Portal de Periódicos da CAPES

Endocrine intervention for transsexuals

2003; Wiley; Volume: 59; Issue: 4 Linguagem: Inglês

10.1046/j.1365-2265.2003.01821.x

1365-2265

Andrew R. Levy, Anna Crown, Russell Reid,

Reproductive Health and Technologies

Gender identity disorder, gender dysphoria, transgenderism and transsexualism are terms used to describe an individual's wish to live and be accepted as a member of the opposite sex. The condition manifests as discomfort with one's phenotypic sex and a wish to have treatment to make one's body correspond to that of the preferred sex. In transsexualism, genotype and somatic differentiation do not conform to brain programming as male or female. The persistent cross-gender identification that results transcends a desire for any cultural advantages of being the other sex. Transsexualism is distinct from sexual orientation, and transsexuals, like nontranssexuals, may be heterosexual, homosexual, bisexual or asexual (Schilder et al., 2001). All strata of society are affected by gender dysphoria (Hoenig & Kenna, 1974) and the similar prevalence of 1 in 13–15 000 males and 1 in 30–35 000 females across Western (van Kesteren et al., 1996a) and Eastern Europe, Singapore and the Indian subcontinent suggests that the influence of culture on the underlying condition is relatively small (van Kesteren, 2002). In Southeast Asia amongst the Hindu and Buddhist population, transsexual individuals often evoke relatively little concern on the part of those affected and their families as the condition is attributed to residues of a previous life as a member of the opposite sex (Stevenson, 1977). In other societies, the challenges that people with gender dysphoria face may transcend those faced by lesbian, gay and bisexual individuals, and they may be at greater risk of harassment and violence which in turn leads to an increased incidence of depression, substance abuse (Seal et al., 2001) and suicide (Lee, 2000; Lombardi et al., 2001; Schilder et al., 2001; Cochran et al., 2002). These risks are further compounded by marginalization from society (Landen & Innala, 2000) and inconsistent legislation with, variously, lack of precedents for changing the sex of birth certification, confused marital rights, criminalization of genital contact between people of the same sex and exclusion of sex reassignment surgery from health service funding and health insurance (Gordon, 1991). In this review, we discuss predisposing factors to transsexualism, and the diagnosis, assessment and hormone treatment of transsexual patients. We also discuss the risks and organ-specific effects of cross-sex hormone treatment. We will briefly summarize the current legal position of transsexuals in the UK and recommendations for the culturally competent delivery of transgender health care. Stable transsexualism in adulthood is a rare co-morbidity of a physical intersex state or genetic abnormality (Meyer et al., 1986) and its ICD10 classification specifically excludes intersex and sex chromosome conditions and serious mental disorders from the diagnosis. A proportion of XY patients who present with genital ambiguity in infancy or early childhood may subsequently become dissatisfied with their sex of rearing (Migeon et al., 2002; Wisneiwski & Migeon, 2002), but interest in intersex states within the transsexual community probably relates more to funding streams than biology and the conceptual need to have a physical condition rather than a more nebulous 'psychological' one. In support of the place of early biological programming (Davenport, 1986; Landen et al., 1998b; Wisniewski et al., 2000), inconsistent reports suggest that increased androgen exposure in utero and early postnatal life experienced, for example, by females with congenital adrenal hyperplasia may result not only in increased male-type posturing, interests and behaviour (Dittmann et al., 1990; Hines & Kaufman, 1994; Berenbaum, 1999) but in an increased prevalence of gender dysphoria in adulthood (Meyer-Bahlburg et al., 1996; Zucker et al., 1996; Hines, 1998). There is also some evidence that antenatal exposure to anticonvulsants, which in animal models alter steroid hormone levels, may enhance the risk of subsequent gender dysphoria in man (Dessens et al., 1999) and the suggestion, confounded to some extent by small numbers available for study and various other forms of bias, that male-to-female transsexuals have female sexually dimorphic brain characteristics (Kruijver et al., 2000). Gender dysphoria is a self-diagnosis, often