Variations in Volume: Breast Size in Trans Women in Relation to Timing of Testosterone Suppression

Abstract

Transgender women are individuals who have a female gender identity but a male sex assignment. The discrepancy between their gender identity and their assigned sex at birth may cause gender dysphoric feelings. This can result in a negative body image, low self-esteem, and even depression (1-3). An increasing number of trans women are seeking medical care to induce feminization and minimize these feelings of gender dysphoria (4).

Currently the medical treatment for transgender women consists of a combination of anti-androgens and estrogen therapy (5). This therapy, also known as gender-affirming hormone therapy (GAHT), induces changes in fat distribution, softening of the skin, and breast development (6-8). The latter is considered an important treatment outcome by trans women (9).

Several studies have shown that breast development in adult trans women is different from that in cis women (female sex assignment at birth and female gender identity). In trans women, breast volume is smaller and the breasts are located more caudally and more laterally on the chest (10, 11). Breast volume is typically expressed in cup sizes, which indicate the difference between the breast and chest measurements and range from AA to G. After 3 years of GAHT, 73% of the trans women had a cup size smaller than A (10). Even though more than half of trans women are satisfied with their breasts after hormone therapy, dissatisfaction with breast size often results in a wish for breast augmentation surgery (10, 12, 13). Recent research showed that 41% of trans women treated at the Amsterdam gender identity clinic have had breast augmentation and 40% was thinking about having it in the future (13).

The differences in breast development between trans and cis women could be the result of many factors. One of the hypotheses is that exposure to testosterone during puberty might negatively impact breast development in trans women. Testosterone may lead to irreversible changes in the mammary gland and/or in the musculoskeletal structure of the thorax, such as the expansion and elongation of the chest and widening of the sternum resulting in more laterally located nipples (14).

When gender dysphoria is diagnosed at an adolescent age, puberty is typically still ongoing. Adolescents can be treated with a gonadotropin-releasing hormone agonist (GnRHa) with addition of estrogen supplementation once they reach an age at which they can give informed consent for sex hormone treatment with partially irreversible effects (5). Puberty suppression (PS) with GnRHa suppresses the production of testosterone, effectively halting the ongoing male puberty (15). This prevents undesired physical changes, such as lowering of the voice and the growth of facial and body hair. Introducing GnRHa early in puberty may also prevent structural changes in the chest that could potentially affect breast development.

In addition to early suppression of testosterone production, other treatment factors might also impact breast development. One could hypothesize that the use of GnRHa or other anti-androgens prior to the start of estradiol may have a positive effect on breast development. Furthermore, the use of a gradually increasing dose of estradiol in trans girls might also be beneficial for breast development as compared to the direct start with an adult estradiol dose in trans women. In girls with Turner syndrome, a rapid increase of the estrogen dose during pubertal induction has been reported to have a negative impact on breast development (16).

Research on breast development in trans girls is scarce compared to research in adult trans women. Hannema et al described breast development in transgender girls using Tanner stages (6). However, breast volume was not examined. Furthermore, several studies describe the efficacy and safety of GnRHa (17), but studies on the differences in outcome between early vs late initiation are limited (8). Understanding the role of the timing of GnRHa in breast development could improve care for transgender women.

With this study we aimed to identify the impact of the timing of PS on breast development and satisfaction with breast development in transgender girls. Furthermore, the influence of a phase of androgen suppression prior to estradiol and the tempo at which the estradiol dose was increased on breast development was evaluated by comparing breast volume and satisfaction between trans women who had PS from early puberty, from late puberty, or who had no PS prior to GAHT. We hypothesized that mimicking physiological female puberty with an early initiation of PS and a gradual dose increase of estradiol will result in larger breast size.

Methods

Participants

In this study, participants from 2 distinct study cohorts were included. One consists of trans women who initiated hormonal treatment during adolescence, from whom cross-sectional data were collected. The other cohort consists of trans women who started treatment during adulthood, from whom baseline data from a clinical trial were used.

