (SD) for continuous measures, and N(%) for categorical measures. Rows in bold have differences significant at p  < 0.05 Bleeding characteristics and ultrasound findings Data are presented as mean (SD) for continuous measures, and N(%) for categorical measures. Rows in bold have differences significant at p  < 0.05

Gender affirming hormone therapy is an essential component of care for many transgender and gender-diverse individuals. Testosterone hormone therapy for transgender and gender-diverse individuals seeking masculinization typically involves the administration of exogenous testosterone to promote the development of masculine secondary sex characteristics while suppressing feminine characteristics [ 1 , 2 ]. The physical effects of exogenous testosterone therapy vary among individuals. Desired changes associated with testosterone therapy may include voice deepening, facial and body hair growth, increased lean mass, shifts in fat distribution, clitoral enlargement, and cessation of menstruation [ 1 , 3 ]. Many of these changes can help transgender and gender-diverse individuals feel more aligned with their gender identity. Testosterone therapy can also lead to physiologic changes in structures such as the endometrium and ovaries. Given this, testosterone may alter uterine bleeding patterns. Amenorrhea is often a desired effect for individuals on gender affirming testosterone. Gender affirming hormone therapy with testosterone can help individuals achieve amenorrhea, often within a few months of initiating therapy [ 4 ]. Furthermore, Pappas et al. found that testosterone alone can help achieve amenorrhea in 90.8% of transmasculine individuals without the need for additional hormones [ 5 ]. However up to 33% of individuals on gender-affirming testosterone experience a return of uterine bleeding on testosterone even after initial amenorrhea [ 6 ]. This can exacerbate gender dysphoria and negatively impact mental health [ 7 , 8 ]. The causes of uterine bleeding in individuals on testosterone are unclear and likely involve a combination of causes. These include factors common to all individuals with uteri, such as polyps and fibroids, as well as causes related to other forms of menstrual suppression, including incomplete ovarian suppression leading to endometrial proliferation, or atrophic bleeding [ 5 , 9 – 12 ]. Ultrasound imaging is commonly used to evaluate abnormal uterine bleeding in patients, including in transgender and gender-diverse patients on testosterone. Endometrial thickness, ovarian volume, and masses or lesions on the uterus or ovaries can be evaluated using ultrasound [ 13 ]. Although ultrasounds are routinely performed for the purposes of uterine bleeding on testosterone, there has been little data to support this practice. Existing studies describe a spectrum of ultrasound findings in this population, but none have specifically compared findings between those who do and do not experience uterine bleeding [ 11 , 12 ]. This gap in knowledge limits clinicians’ ability to counsel patients on whether an ultrasound may be useful for their evaluation and management. Ultrasounds can be difficult for individuals who may have dysphoria associated with their reproductive organs or genitals. Pelvic ultrasounds can be particularly distressing for transgender and gender-diverse individuals, both emotionally and physically [ 10 ]. Therefore, it is essential to better understand the utility of ultrasounds for uterine bleeding on testosterone and whether the potential benefits outweigh the emotional and physical distress they may cause. The goal of this study was to evaluate differences in pelvic ultrasound findings among transgender and gender-diverse individuals on testosterone. Our primary research question examined whether individuals on gender-affirming testosterone exhibit differences in endometrial stripe thickness based on the presence or absence of uterine bleeding within three months of ultrasound evaluation. Our hypothesis was that there would be no difference between the findings in those done for individuals with and without uterine bleeding. By determining whether there are differences in pelvic organ characteristics via ultrasound, this study aims to influence health monitoring protocols for transgender and gender-diverse patients on testosterone with abnormal uterine bleeding.

Our study found no significant differences in pelvic ultrasound findings between transgender and gender-diverse individuals on testosterone with and without uterine bleeding, suggesting that routine ultrasound has limited diagnostic utility in this population. Given the multifactorial nature of bleeding on testosterone, including hormonal and metabolic influences, a trial of medical management may be preferable before imaging in low-risk individuals. While ultrasound remains an important tool for evaluating structural causes of abnormal uterine bleeding, its role in transgender and gender-diverse individuals on testosterone requires further investigation. Reducing unnecessary ultrasounds may help minimize barriers to care, including gender dysphoria and medical mistreatment concerns, while ensuring appropriate use of imaging when clinically indicated. Future research should work toward standardizing protocols for evaluating bleeding in transgender and gender-diverse individuals on testosterone. A prospective study that concurrently monitors testosterone levels and ultrasound imaging findings would provide critical insights into the relationship between testosterone therapy, pelvic organ changes, and both achievement of amenorrhea and resumption of undesired uterine bleeding. Such an approach could elucidate the mechanisms behind these changes and help inform more tailored management strategies for patients experiencing unwanted uterine bleeding on testosterone.