supported by friends and family, with no supporting tests other than persistence of dysphoria for at least 2 years alleviated by cross-gender identification. It is amenable to hormonal and surgical palliation but is essentially an incurable condition associated in a significant proportion of cases with social isolation and a degree of mental anguish to which even the most judicious medical and surgical management may add physical discomfort and infertility. As the effects of treatment are often irreversible, the process of psychological evaluation to determine whether treatment will facilitate the patients' comfort and integration into society or leave them still incapacitated by social phobias is far-reaching, and involves input from diverse sources such as ethicists, psychologists, speech and beauty therapists, social workers and personal shopping assistants as well as psychiatrists, nurses and various medical and surgical specialities (Monstrey et al., 2001). Ideally a clinic structure that encompasses the gamut of clinical governance mechanisms such as supervision, audit and continual professional development for staff should be available to optimize integration of these specialist areas. The objective of treatment is to rehabilitate the individual as a full member of society in their preferred gender role – the most perfect medical and surgical palliation being of no use 'als ze achter de geraniums blijven zitten' (if they feel compelled to stay behind the geraniums). Widely used (but not necessarily evidence-based) standards of care for patients with gender dysphoria are promulgated by 'The Harry Benjamin International Gender Dysphoria Association' (Benjamin, 1967; Meyer et al., 2001; internet search terms: 'harry benjamin'). These standards of care, regarded as authoritative by providers but as restrictive by many transsexuals, involve a 'real life' trial lasting months or years in which the preferred gender roll is adopted in every aspect of life. If the process is perceived as excessively lengthy (i.e. beyond 1 year) there is a significant risk that unsafe practices, such as buying hormones on the street or internet, or seeking treatment from unscrupulous physicians or unskilled surgeons may be pursued (Lombardi, 2001). For many transsexuals, demonstrating that their preferred gender role improves their social and psychological function in society is the hardest part of the process, as they are particularly exposed to ridicule and discrimination before treatment produces noticeable changes. Assessment by at least two senior specialists is recommended, often involving a number of in-depth interviews over several months during which psychiatric (Levine, 1980) and psychological state, personality profile, physical and emotional development, school and employment background, adult relationships, physical health and current level of social functioning are addressed. The process is not an inevitable escalator of intervention (Levine & Lothstein, 1981) and many who enter the process do not proceed to full gender-reassignment surgery (Landen et al., 2001). Others are content to step in and out of medical management to suit their needs, and protocols to optimize care should be sufficiently flexible to incorporate this tendency in a nonpunitive manner. It is particularly important to allow young people flexibility to question their gender identity (Lombardi, 2001). Within the bounds of confidentiality, an objective view from the biological spouse, close friend or others within an intimate relationship may be sought to help disclose issues contributing to confusion or distress within the person's mind. Families of transsexual patients may also benefit from receiving counselling in their own right, as they too face difficult changes. The positive support of a partner or of older children can be invaluable, but for transsexuals who become alienated from their family in the process of revealing their sexual identity, unofficial 'families of attachment' that replace families of origin, should not be overlooked or ignored (Schilder et al., 2001). The recommendation that in addition to baseline cholesterol, urea and electrolytes, glucose and liver function tests, physical assessment should include a full examination of secondary sexual characteristics needs to be approached with care, as such scrutiny may be particularly traumatic for this group of patients. Aside from potentially disclosing evidence of self-treatment, which may be as much to do with delay, desperation and lack of