Individuals were eligible for the cross-sectional cohort if they had an age of ≥ 16 years, had started PS before the age of 18 years, and had received estrogen treatment for at least 2 years and no more than 10 years. Exclusion criteria were physical or mental disabilities, participation in other studies, and deviation from the treatment protocol such as temporary cessation of hormonal medication or use of progesterone. Furthermore, individuals who had undergone breast augmentation were not included. However, if they approved, data on age at start of treatment and estradiol dose were collected from their file and they were asked about their bra cup size before augmentation. Eligible trans women with an appointment with the clinician or a mental health professional at the outpatient gender identity clinic in Amsterdam were invited to participate in this cross-sectional study. The included participants were divided into 2 groups based on pubertal stage according to Tanner at start of PS (18). Participants who had started PS at Tanner Genital (G) stage 2 or 3 were categorized as “early PS” and participants with Tanner stage G4 and G5 as “late PS.”

For the adult group, baseline data from the first 20 individuals who participated in a separate clinical trial and approved to the use of their data in other studies were also used in this study (19). They met the following inclusion criteria: age > 18 years at start of GAHT, a minimum of 2 years and a maximum of 10 years of estradiol therapy, and gonadectomy performed. Exclusion criteria were breast augmentation (17%), a body mass index (BMI) > 30 kg/m², and the previous or current use of progesterone. This group had not received PS prior to GAHT.

Treatment

Treatment was initiated after the diagnosis of gender dysphoria was confirmed by a mental health professional according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth and later Fifth Edition (20, 21). Participants were treated according to the Endocrine Society clinical guideline (5). In adolescents, treatment was initiated with PS which consisted of subcutaneous or intramuscular triptorelin (Decapeptyl-CR 3.75 mg every 4 weeks or Pamorelin 11.25 mg every 10-12 weeks). Estradiol supplementation was started from the age of 15 to 16 years. For early pubertal starters, oral 17ß-estradiol was initiated at a dose of 5 µg/kg/day and increased every 6 months by 5 µg/kg/day until reaching the adult dosage of 2 to 6 mg/day. Late pubertal starters had a starting dose of 1 mg/day and increased every 6 months to the adult dosage of 2 to 6 mg/day. Some individuals were temporarily given a high dose of estradiol (6 mg) or 100 to 200 µg ethinyl estradiol (EE) in order to reduce their growth. GnRHa treatment was continued alongside the estrogen supplementation, unless a gonadectomy was performed after the age of 18 years. In the adult starters, treatment consisted of androgen suppression with cyproterone acetate (25-50 mg/d) or spironolactone (50-100 mg/3 times daily) and oral (4-6 mg/d) or transdermal (100 µg twice/week) estradiol initiated simultaneously; androgen suppression was discontinued after gonadectomy.

Data Collection

Questionnaires

After their scheduled appointment in the outpatient clinic, participants filled out 2 questionnaires using CASTOR EDC (Version 1.6). Their satisfaction with their breasts was queried using questions derived from the questionnaire used by De Blok et al (10). The questionnaire used a 4-point Likert scale (very dissatisfied, somewhat dissatisfied, somewhat satisfied, very satisfied) to answer the following questions: How do you rate the 1) size, 2) shape, and 3) symmetry of your breasts and the 4) size and 5) shape of your nipples? For the early and late pubertal group, an additional question (which could be answered with yes, no, or maybe) was: 6) Would you like to have a breast augmentation?

Self-esteem was evaluated with the questionnaire by Rosenberg (22), which was translated to Dutch by Linden et al (23). Self-esteem was rated as low (0-14 points), average (15-24 points), or high (25-30 points).

Measurements

Besides measurements of height and weight, a bioelectrical impedance analysis (BIA) was performed using a Tanita MC-780 MA (Tanita Europe B.V., Amsterdam, The Netherlands) to determine fat percentage.

Laboratory measurements

Serum estradiol concentrations measured on the day of the study visit were collected if available. If unavailable, laboratory results from 6 months prior to or following the study visit were utilized. Serum estradiol concentrations were measured using liquid chromatography tandem mass spectrometry (VUmc, Amsterdam, The Netherlands) with an interassay coefficient of variation of <7% and a lower limit of quantitation of 20 pmol/L.

3D scan

A 3D image was generated of the breasts and thorax using the Artec Leo scanner. The multidimensional image was analyzed using the following 3D software systems: Artec Studio 15 professional (Version 15.1.2.60), GOM Inspect Pro Software 2021, and Autodesk Meshmixer (Version 3.5.474). As described by Kovacs et al (24), the breast volume was determined by establishing the breast borders and generating a posterior wall on the breast. First, the natural shape of the breast was mapped using the folds of the breast as a guideline and removed from the chest. Secondly, the remaining gap in the chest was closed using computer software which followed the natural shape of the pectoralis major muscle. This then doubles as the posterior wall of the removed breast, which enables to calculate breast volume. Cup sizes were determined based on breast volume and categorized as cup ≤ A (< 150 mL), B (150-299 mL) and C (300-449 mL).