Our study aimed to compare pelvic ultrasound findings between transgender and gender-diverse individuals on testosterone who experienced uterine bleeding and those who did not. We found no significant differences in ultrasound findings between individuals with and without uterine bleeding within three months of imaging. This suggests that ultrasound may have limited diagnostic value in evaluating uterine bleeding among low-risk transgender individuals receiving testosterone therapy. Research on ultrasound differences in individuals on gender-affirming testosterone therapy with and without uterine bleeding is limited. Most existing studies focus on general anatomical changes due to testosterone rather than how these findings vary based on bleeding status [ 11 , 12 ]. Our study contributes valuable insight into the management of uterine bleeding in transgender and gender-diverse patients on testosterone therapy. Uterine bleeding is a well-documented phenomenon among transgender and gender-diverse individuals on testosterone therapy, despite its intended suppressive effects on menstruation [ 6 ]. Existing research suggests that the persistence of uterine bleeding despite testosterone therapy may be multifactorial. Potential factors leading to uterine bleeding may include differences in hormone levels, endometrial activity, and body mass index (BMI). Grimstad et al. found that individuals with uterine bleeding had lower mean serum testosterone levels and higher mean estradiol levels compared to those without bleeding [ 6 ]. Similarly, da Silva et al. reported that lower testosterone levels were significantly associated with persistent uterine bleeding [ 9 ]. This is inconsistent with our results given that the mean testosterone level taken most recently before ultrasound evaluation for those with uterine bleeding was significantly higher than those without uterine bleeding. This could have been due to increased testosterone levels prior to ultrasound as an attempt at managing the bleeding, however in our cohort the majority did not have a recently increased dose, and those who did did not have higher levels than those who had not had recently increased doses. This may have had to do with selection bias (those who do not bleed after increase dose don’t get ultrasounds) but may also have had to do with the clinicians involved who favor earlier evaluation (including ultrasound) prior to trialing a dose increase. It is known that testosterone therapy can lead to atrophy of the endometrial lining that causes reduced endometrial thickness. The endometrium in transmasculine individuals on testosterone can exhibit a range of histological patterns, from atrophic to proliferative states, and thicknesses that can include thin linings (e.g., 2 mm) to thickness more typical of menstruating individuals (e.g., 6 mm) [ 14 – 16 ]. According to a study by Asseler et al., transmasculine individuals on testosterone have significantly thinner endometrial linings compared to cisgender women, with a median endometrial thickness of 3.9 mm versus 4.9 mm, respectively [ 12 ]. Grimstad et al. also demonstrated that long-term consistent testosterone use often leads to near-complete endometrial atrophy on ultrasound [ 14 ]. This reduced endometrial thickness evident on ultrasound suggests a lack of endometrial proliferation due to testosterone exposure. Despite the endometrial atrophy and the majority achieving amenorrhea, endometrial activity persists in many transmasculine individuals on testosterone. Grimstad et al. noted that active endometrium was found in the majority of patients undergoing hysterectomy, indicating that endometrial proliferation can continue despite testosterone therapy [ 14 ]. This suggests that the presence of uterine bleeding does not necessarily correlate with significant differences in endometrial thickness or other ultrasound findings. This is supported by our results that there was no significant difference in endometrial thickness on ultrasound for individuals who were and were not experiencing bleeding. Although our study found no significant difference in BMI between participants with and without bleeding, previous research has identified higher BMI as a risk factor. Individuals with a higher BMI are more likely to experience persistent bleeding and may require additional interventions, such as oral progestogens, to achieve amenorrhea [ 9 ]. These findings highlight the multifactorial, and poorly understood nature of how and why individuals taking testosterone experience uterine bleeding. Current evaluation approaches often include hormonal assessments and pelvic ultrasound to rule out endometrial pathology. However, while hormonal and metabolic factors may contribute to uterine bleeding, our findings suggest that routine ultrasound evaluation may have limited diagnostic utility in this population. Given the lack of standardized guidelines for managing uterine bleeding in this population, our study adds to the ongoing effort to refine diagnostic strategies and optimize clinical care. Ultrasound is a primary imaging modality used for evaluating abnormal uterine bleeding, although research supporting this is primarily from the cisgender population. Ultrasound is especially useful for assessing the endometrial lining and detecting structural causes of abnormal uterine bleeding, including polyps, adenomyosis, fibroids, hyperplasia, and malignancy [ 16 ]. One important reason that clinicians perform ultrasounds is to exclude the rare but concerning finding of an endometrial mass, which would be a concern for malignancy [ 17 ]. Thus, the American College of Radiology (ACR) recommends TVUS as an initial imaging modality for abnormal uterine bleeding, which can be supplemented with transabdominal ultrasound for a comprehensive pelvic assessment [ 18 ]. However, it is likely that causes of uterine bleeding in transgender and gender-diverse people taking testosterone are likely different, and perhaps less likely to be visible, than causes in most cisgender individuals experiencing abnormal bleeding. Most studies suggest that testosterone does not increase the risk of endometrial malignancy [ 14 , 15 ], further calling into question the necessity of routine ultrasound in otherwise healthy individuals without additional risk factors. While differences seen in ultrasound findings might reasonably be expected to help provide insight into why some patients on testosterone experience persistent uterine bleeding, our data does not suggest that it has clear utility in this population. Underlying structural pathology detectable by pelvic ultrasound may not be the primary driver of uterine bleeding in this population. Instead, factors such as hormonal levels, endometrial activity, and BMI may play more significant roles. Therefore, while ultrasound is invaluable for ruling out structural causes of bleeding, it may not be as useful in preliminary management in the population on gender-affirming testosterone. Based on our data, clinicians may reasonably opt to forgo a pelvic ultrasound as primary screening in otherwise young healthy transgender and gender-diverse patients with no other concerns and instead trial a medical approach to management. Should bleeding persist, this would raise concerns for alternative structural causes, at which point ultrasound would still be indicated. Avoiding unnecessary ultrasound would also help to limit exposure to a source of significant dysphoria for some transgender and gender-diverse individuals. Even aside from the dysphoria often associated with a vaginal probe, multiple studies have shown that transgender and gender-diverse patients often avoid necessary medical care due to fears of mistreatment [ 19 ]. Furthermore, more than 70% of transgender and gender-diverse patients report having experienced at least one negative interaction within radiology settings [ 14 ]. Therefore, in addition to limiting the use of ultrasounds when they are not clearly indicated, it is essential for referring providers to recognize and address the potential discomfort transgender and gender-diverse patients may experience during ultrasound evaluations for abnormal bleeding [ 20 ]. Our study has several limitations. There is the potential for selection bias, as ultrasounds were conducted based on specific clinical indications rather than being systematically performed across the entire cohort of transgender and gender-diverse individuals who were taking testosterone and seen in our clinic. Additionally, this study did not account for certain confounding factors. For example, the testosterone dose was not standardized and there was likely a variation in the adherence to therapy. Not all bleeding may have been uterine in origin, as vulvovaginal atrophy can also be a cause of bleeding. An additional limitation of this study is its reliance on chart review, which is subject to variability in visit note documentation. This may have led to certain aspects of a patient’s history being incompletely captured, resulting in potential data gaps. The relatively young age range of our cohort (age 16–22) may influence the generalizability of our findings, as ultrasound characteristics and clinical findings may vary across different age groups, and our clinic populations tend to be less racially diverse than the transgender and gender-diverse population, although these demographics were not assessed in this study. It is important to acknowledge that our results do not account for the rare possibility of detecting a mass or lesion on ultrasound. However, it is important to balance the low likelihood of a missed mass against the possible consequences of recommending a potentially dysphoric test that may lead to disengagement from care. This is especially relevant when that test may be unlikely to provide useful information in a young, otherwise healthy study population who are experiencing uterine bleeding on testosterone. Alternative diagnostic approaches may be more appropriate in this population. Further research is warranted to explore more effective and patient-centered strategies for managing uterine bleeding and determining when, and if, ultrasound is appropriate.