information as a genuine desire to take matters into their own hands, clinical examination, like measurement of sex hormone levels, is unlikely to yield anything more than confirmation of biological sex. A clear explanation of the irreversible effects of hormone therapy on body habitus and formal consent is a necessary precaution. There are very few well validated efficacy data for different treatment regimens, a void that is filled, perhaps understandably, by case reports and internet autobiography which should be recognized as such and appraised with appropriate care. Recommendations in this review are derived from published sources and personal experience (RR) of managing over 3000 transsexual patients over the last 20 years. Female-to-male. For female-to-male transsexuals, the usual hormone treatment is Sustanon 250 mg intramuscularly (i.m.) every 2 weeks, or testosterone enanthate (Primoteston Depot) if the patient is sensitive to peanut (arachis) oil. Male-to-female. Optimal hormone treatment for male-to-female transsexuals is guided by the development of the desired mental changes and by the onset and maintenance of a feminized physical state that the male-to-female transsexual finds acceptable. The effects of different hormone preparations are as variable in XY females as in XX females in terms of breast development, mood changes and vasomotor stability. Treatment often requires modification depending on perceived efficacy and the patient's beliefs, which are almost invariably informed by the many excellent trans gender internet sites and local support networks (internet search terms: 'gender trust', 'press for change', 'gires', 'looking glass society', 'renaissance transgender', 'mermaids', 'ftm'). A typical regimen is Estraderm TTS 50 or 100 (delivering 50 or 100 µg oestradiol/24 h transdermally when applied twice weekly, reduced to 50 µg/24 h postgender-reassignment surgery), oral ethinylestradiol 100–150 µg daily in divided doses (50 µg twice a day or three times a day, reducing to 50 µg daily postsurgery) or oral oestradiol valerate (Progynova) 4–6 mg daily (2 mg twice a day or three times a day, reducing to 1–2 mg postgender-reassignment surgery). Whilst well tolerated, the relatively short action of oestradiol valerate (the pharmacological equivalent of natural 17-β-oestradiol) is sometimes less effective in producing adequate feminization. As insufficient oestrogen and possibly also excessive oestrogen may be associated with vasomotor symptoms, adjustment of the dose in both directions may be necessary. In some centres, spironolactone 100–300 mg daily (Prior et al., 1989) or cyproterone acetate 50–100 mg daily (van Kesteren et al., 1997) is routinely added to transdermal or oral oestrogen treatment, whilst in others, the addition of cyproterone is thought to confer a risk of idiosyncratic severe hepatotoxicity while adding relatively little to the antiandrogenic effects of oestrogens alone. Many patients, however, seem to appreciate the further feminizing effects of the combination of oestrogen and cyproterone acetate and will, even if this is not observed objectively, surrender to the psychological compulsion to do everything in their power to feminize. Although there is no absolute requirement for cyproterone or any other progestogenic drug, some male-to-female transsexuals find that their libido is maintained and breast growth augmented by the addition of dydrogesterone (such as Duphaston 10 mg twice a day) or the slightly more virilizing medroxyprogesterone acetate (Provera 5–10 mg daily). Continuation of cyproterone treatment after gender reassignment surgery may continue to discourage further body hair growth. The 5-α reductase inhibitor finasteride 2·5–5 mg orally daily and local minoxidil solution 2–5% (Regaine) are occasionally used if the person is distressed by scalp hair loss and finds hair pieces difficult to tolerate. The low intrinsic morbidity and mortality of the relatively young population that tend to be treated for gender dysphoria, the comparatively small numbers studied and the inevitable lack of placebo controlled, prospective trials, complicates the distinction between the risks of gender dysphoria itself (van Kesteren et al., 1997; Kenagy, 2002) and the risks attributable to cross-sex hormone treatment. In a retrospective study of 303 male-to-female transsexuals treated with cross-sex hormones, the death rate due to increased numbers of suicide and death of unknown cause was five times that of a reference population (Asscheman et al., 1989). Depressive