Pelvic Pain in Transgender People Using Testosterone Therapy

Pelvic pain in transgender and gender diverse (herein referred to as trans) people presumed female at birth, who are using testosterone as gender-affirming hormone therapy (GAHT), is poorly understood. 1 Understanding adverse effects of testosterone therapy is important, given that trans people comprise an estimated 0.5%–4.5% of the adult population. 2–4 Furthermore, there is an increasing demand for gender-affirming health care globally. 5–7 Testosterone GAHT is very effective at inducing masculinizing physical changes, including significant genital and reproductive system effects, an increase in body and facial hair, deepening of the voice, increase in muscle mass, and a decrease in fat mass. 8 Menstrual cessation, one of the most desired aspects of testosterone GAHT, typically occurs within the first 6 months of therapy, although breakthrough bleeding is not uncommon. 9 Clitoral enlargement, vulvovaginal atrophy, and increase in libido are also frequently observed. 10 , 11 Endometrial changes can be varied regardless of menstrual cessation with either proliferative (in 40%) or atrophic endometrium (in 50%). 12 No significant histopathological change appears to occur in the ovaries. 13 Pelvic pain in people with a uterus and ovaries is extremely common in the general population with 17%–81% reporting dysmenorrhea, 8%–22% reporting dyspareunia, and 2%–24% reporting noncyclical pain. 14 Chronic pelvic pain persisting beyond 6 months affects 1 in 7. 15 Causes are multifactorial and rarely reflect a single pathological process. 16 There is a considerable economic burden on people experiencing chronic pelvic pain and on health care systems worldwide. 17 Diagnosis and management can be challenging and require an individualized approach. 18 As clinicians (gynecologists, endocrinologists, and physiotherapists), we have seen increasing numbers of trans individuals on testosterone seeking assistance to relieve symptoms of pelvic pain. However, there is a paucity of data regarding pelvic pain in trans individuals using testosterone GAHT. 1 , 19 Given that estradiol deficiency may lead to atrophic vaginitis, we hypothesized that pelvic pain in trans people using testosterone would be predominantly lower abdominal and would worsen with symptoms of genital dryness or penetrative sexual activities. Furthermore, we hypothesized that pre-existing endometriosis, vulvodynia, or vaginismus would be risk factors. Given the limited research on the prevalence and/or the characteristics of pelvic pain experienced by individuals using testosterone GAHT, this was an exploratory study aiming to identify the characteristics of pelvic pain in trans people using testosterone GAHT and to explore potential factors associated with experiencing pelvic pain after commencing testosterone GAHT.

Of the 670 trans people presumed female at birth and using testosterone GAHT, who were invited to complete the survey, a total of 506 responded. After removing duplicates and incomplete responses, 486 valid responses remained. Detailed characteristics are outlined in Table 1 . The median age of the 486 respondents was 27 years (23–34 years). Twenty respondents were between 50 and 67 years of age (presumed postmenopausal). Twenty (4.1%) of the respondents identified as Aboriginal or Torres Strait Islander, 7 (1.4%) reported having a variation of sex characteristics (intersex), and a further 45 (9.3%) indicated they did not know if they had a variation of sex characteristics. The median duration of testosterone GAHT use was 32 months (14.0–60.5 months) with intramuscular testosterone undecanoate injections, the most used formulation. Characteristics of the Study Sample Multiple responses allowed for this question so total responses do not sum to 100%. A total of 351 (72.2%) of the study sample experienced pelvic pain after starting testosterone therapy. Of those 351 respondents, 316 (90%) reported pelvic pain “sometimes” and 35 (10%) reported pelvic pain “always or almost always.” Of the 20 respondents older than 50 years and presumed postmenopausal, 9 (45%) reported pelvic pain “sometimes” and 2 (10%) reported pelvic pain “always or almost always.” A majority of the 351 participants ( N  = 345, 98.3%) reported some form of pelvic pain before starting testosterone therapy. This included 190 (65.5%) who “always or almost always” and 89 (30.7%) who “sometimes” experienced pain around menstruation, 48 (14.3%) who “always or almost always” and 102 (30.5%) who “sometimes” experienced assumed ovulation pain, and 31 (9.3%) who “always or almost always” and 140 (41.9%) who “sometimes” experienced pain between menstrual periods. The most common description of pain was cramping (described by 72.6%), followed by aching (58.1%), stabbing (39.9%), and sharp (33.9%). The pain was most commonly located in the hypogastric region (described by 87.2% of respondents) ( Fig. 1 ). The median score for pain severity after commencement of testosterone GAHT on a scale from 0 to 10 (most severe pain) was 6.2 (4.0–7.7). This was similar to the median score of 6.7 (5.5–7.8) reported for pain severity around menstruation before commencement of testosterone GAHT. Consistent with this, the median score in response to the question “How does the pelvic pain you experience using testosterone compare to the pelvic pain you experienced before starting testosterone?” (0 = much less severe, 5 = about the same to 10 = much more severe), was 4.3 (2.2–6.8). Diagram of abdominal and pelvic regions ( N  = 351). Location of pelvic pain selected by number of respondents describing pelvic pain since starting testosterone therapy. Multiple responses allowed for this question so total responses do not sum to 100%. Multivariable regression demonstrated that there were higher odds of reporting pelvic pain after starting testosterone in people also experiencing persistent menstruation (OR = 4.46 [1.33–14.97]), or who had a current or