Statistical Analyses

Statistical analyses were performed using STATA version 15.1 (StataCorp, College Station, Texas, USA). Data are presented as mean with SD when distributed normally and as median with interquartile range (IQR) in case of non-normal distribution. The mean breast volume of both breasts was calculated and used in further analyses. Linear regression analyses were used to estimate the effect of timing of puberty suppression on different aspects of breast development. The analyses for breast characteristics were adjusted for total body fat percentage as a continuous variable.

Ethics

The study protocol was assessed by the Ethical Review Board of the Amsterdam UMC, VU University Medical Center Amsterdam who determined that the Medical Research Involving Human Subjects Act (WMO) does not apply to this study. The clinical trial in adult subjects was approved by the Ethical Review Board of the Amsterdam UMC, VU University Medical Center Amsterdam. Written informed consent was acquired from all participants.

Results

Between July 2021 and April 2022, a total of 95 trans women who started treatment during adolescence were approached to participate in this study. Twelve individuals (13%) had undergone breast augmentation and 21 individuals refused to participate. Eleven individuals could not be reached. Other individuals could not participate due to physical or mental disabilities (n = 3), temporary cessation of hormonal medication (n = 3), participation in other studies (n = 2), and progesterone use (n = 1). A total of 42 individuals participated in the study of whom 2 were excluded due to poor scan quality. Of those included, 23 initiated PS at Tanner stage G2 or G3 (early PS group) and 17 subjects initiated treatment at Tanner stage G4 or G5 (late PS group). Additionally, data from 20 participants who were included in the clinical trial in adults between March 2021 and June 2021 and who agreed to the use of their data in other studies were used.

Participant Characteristics

Participant characteristics are presented in Table 1. As expected, the early PS group was younger at start of PS and GAHT and duration of monotherapy with GnRHa was 1.5 years (95% CI, 1.0 to 1.9) longer in this group compared to the late PS group. Age at inclusion was comparable between the early and late PS group but the adult starters were 18.7 years older (95% CI, 14.5 to 22.9) compared to the early group. Duration of GAHT at time of inclusion was 4.5 ± 1.7 years and comparable in the 3 treatment groups. Compared to the early PS group, subjects in the late PS group and adult group had a higher BMI by 4.6 kg/m² (95% CI, 2.0 to 7.4) and 6.6 kg/m² (95% CI, 4.2 to 9.1) respectively and a higher fat percentage by 9.0% (95% CI, 4.4% to 13.5%) and 9.2% (95% CI, 4.7% to 13.6%), respectively. BMI and fat percentage were comparable between the late PS and adult group.

Breast Volume

The median breast volume was 115 mL (IQR, 68 to 203) which corresponds to a cup size < A. Figure 1 shows the distribution of cup sizes in the 3 treatment groups. The difference in breast volume between the early PS group compared to the late PS group was −47 mL (95% CI, −104 to 9) but this difference was only −4 mL (95% CI, −75 to 67) after correction for fat percentage. Breast volume in the adult group was comparable to early PS (27 mL; 95% CI, −28 to 83) and late PS (−20 mL; 95% CI, −80 to 40). After correction for fat percentage and duration of GAHT, differences were −36 mL (95% CI −104 to 31) between adult and early PS and 41 mL (95% CI −103 to 20) between adult and late PS. There was no relation between breast volume and serum estradiol concentrations (−0.5 mL per 10 pmol/L [95% CI, −2.0 to 0.9]).

Figure 1.

Cup sizes of individuals in the early PS, late PS, and adult group in percentages of the total number of participants in the group.

Open in new tabDownload slide

A total of 13 subjects were treated with EE 100 to 200 µg in order to limit adult height for a median of 1.6 years (IQR, 0.9 to 1.8) of whom 10 started PS in early puberty. Breast volume was 81 mL (95% CI, −138 to −23) smaller compared to those who were not treated with EE. After correction for fat percentage, this difference slightly decreased to 71 mL (95% CI, −133 to −8).