mood changes were 15 times more common and thromboembolic events were increased 45-fold compared to the general population (Asscheman et al., 1989). In the same study, androgen treatment in 122 female-to-male transsexuals was associated with weight increase and acne. In both groups it was not clear whether persistent liver enzyme abnormalities should be attributed to hepatitis B and alcohol abuse rather than the effects of cross-sex steroids. However, much of the morbidity was minor and reversible with appropriate therapy or temporary discontinuation of hormone treatment. In a retrospective, descriptive study of 816 male-to-female and 293 female-to-male transsexuals treated with cross-sex hormones, there was no increase in mortality overall or mortality that could be attributed to cross-sex hormone treatment compared to an age and gender adjusted general Dutch population (van Kesteren et al., 1997). In male-to-female transsexuals there was an increased risk of suicide and HIV infection (a ninefold and sixfold increase in risk, respectively: n= 13) but no change or a slight reduction in risks from malignancies (risk halved: n= 7), myocardial infarction [standardized mortality ratio (SMR) 0·7: n= 6], chronic obstructive pulmonary disease and pancreatitis (SMR 0·8: n= 2 and 1, respectively). Morbidity risks in this group relate to hepatitis B (44-fold increased risk: n= 12) and cholelithiasis (14-fold increased risk: n= 8), with a possible increase in risk of stroke (1·7-fold increase: n= 6) and no detected change in risk of prostatic carcinoma, hypertension or myocardial infarction (van Kesteren et al., 1997). As only two deaths were observed in the female-to-male transsexual group (n = 293), relative mortality resulting from male hormone treatment in genetic females remains unclear (van Kesteren et al., 1997). The most significant cross-sex hormone treatment risk relates to spontaneous venous thromboembolism particularly during the first year of oral oestrogen treatment in male-to-female transsexuals. The 2·6% incidence of venous thromboembolism, falling to 0·4% per annum thereafter (Asscheman et al., 1989; van Kesteren et al., 1997) compares to a 0·005–0·01% per annum spontaneous incidence in the general (young) population, associated with a 1–2% case fatality rate from pulmonary embolism (Anderson et al., 1991; Rosendaal, 1997; O'Connell & Perry, 2001). Compared to postmenopausal women who experience a two- to threefold increase in risk of spontaneous, symptomatic venous thromboembolism during current sex hormone replacement therapy (from approximately 0·1% per annum risk over the age of 60 years (Rosendaal, 1997)) this represents a highly significant excess in absolute risk in a much younger population who may in addition be subject to the additional risk of major, pelvic surgical procedures (Daly et al., 1996; Grodstein et al., 1996; Jick et al., 1996; Perez Gutthann et al., 1997). Of the cases of venous thromboembolism reported by van Kesteren et al. (1997) only one of the 45 cases occurred during transdermal rather than oral oestrogen use, and in this case the patient had previously suffered a thromboembolism while taking oestrogens orally (van Kesteren et al., 1997). This potential treatment route-related difference in risk has not yet been confirmed by other studies (Perez Gutthann et al., 1997). Concurrent thrombophilic mutations, particularly factor V Leiden and antithrombin deficiency substantially increase the risk of venous thromboembolism conferred by oestrogen treatment (Vandenbroucke et al., 1994). The absolute risk, however, remains relatively low and may decrease with duration of use (Lidegaard et al., 1998). The risk should also be offset against the potential hazards of inadequately treated gender dysphoria, and morbidity associated with sex hormone deficiency such as hot flushes and osteoporosis. Furthermore, the risk of warfarin treatment in a young, well-informed population with low co-morbidity is relatively low. Screening for deficiencies of protein C, protein S, antithrombin III and factor V Leiden may theoretically be useful to reduce the thromboembolism risk but is more likely to result in transsexuals being needlessly deprived of the treatment that they require and is not generally recommended (van Kesteren, 2002). The recommendation that oestrogen (at least in oral contraceptive formulations for XX women) should be stopped before elective major surgery are at odds with guidance from several specialist groups (Thromboembolic Risk Factors (THRIFT) Consensus Group, 1992; RCOG Working Party, 1995; Anonymous, 1997) and as might be expected, there are no data available for specific thromboembolic risks of major elective surgery in male-to-female transsexuals taking oestrogen at high dose. Male-to-female. Histomorphometric studies of trans-ilial bone biopsies from 23 male-to-female transsexuals (mean age 38 years) following 8–14 months' treatment with ethinylestradiol 100 µg/day and cyproterone acetate 100 mg/day did not show any differences in bone volume or surface, trabecular thickness, osteoclast number or eroded surface between transsexuals and controls (Lips et al., 1989). Reduced osteoid volume, surface and thickness in these subjects suggests that oestrogen and antiandrogen treatment in male-to-female transsexuals may suppress bone turnover but is not associated with bone loss (Lips et al., 1989). In a study of 20 male-to-female transsexuals exposed to cross-sex hormones for 28–63 months, oestrogen treatment (ethinylestradiol 100 µg daily with cyproterone acetate 100 mg/day (50–100 µg oestradiol without cyproterone after gonadectomy)) prevented bone mineral density loss induced by testosterone privation (van Kesteren et al., 1998). Further small studies of 28 preoperative, male-to-female transsexuals (Reutrakul et al., 1998) and 10 male-to-female transsexuals (Schlatterer et al., 1998) produced similar results, suggesting that oestrogen treatment for 2 years or more maintains or increases bone mineral density compared to untreated healthy adult male controls. Female-to-male. A bone histomorphometric study of 15 female-to-male transsexuals (mean age 30 years) who had undergone hysterectomy and bilateral ovariectomy, after an average of 39 months treatment with 250 mg testosterone i.m. every 2 weeks, showed intact trabecular bone structure and increased cortical thickness with low bone turnover indices compared to 11 healthy men and eight postmenopausal women (Lips et al., 1996). A further small study of 10 female-to-male transsexuals also failed to identify any adverse effects of long-term, cross-sex hormone treatment after native sex hormone reduction (Schlatterer et al., 1998). The benign nature of oestrogen reduction and testosterone replacement on bone mineral density is not a universal finding, and a study of 35 female-to-male transsexuals reported after 1 years' treatment (van Kesteren et al., 1996b), 19 of whom were followed up for 28–62 months (van Kesteren et al., 1998) suggests that testosterone [Sustanon 250 mg i.m. every 2 weeks (switched to testosterone undecanoate – Restandol – in seven of 19 injection-intolerant people postgonadectomy)] is unable to prevent bone mineral loss associated with a decline in oestrogen levels. This lack of efficacy at the bone interface is at odds with the beneficial effects of more modest male hormone supplements in hypogonadotrophic genetic males (Behre et al., 1997). As individual changes in bone mineral density are highly variable and the numbers studied generally small, long-term data on osteoporosis risk in the female-to-male transsexual group and further evidence to confirm the potential place of LH measurement as a marker of adequate hormone substitution (van Kesteren et al., 1998) are very much needed. Increase in breast size usually begins 2–3 months after the start of female sex hormone treatment in male-to-female transsexuals and continues for 2 years (Meyer et al., 1986; van Kesteren, 2002). Unfortunately, only one-third of transsexuals achieve more than a B cup and with 45% not advancing beyond an A cup, at least 60% require breast augmentation to achieve the appearance they desire or at least find acceptable. Breast development seems to be more pronounced in subjects with higher body mass indices and it may be worth suggesting to slender male-to-female transsexuals that they do not make too stenuous efforts to avoid the modest oestrogen-induced gain in weight that is often experienced (van Kesteren, 2002). Various attempts to hasten transition to the preferred gender have been reported (Kay & Saad, 1983; Wylie, 2000), the most common being ingestion of higher quantities of hormones than prescribed. Sudden cessation of oestrogen treatment can be associated with galactorrhoea (van Kesteren, 2002). There are case reports of breast carcinoma in hormonally treated male-to-female transsexuals (Symmers, 1968; Ganly & Taylor, 1995) but none in the van Kesteren et al. (1997) series of 816 hormonally and surgically treated male-to-female transsexuals. In a preliminary study of 142 male-to-female transsexuals treated with oestrogen and cyproterone for from 6 months to 9 years, variable increases in circulating prolactin but no specific risk of inducing autonomous prolactin secretion was found (Gooren et al., 1985). The same group reported a more extensive series of 214 oestrogen- and cyproterone-treated male-to-female transsexuals in which modest increases in prolactin were thought to be universal. In this population prolactin increased to over 1000 mU/l in 20% (Asscheman et al., 1988; 14% in the van Kesteren et al., 1997 series) and remained persistently elevated in 15 (7%). Of these, five (2%) were believed to show pituitary enlargement on computed tomographic (CT) scanning but again, prolactinoma development was not found (Asscheman et al., 1988). Autonomous prolactin secretion in a 26-year-old male-to-female transsexual who in addition to the prescribed 100 µg oral ethinylestradiol daily surreptitiously injected herself with 100 mg oestradiol-17-undecanoate intramuscularly for at least 6 of the 10 months of treatment has been reported (Gooren et al., 1988), and a histologically proven prolactinoma reported in a 33-year-old male-to-female transsexual who had received exogenous oestrogen for the previous 17 years (Kovacs et al., 1994). Given the high frequency of occult prolactinoma formation (Molitch, 1997) and the apparent rarity of prolactinoma in genetic males during cross-sex hormone therapy, a direct link between exogenous oestrogens and prolactinoma induction in humans cannot be drawn. Contrary to early conclusions (Asscheman et al., 1988; Goh & Ratnam, 1990), safety data about adverse pituitary effects of oestrogens in genetic males are reassuring. Female sex hormones adversely affect fertility in most XY individuals (Thiagaraj et al., 1987; Venizelos & Paradinas, 1988; Lubbert et al., 1992; Handelsman et al., 2000). Particularly for male-to-female transsexuals who may be sexually orientated towards women post-treatment, there is an argument that the same options offered any other patient who risks losing their germ cells should be made available to this group (Brothers & Ford, 2000; De Sutter, 2001). As far as freezing unfertilized ova in female-to-male transsexuals before treatment, the chance of a successful pregnancy resulting from the use of such ova is currently believed to be very low. The rate of development, extent and potential reversibility of fertility reduction in female-to-male transsexuals receiving testosterone is also unclear at the present time. Exogenous androgen in female-to-male transsexuals may induce changes to ovarian histology reminiscent of polycystic ovarian syndrome (Amirikia et al., 1986). It is not known whether these and associated metabolic changes are analogous to those seen in sporadic polycystic ovarian syndrome, given differences in mean body mass index, glucose tolerance and degree of androgen excess, or whether the potential risks associated with one are applicable to the other (Schildkraut et al., 1996; Solomon, 1999). As the association between polycystic ovarian disease and the risks of endometrial and ovarian malignancy are not entirely clear (Pierpoint et al., 1998; Balen, 2001), recommendation for hysterectomy and bilateral salpingoophorectomy, generally carried out around 15 months after the start of testosterone treatment in female-to-male transsexuals (van Kesteren et al., 1998), has yet to be fully justified, at least on the grounds of enhanced risk of malignancy. Individual cases of prostate cancer in male-to-female transsexuals have been reported (Thurston, 1994; van Kesteren et al., 1996c; van Haarst et al., 1998). The extent of potential reduction in risk associated with female hormone treatment and androgen suppression is unknown. Transient elevation of liver enzymes that cannot be attributed to any other cause is found in approximately 3% of male-to-female transsexuals treated with oestrogens and persists for more than 6 months in half of these (van Kesteren et al., 1997). Case reports of hepatomas (Paradinas et al., 1977) and spontaneous rupture of a liver cell adenoma (Bird et al., 1979) have been reported in female-to-male transsexuals treated with methyltestosterone, a form of male hormone no longer used. Plasma levels of homocysteine, believed to be a modest (Homocysteine Studies Collaboration, 2002) or perhaps more significant (Graham et al., 1998) independent cardiovascular and cerebrovascular risk factor (Haltmayer et al., 2002) are reduced in hormone-treated male-to-female transsexuals and increased in androgen treated female-to-male transsexuals (Giltay et al., 1998). In practice, the additional risk of heart disease in people treated with cross-sex hormones is not known but available data suggest that there is not a substantial increase (van Kesteren et al., 1997). Androgen treatment in female-to-male transsexuals results in the induction of facial hair growth and increased sebum production that is in many cases evident within 4 months and continues to develop beyond one year (Giltay & Gooren, 2000). In male-to-female transsexuals, reduction in facial and truncal hair shaft diameter reaches a maximum after 4 months treatment with cross-sex hormones but does not progress further. Skin sebum production falls rapidly to almost undetectable levels but hair growth in length remained largely unaffected or responds only very slowly (Giltay et al., 2000). Thus male-to-female hair reduction is almost invariably inadequate with hormonal treatment alone and depends on physical means. Electrolysis is effective but uncomfortable, potentially scarring (particularly if the pain induces flinching) and often too slow to be practical as a sole treatment. Laser hair removal is more rapid, provided hair colour is dark (and skin colour is not), and becomes more comfortable as successive treatments reduce the density of hair shafts and hence the area of 'burn'. For both, sedatives, simple analgesics and the preapplication of anaesthetic creams such as Emla or Ametop 1–5 h before treatment may be required. Small amounts of remaining hair can be rubbed away with proprietary disposable abrasive pads (Silkymits©) and body hair treated with mechanical depilation, waxing, sugaring or shaving. Effective treatment of genital hair is also required prior to penoscrotal inversion vaginoplasty, but for this procedure, subcutaneous local anaesthetic nerve blocks or even a sitting spinal with heavy bupivacaine may be required. A variety of cosmetics, some applied in two parts, can provide excellent camouflage until such time as the persons perception of their appearance becomes acceptable (internet search terms: 'cosmetic camouflage' or 'dermablend'). Magnetic resonance imaging (MRI) analysis of regional fat deposition in 20 male-to-female transsexuals before and 1 year after cross-sex hormone treatment demonstrates a significant increase in subcutaneous and visceral fat depots and a decrease in thigh muscle area (Elbers et al., 1999). In 17 female-to-male transsexuals a transient (Elbers et al., 1997b) reduction in subcutaneous fat and increase in thigh muscle area with increased visceral fat is seen (Elbers et al., 1999). Adipose tissue changes in male-to-female transsexuals, particularly increased fat around the upper thighs, is often insufficient to confer true female habitus and if the profile remains disturbing, surgical intervention may be required. The sex differences in circulating leptin levels, with women having higher levels than males irrespective of body habitus, are reversed by cross-sex hormone treatment (Elbers et al., 1997a). Androgen treatment produces an improvement in spatial ability in female-to-male transsexuals that persists during treatment and for several weeks after treatment withdrawal (Slabbekoorn et al., 1999). In a study of 35 female-to-male transsexuals, aggression and sexual motivation were increased by testosterone treatment (van Goozen et al., 1995). In 15 male-to-female transsexuals examined in the same study, oestrogen treatment reduced libido and the proneness of recipients to anger and aggression (van Goozen et al., 1995). In normal young men, oral levonorgestrel-induced reduction in circulating testosterone (and oestrogen) over 8 weeks adversely affected verbal memory (Cherrier et al., 2002). Many male-to-female transsexuals report that oestrogen treatment is associated with a calming, almost antidepressive effect (Reid, 2002) and there is some support for the use of oestrogens alone as mood modulators in perimenopausal and postmenopausal women, and to augment the action of antidepressants (Halbreich & Kahn, 2001; Grigoriadis & Kennedy, 2002). Male hormone replacement in elderly men may also be associated with mood modulation (Seidman & Walsh, 1999) but the potential role of testosterone as an antidepressant is less well studied and the literature less persuasive. In a group of 30 female-to-male transsexuals, mean clitoral length of around 4–4·5 cm was achieved 1 year after the start of treatment with testosterone cypionate (200 mg every 2 