previous diagnosis of PTSD (OR = 2.50 [1.07–5.85]) ( Table 2 ). Pelvic pain was also positively associated with pain with orgasm (OR = 32.72 [10.65–100.52]). A total of 11.3% ( N  = 39) reported that pelvic pain “always or almost always” resulted in them stopping sexual activity or made them consider alternative ways/methods of being sexually active, and 39.2% ( N  = 135) reported this was “sometimes” the case. Factors Associated with Pain After Starting Testosterone Therapy “Unsure/Prefer not to say” was treated as missing and excluded from analysis. Bold values denote statistical significance at the p  < 0.05 level. Odds ratio (95% CI) and p -values from logistic regression mutually adjusted for all factors presented in the table. CI, confidence interval; IQR, interquartile range; PTSD, post-traumatic stress disorder. The most reported treatments used for relieving pelvic pain included analgesic medication “pain killers” (56.7%, N  = 199), with nonsteroidal anti-inflammatory drugs (NSAIDs; Ibuprofen, Diclofenac and Aspirin) and paracetamol reported most as being helpful ( Table 3 ). Of the small number of participants ( n  = 26) who had undergone hysterectomy, pelvic pain was the most common indicator for surgery (61.5%), followed by gender dysphoria (46.2%). A majority (72%) reported reduction in pain, while 20% reported little to no change in their pelvic pain and two reported an increase in pelvic pain following hysterectomy. However, given the small sample, inferences are limited. Treatments for Pelvic Pain in People Using Testosterone Multiple responses allowed for this question so total responses do not sum to 100%. IUD, intrauterine device.

Pelvic pain after initiation of testosterone therapy was reported by 72.2% of trans people responding to this online survey. Cramping pain in the suprapubic (hypogastric) regions were the most common descriptors. Factors associated with increased odds of reporting pain with testosterone were persistent menstruation, pain with orgasm, and current or previous diagnoses of PTSD. NSAIDs were most commonly used to relieve pain. Of the small number of participants who had undergone hysterectomy, a majority reported reduction in pain. Consistent with our hypothesis and the only other published study describing pelvic pain in trans people using testosterone GAHT, ∼70% of respondents reported pelvic pain with the most common description as cramping, most commonly in the hypogastrium or suprapubic region, followed by right and left iliac regions. 1 Although our exploratory study cannot determine causation or mechanisms of pain, the increased likelihood of reporting pain in people with persistent menstruation and orgasm may suggest increased pelvic floor muscle dysfunction. 20 It is known that levator ani, the collective group of pelvic floor muscles consisting of the deeper layer of pubococcygeus, the iliococcygeus, and the puborectalis, as well as superficial perineal muscles such as the bulbospongiosus, are enriched with androgen receptors and exquisitely androgen sensitive in male humans, and in male rodents. 21 Androgen-receptor knockout mice, and men who have androgen deprivation therapy for prostate cancer have a marked reduction in the size of the levator ani muscle. 22 , 23 Androgens have been shown in female humans and rats to have anabolic effects on pelvic floor muscles. 24–26 As such, it is plausible that testosterone GAHT may affect the pelvic floor and perineal muscle, and therefore may contribute to the experience of pain. While pain from endometriosis has been described to worsen with orgasm and penetrative sexual activity 27 and could explain pelvic pain after commencing testosterone therapy, previous research has suggested that this is unlikely. 19 A review of 67 people who had gender-affirming hysterectomy, among whom 51% had reported pelvic pain, found intraoperative endometriosis in 32% of those who reported pain as well as in 22% of those without pain. 19 Our analysis did not show an association between having a diagnosis of endometriosis, vaginismus, vulvodynia, or genital dryness and pelvic pain. Participants who reported persistent menstruation had significantly higher odds (OR = 4.46) of reporting pain after starting testosterone. Individuals who do have persistent vaginal bleeding would generally be continuing to have typical rises and falls in estradiol and progesterone concentrations over the course of a menstrual cycle. Pelvic pain in people with persistent menstruation may arise from factors such as dysmenorrhea from myometrial contractions, the release of inflammatory mediators that cause endometrial shedding, or pelvic floor muscle dysfunction. Menstruation with fluctuations in estradiol and testosterone levels have been demonstrated to affect pelvic floor muscle activity with significantly higher levels of muscle tone in the luteal phase relative to the follicular and ovulatory phases. 28 Sexual dysfunction among trans people has been reported to be very common, likely compounded by increased sexual desire after commencing testosterone therapy. 29 However, there is little published research. 30 Those who had pain with orgasm had significantly elevated odds (OR = 32.72) of reporting pelvic pain after starting testosterone GAHT. Pain with orgasm or dysorgasmia is one of the least understood and poorly studied areas in sexual medicine, involving a complex interplay of psychological, neural, vascular, and endocrine factors. Dysorgasmia may be the result of bladder neck contractions, uterine neuroinflammation, uterine contractions, and/or pelvic floor musculature dystonia. 31 , 32 As such, treatment for sexual pain would be best tailored to the individual with a multidisciplinary approach involving gynecology, physical therapy, pain management, sexual therapy, and mental health professionals who specialize in chronic pain. 33 Further research is needed on etiologies and evaluating multimodal approaches. There is a known clear association between chronic pain and PTSD in people of all genders, explained by both genetic and environmental factors. 