Because this effect of EE might have confounded the analysis of effect of treatment group, breast volume was also compared between the 3 treatment groups after exclusion of individuals treated with EE, showing no statistically significant differences (early vs late, −25 mL [95% CI, −97 to 47]; early vs adult, 6 mL [95% CI, −61 to 73]). After correction for fat percentage, the differences remained not statistically significant (early vs late, 9 mL [95% CI, −70 to 88]; early vs adult, 53 mL [95% CI −33 to 129]).

Satisfaction and Self-Esteem

The majority of the participants (68%) were satisfied with the size of their breasts. The satisfaction scores for all treatment groups are shown in Fig. 2. Figure 3 shows breast sizes per satisfaction score. There was no relation between breast volume and satisfaction.

Figure 2.

Satisfaction with breast volume in the early PS, late PS, and adult group.

Open in new tabDownload slide

Abbreviation: PS, puberty suppression.

Figure 3.

Breast volume by satisfaction with breast size and treatment group (indicated with white, gray, and black circles for early PS, late PS, and adult groups).

Open in new tabDownload slide

The perceived self-esteem score was 19.2 ± 5.0 and was classified as average. The early PS group scored 2.9 points higher than the late PS group (95% CI, 0.2 to 5.6). Self-esteem in adult starters was comparable to both the early PS and late PS group (adult vs early, −1.2 [95% CI −4.2 to 1.9]; adult vs late, 1.7 [95% CI −1.5 to 5.0)]. Breast volume and satisfaction with breast size had no relation with self-esteem.

Breast Augmentation

Twenty percent of the participants from the adolescent group (4 early PS and 4 late PS) wanted to have breast augmentation in the future and 43% (10 early and 7 late PS) indicated they might want breast augmentation. The desire for breast augmentation surgery was comparable in both groups. A correlation was found between the desire for breast augmentation and breast volume, as the women who wanted breast augmentation had a smaller breast volume by 85 mL compared to the women who did not want surgery (95% CI, −162 to −9).

A total of 12 individuals who were approached for the cross-sectional study were excluded because they had undergone breast augmentation, of whom 7 (58%) had initiated PS at Tanner stage G4 or G5. They had received estradiol for a median of 2.7 years (IQR, 2.0 to 3.7) before the surgery was performed. The majority of the subjects had a cup size A or smaller (n = 9) before surgery. There was no difference in cup size between those who had started PS in early vs late puberty. Cup sizes were comparable to the 40 participants who were included in the early and late PS groups.

Discussion

This study is the first to describe breast volume in transgender women who were treated with GnRHa prior to estradiol treatment. Our hypothesis stated that early initiation of PS would have a positive effect on breast development. However, we found no difference in breast volume between the trans women who started GnRHa treatment in early puberty compared to those who started in late puberty or in adulthood.

The mean breast volume was 115 mL, corresponding to a cup size < A which is similar to findings from a previous study that reported a mean breast volume of 100 mL in 69 transgender women (10). In contrast to our hypothesis, transgender women who started GnRHa treatment in late puberty showed a trend toward a larger breast volume by 47 mL compared to those who started in early puberty. However, after correction for fat percentage, this difference was no longer present. These findings refute the hypothesis that prolonged pubertal exposure to testosterone contributes to the relatively modest breast growth reported in transgender women. The finding that breast volume of the PS groups was comparable to that in the adult starters also suggests that early testosterone blocking has no impact on breast volume.

The 13 individuals treated with EE had a smaller breast volume compared to those treated with estradiol. This suggests an unfavorable effect on breast volume of supraphysiologic doses of estrogen in the first phase of hormonal treatment, or it might indicate a negative effect of EE specifically, as has been suggested for breast development in Turner syndrome (16). The fact that breast volume in the adult group, who immediately started with a full maintenance dose of 2 to 6 mg estradiol, was similar to that in the PS groups in whom a more gradual dose increase is used, argues against an important role of the tempo of estradiol dose increase in determination of breast volume. However, other aspects that may be affected, such as breast shape, were not assessed in the current study.