weeks, i.m.; Meyer et al., 1981, 1986). Further increases in androgen dose had no additive effect (Meyer et al., 1981). Since 1966, transsexuals have had full protection from discrimination in the workplace. In July 2002, the European Court of Human Rights held that British transsexuals must be able to change the gender on their birth certificate and to get married (Christine Goodwin vs. The United Kingdom). At the time of writing, British law still has to comply with the ruling. In the meantime, there is a sound legal basis for believing that the courts would follow the decision of the European Court of Human Rights and hold transsexual people as being of their new sex (Dr Stephen Whittle, Reader in Law, Manchester Metropolitan University, personal communication). Transsexuals frequently report adverse experiences of health care, from insensitivity and ignorance to discrimination and hostility. Reported attitudes of doctors range from phobic to judgemental and patriarchal (Lombardi, 2001; Schilder et al., 2001). Recommendations of ways to improve transgender health care centre on acknowledging the transgender individual's identity. Rather than becoming fixated on technical and medical aspects of their care, the 'years of isolation and struggle' experienced by many individuals and the daily stress of living in a stigmatized and marginalized community should be acknowledged. Real regret in people who have undergone palliative treatment for transsexualism is rare (< 0·5%) and, rather than misdiagnosis, usually relates to unfavourable life circumstances such as early marriage, adverse social situation and lack of support (Landen et al., 1998a), poor adaptive skills, late start of treatment and limited intelligence (van Kesteren et al., 1997; Smith et al., 2001). A history of a previously unsuccessful long-term homosexual relationship or sexual abuse in childhood also seems to increase the risk that inappropriate treatment will be demanded (Reid, 2002). Peer pressures within the transsexual community may also push vulnerable male-to-female transsexuals toward surgery to distinguish themselves from transvestites – heterosexual males for whom periodic cross-dressing induces sexual arousal (Frohwirth et al., 1987). With respect to the quality of surgical outcome, in an unpublished survey of around 200 male-to-female transsexuals carried out by 'Press for Change', even those who reported poor outcomes of gender reassignment surgery did not regret their decision to undergo the procedure as the structural reconfiguration rather than the quality of change is one of the desired endpoints, perhaps reinforced by misunderstood anatomy and well-intentioned reassurances (Claire McNab, editor of 'Press for Change', personal communication; McNab, 2002). Satisfaction with sexual activity in both male-to-female and female-to-male transsexuals tends to increase markedly with cross-sex hormone treatment and gender-reassignment surgery even if sexual functioning may be technically imperfect (Lief & Hubschman, 1993). In summary, transsexuals have persistent cross-gender identities, usually without any predisposing factors. Self-diagnosis is confirmed by psychological assessment, which includes a trial period living in the chosen gender before consideration of hormonal treatment and surgery. The choice of cross-sex hormone treatment is generally straightforward and complications, principally venous thromboembolism in male-to-female and potentially osteoporosis in female-to-male, are highly amenable to treatment. Although more evidence would be welcome, adequately treated gender dysphoria is likely to be safer than the untreated condition, which is associated with an enhanced risk of depression and suicide. Reassuringly, few transsexuals regret undergoing treatment. Doubts about the integrity of transgender individuals and the authenticity of gender dysphoria as a diagnosis, lack of approbation from peers and perhaps personal phobias may lead some members of the medical profession to withhold treatment or prescribe inadequate doses of cross-sex hormones on perceived safety grounds. Transsexual individuals require long-term assistance to optimize cross-sex hormone treatment and should not be subject to discrimination when they seek health care. The perception within some parts of the transgender community that the nonpsychiatric medical profession are not interested in their plight should be laid to rest. We would like to thank Ms Melissa Jenkins and Drs Sidney Crown, Andy Black and Stephen Whittle for helpful advice.