34 , 35 We observed that people who had a current or previous diagnosis of PTSD had higher odds (OR = 2.50) of reporting pelvic pain after starting testosterone GAHT. Research has previously shown that in cisgender women with chronic pelvic pain, there is an increased prevalence of abuse experiences, high number of major life events, and diagnosis of PTSD (but not depression). 36 , 37 One study in 107 cisgender women suggests that temporally, PTSD appears to precede the diagnosis of chronic pelvic pain supporting a partial role for PTSD or its trauma-related trigger in the pathophysiology of chronic pelvic pain. 38 Notably, previous studies in people presumed to be female have suggested an independent association between chronic pelvic pain and anxiety, depression and mixed anxiety, and depressive disorder. 39 This was not observed in our analyses, which potentially may be related to the high prevalence of depression and anxiety among trans people overall. 40 Treatment of pelvic pain can be challenging in the general population. 18 A multidisciplinary biopsychosocial approach that addresses contribution of various factors to the individual is needed. 41 This may include medical therapies, pelvic floor physical therapy, addressing sexual function, hypersensitivity to pain, and psychological factors such as PTSD. 41 Respondents to this survey had reported various self-management strategies. Over-the-counter pain-relieving medications in the form of paracetamol, NSAIDs, and heat were the most frequent strategies reported to manage pelvic pain. In a Cochrane review of management of dysmenorrhea, NSAIDs and heat were recommended as first-line treatment to alleviate pain symptoms produced by the release of prostaglandins from the endometrial lining. It is recommended that these strategies are initiated 48 hours before onset of menses. 42 Irregular bleeding and amenorrhea may explain the ineffectiveness of these strategies in this population, with limited warning of the onset of breakthrough bleeding episodes. 9 Many trans people seek hysterectomy and/or oophorectomy as part of gender affirmation, or due to pelvic pain or ongoing or abnormal bleeding. Of the individuals in this study who had a hysterectomy, 72% reported relief of pelvic pain symptoms after hysterectomy. While the overall number of respondents undergoing a hysterectomy and/or oophorectomy was too small for meaningful statistical analyses, surgery would indeed cure persistent menstruation, which was much more likely in people reporting pain after commencing testosterone therapy. Moreover, cisgender women have also reported resolution or decrease in pelvic pain following hysterectomy. 43 It must be noted that some individuals in this study reported little to no change, or an increase in pelvic pain following hysterectomy. Further research is warranted. While further studies need to evaluate the possibility of high pelvic floor muscle tone as a causative factor for pelvic pain in trans people after starting testosterone for gender affirmation, a recent systematic review of pelvic floor physical therapy to release myofascial trigger points found positive beneficial effects, particularly in people with chronic pelvic pain and dyspareunia. 44 This systematic review did not specifically include studies involving trans people, but did include both men and women with a range of conditions, including dyspareunia, provoked vulvodynia, interstitial cystitis, painful bladder syndrome, chronic prostatitis, or chronic pelvic pain syndrome. 44 Given the lack of current treatments available to alleviate often debilitating pelvic pain in trans people on testosterone therapy, pelvic floor physical therapy may be a low-risk treatment strategy. 41 A pelvic floor muscle down-training program, which focuses on the quality of the muscle function and relaxation phase of the contraction can be particularly helpful in this clinical setting. This study has multiple limitations. As this was an online survey recruited through nonprobability sampling, this may have encouraged a greater proportion of responders who were younger individuals and may not be representative of the broader trans community. Of the 670 people participating in the larger TRANSform study, who indicated they were using testosterone therapy and were invited to participate in this testosterone and pain study, 486 responded, corresponding to a response rate of 72.5%. Given that potential participants were invited to participate in a study titled “Pain experiences in trans men and trans masculine people using testosterone survey,” there may well have been responder bias, with individuals experiencing pain syndromes more likely to respond and overrepresenting the proportion of individuals on testosterone experiencing pelvic pain. Furthermore, as participants were asked to recall experiences from before commencing testosterone, there is the possibility of recall bias. Medical conditions were self-reported, and we were unable to confirm diagnoses. We also acknowledge that, although a number of participants indicated that they had or were unsure about whether they had a variation of sex characteristics (intersex), participants were not asked whether they had a uterus and ovaries as part of their birth anatomy. In addition, questions regarding penetrative sexual activities did not specify vaginal or anal penetration. Standardized questionnaires for sexual dysfunction were not included, as these are not validated for trans and gender diverse populations. The impact of testosterone therapy on sexual function warrants further investigation. Despite the limitations, this survey is the largest study to date exploring pelvic pain in trans people using testosterone therapy and is hypothesis generating for future studies examining the pathophysiology of, or the effectiveness of interventions on pelvic pain.