The current study has not revealed an explanation for the relatively low breast volume observed in transgender women. Reduced Tanner breast stage compared to women in the general population and reduced satisfaction with breasts has also been reported in women with XY differences of sex development (DSD), Turner syndrome, and XX congenital adrenal hyperplasia, although breast volume was not assessed (25). Those who needed pubertal induction were found to have lower Tanner breast stage than those with spontaneous puberty, suggesting that perhaps current hormone replacement strategies are not optimal to induce breast development. In addition, it has been suggested there may be a critical period for breast development, supported by the finding that in women with DSD, Tanner breast stage was negatively associated with age at diagnosis (25). Estrogen treatment in the current study started beyond the age of physiological puberty in girls in the general population. Furthermore, the possible role of sex chromosomes or prenatal testosterone exposure on breast development deserves further study, for example by comparing breast volume in women with various forms of DSD. In addition to estradiol, progesterone is critical for breast development during female puberty; estradiol predominantly controls the expansion of breast tissue and progesterone stimulates differentiation of the mammary tissue (19). Whether progesterone could be important for breast development in feminizing GAHT is under debate. The absence of progesterone exposure might have contributed to the limited breast volume increases observed in this study, regardless of the use of GnRHa. Results of an ongoing trial should clarify if progesterone treatment is beneficial for breast growth in trans women (19). However, the large variability in breast volume among trans women, with some having a C-cup or in other studies even an E-cup (10), prompts the question what other factors, such as genetic factors, might affect breast development.

Despite the median cup size < A, two-thirds of the participants were satisfied with the size of their breasts. This might suggest that breast development in itself, can have a lot of value for trans women. However, a significant number of individuals (63%) were considering breast augmentation in the future with those having smaller breast volumes being more likely to contemplate the procedure. Considering that 13% of all individuals approached for this study had undergone breast augmentation, these findings are comparable to those reported in a survey study in 403 trans women where 54% considered breast augmentation and 9% had already had breast augmentation (26).

Although the mean perceived self-esteem score was average in all groups, the early PS group scored slightly higher by 2.9 points. As self-esteem is a multifactorial outcome and differences in breast volume were marginal, it is understandable that breast volume or satisfaction did not influence self-esteem. It is, however, likely that secondary sex characteristics like voice deepening were more pronounced in the late PS group, which may have contributed to a marginally lower self-esteem score. The fact that trans women from the adult group did not have a lower self-esteem score could be due to their older age, as prior research has shown that self-esteem increases significantly up to the age of 30 (27).

A strength of this study is the 3D technique used to calculate breast volume instead of using the less reliable breast and chest circumferences. This technique is fast, noninvasive, and provides the most objective calculation of breast characteristics (28, 29). Another strength is the inclusion of 3 different treatment groups of similar number of participants and with a comparable GAHT treatment duration, with breast volume and satisfaction assessed in the same manner. The significantly higher age at inclusion of the adult group is not expected to influence breast volume.

A limitation of this study is the possible selection bias due to the cross-sectional design of the study. Firstly, 13% of the approached trans women had undergone a breast augmentation and were therefore excluded. However, inquiries about their pre-augmentation cup sizes revealed similar results to those of the participating individuals. Unfortunately, cup size prior to breast augmentation was not available for those excluded from the clinical trial but based on the relatively low number of exclusions due to breast augmentation of 17%, we estimate that the risk of significant selection bias is small. Second, one can argue that those who do not feel satisfied with their body may be less likely to participate in this study. However, only 22% of the approached individuals refused to participate. A future prospective study on breast volume and satisfaction would be useful to confirm our findings.

To conclude, early PS and gradual estradiol dose increase do not appear to result in a larger breast volume after 5 years of GAHT compared to a later start of PS and more rapid estradiol increase, but supraphysiologic doses of EE may lead to reduced breast volume. Generally, transgender women express satisfaction with their breast size; however, those with smaller breasts are more inclined to consider breast augmentation. Since breast volume represents only one of several breast characteristics, future studies should focus on the effects of PS and rate of estradiol dose increase on breast shape and positioning.

Funding

None.

Disclosures

No disclosures.

Data Availability

The data set generated during and analyzed during the present study is not publicly available because of privacy regulations.

References

Abbreviations

 
• BMI

  body mass index

 
• EE

  ethinyl estradiol

 
• GAHT

  gender-affirming hormone therapy

 
• GnRHa

  gonadotropin-releasing hormone agonist

 
• IQR

  interquartile range

 
• PS

  puberty suppression