Pelvic pain occurring after commencing testosterone GAHT is frequently reported by trans people. The increased likelihood of reporting pain in people with persistent menstruation and orgasm, as well as the known androgen sensitivity of the pelvic floor musculature, warrant further research on pelvic floor muscle dysfunction as a contributor. 20 Until further evidence is available, a tailored multidisciplinary trauma-informed approach addressing the needs of the individual with pelvic pain should be provided, which may encompass pain management, sexual function, addressing persistent menstruation, and mental health.

Deidentified participant data are available upon reasonable request from the corresponding author through email ( [email protected] ), provided that the related research is deemed to be of benefit to the trans and gender diverse community, and has undergone Austin Health Human Research Ethics Committee approval in the form of an amendment.

Participants in this study were recruited from a larger longitudinal Australian trans health study known as TRANSform . Inclusion criteria for TRANSform were assessed by three screening questions: (a) currently living in Australia; (b) identification as trans (“is your gender different to what was presumed for you at birth?”); and (c) 16 years of age or older. Participants were recruited using a nonprobability snowball sampling approach with recruitment calls posted on social media (Facebook and Instagram) and shared widely by trans and gender diverse community support groups and organizations in Australia. A total of 670 TRANSform participants who indicated that they were using testosterone therapy for gender affirmation were emailed an individualized link to a survey titled “Pain experiences in trans men and trans masculine people using testosterone survey.” This online cross-sectional survey was open between August 28, 2020, and December 31, 2020. Written informed consent was not obtained; however, the survey preamble outlined that completing the survey implied consent. The survey was designed collaboratively by our core team of researchers (S.Z., A.F.Q.W., T.C., and K.E.), who are members of the Australian trans community, and clinicians specialized in trans health care. Survey data were collected and managed using REDCap electronic data capture tools hosted at The University of Melbourne. The study was completed in accordance with the Declaration of Helsinki as revised in 2013 and received ethical and governance approval by the Austin Health Human Research Ethics Committee (reference No. HREC/57155/Austin-2019), ACON Research Ethics Review Committee (reference No. 2020/03), and the Thorne Harbour Health Community Research Endorsement Panel (reference No. THH/CREP 20-006). A small participation incentive (AUD$5 gift card) was provided for completion. Survey questions are outlined in detail in the Appendix . In brief, demographic data, testosterone formulation, dosage, duration of use, and self-reported testosterone concentrations were obtained. Participants were asked to describe characteristics and location of pelvic pain and rate severity, as well as compare the presence of pelvic pain before and after commencing testosterone therapy for gender affirmation. Potential associated factors were explored, including persistent menstruation; presence of genital dryness; history of hysterectomy or oophorectomy; presence of pain with sexual activities; use of intrauterine device; and known diagnoses of depression, anxiety, post-traumatic stress disorder (PTSD), endometriosis, vulvodynia (pain in the area around the vulva, not necessarily with touch), or vaginismus (involuntary tightening of the muscles around the vagina, not necessarily with penetration). The number of pregnancies (including miscarriages and terminations) and number of live births were also determined. Participant characteristics are reported as frequency (percentage) for categorical variables, and median (interquartile range) as appropriate for not normally distributed data. Logistic regression was used to estimate the effects of possible factors contributing to pain on the odds of experiencing pain after starting testosterone. The factors considered in the regression were selected before performing the analysis based on potential risk factors for pelvic pain from expert opinion (given the lack of published research in this field). Results are reported as odds ratios (OR) with corresponding 95% confidence intervals. This is a complete case analysis with an alpha level of 5% ( p  < 0.05) considered statistically significant. Statistical analyses were performed using R version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria).