© Archives of Sexual Behavior : 10/1998

Authors - Schlatterer, Kathrin; Yassouridis, Alexander; Werder, Klaus von; Poland, Dorette; Kemper, Johannes; Stalla, Gunter K.


Gender dysphoric disorders have been known from antiquity onward across national and cultural boundaries. They have been described in classic literature from Heroditus to Shakespeare (Pauly, 1965; Green, 1966). Historically, no difference was made between transsexualism and transvestism due to the lack of technical opportunities for sexual conversion like hormone application and sex reassignment surgery. Beside the extreme solution of castration, clothing and cosmetics were the only tools available for both groups of individuals. A hormonal sex change became feasible after the discovery of the sex hormones early in the 20th century, and their successful synthesis and commercial manufacture in the 1930s. During the last decades in Western Europe and the United States multistep concepts for the treatment of transsexual patients, in which cross-gender hormone application plays the central role, have been developed and continuously improved. Our strategy for cross-gender hormone treatment was refined during the follow-up of 129 transsexual patients treated in our neuroendocrinological outpatient clinic over the last 5 years (Schlatterer et al., 1996). Of those patients, 88 have been included in the present study. Treatment concepts, psychosocial characteristics, and endocrinological follow-up data are presented. The data show a small incidence of side effects due to our therapeutic strategy.


Sample Population

Transsexual patients were seen by the authors in the Endocrinological Outpatient Clinic of the Max-Planck-Institute between 1991 and 1995 inclusively. During that time 129 patients were referred to us. For the purpose of the present study we reviewed the files of these patients and developed a questionnaire for personal data. The patients were also interviewed. Of the 129 patients treated, 88 (46 M-to-F and 42 F-to-M) provided us with personal information and were therefore included in the present study, 41 patients either refused to take part in the investigation or their addresses could not be ascertained. A complete history for controlled variables and treatment schedules was obtained besides the personal data. Physical examinations were performed regularly.

Cross-Gender Hormone Therapy

The aim of cross-gender hormone therapy in transsexual patients is an assimilation of secondary sex characteristics to the desired sex as quickly as possible by administration of decreased doses of specific hormone recombinants in consecutive time intervals (steps) after beginning therapy. For this purpose F-to-M transsexuals were treated with a 250-mg depot of testosterone applied intramuscularly. Injection intervals varied between 2 weeks in the beginning to 3 to 4 weeks later on. Optimal lifelong therapy was designed individually by regular screening of serum testosterone, while serum estrogen levels were decreasing. Cross-gender hormone therapy for M-to-F transsexuals is more complex. Testosterone synthesis was decreased effectively by application of the antiandrogen cyproterone acetate. In the beginning of therapy we administered a daily dose of 100 mg orally, slowly adapting it to falling serum testosterone levels. Estrogens usually were administered in a two-phase regimen. High-dose pharmacological estrogen was given in the beginning of therapy as an intramuscular depot, generally every 2 weeks. Optimal individual injection intervals were defined according to patient's risk factors, side effects of therapy, and serum hormone levels in intervals of between 1 and 3 months. As soon as secondary sex characteristics had fully developed, therapy was changed from the high-dose application to a lifelong low-dose estrogen substitution therapy. This also has to be carefully monitored and modified if necessary. Antiandrogen therapy at this point is no longer necessary. For more information concerning cross-gender hormone therapy forms, its risks and side effects, see Schlatterer et al., 1996.

Statistical Analysis

Since the majority of the variables considered in the study are of nominal or categorial data structure, analysis of contingency tables was basically used for evaluating the data at hand. Besides the frequency distributions of transsexuals within the levels (categories) of the various variables, tests of significance of the dependence of transsexuality and certain variables based on the chi-square statistic were performed. (Note: dependent upon the values of the cell frequencies the p values of the chi-square statistic were calculated either exactly or approximately with Monte Carlo simulations.) A nominal level of significance [Alpha] = 0.05 was accepted. Various tests of independence (or tests of homogeneity) were performed at a reduced level of significance (Bonferroni correction), in order to keep the Type I error less or equal to 0.05.


Before dealing with the effects of cross-gender replacement therapy on endocrinological findings the psychosocial background, the medical and drug background as well as the age distribution within each transsexuality group was investigated. We proposed to compare our results with those obtained by other studies.

Endocrinological and Side Effects During Cross-Gender Hormone Therapy

The clinical data of our transsexual patients are presented in Table IV (not shown). For 12.5% of our transsexual patients, cross-gender hormone therapy at the time of completion of this study was no longer performed. The number of F-to-M transsexuals with no continued therapy was more than the M-to-F ones (M-to-F: 6.5%, F-to-M: 19.04%). The rest (approximately 80%) of our F-to-M transsexual patients received an intramuscular testosterone application therapy (250 mg every 2-3 weeks). Among the 46 M-to-F patients 32% (i.e., 15 patients) were still in the high-dose pharmacological phase of treatment and received a combination therapy of high dose estrogens together with antiandrogens (estradiol 40-100 mg im every 2 weeks together with cyproterone acetate 10-100 mg daily) whereas 4% (i.e., 2 patients) received high-dose estrogen therapy with estradiol 80-100 mg every 2 weeks alone. Of the M-to-F patients 22% still needed antiandrogen therapy when reducing the estrogen doses to 2-8 mg estradiol daily. In 30% of these patients testosterone serum levels had dropped so far that a combination therapy was pointless. They were administered estradiol 2-8 mg daily alone. For different reasons a minority of 2% of our M-to-F transsexual patients received therapy with natural, unconjugated estrogens. The same number of patients was administered a combination of 2 mg cyproterone acetate and 35 microgram ethinylestradiol.

Under the hormone therapy 26% of the M-to-F transsexuals showed no side effects, whereas the number of F-to-M transsexuals without side effects was significantly higher (42%). Although some of our patients showed more than one side effect, we present in Table IV (not shown) the absolute frequency for the incidence of single side effects only. The most common side effect observed by the M-to-F patients was the development of hyperprolactinemia: 24 M-to-F transsexuals showed this symptom and in 4 we also found transient elevated levels of prolactin. These 4 patients performed mechanical compression of their breasts. The nonpersisting elevation of prolactin levels could be traced to this manipulation procedure of the breast. The incidence for transient hyperprolactinemia with normalizing levels of prolactin after dose adjustment was in the range of those found in studies already performed for estrogen-treated M-to-F transsexuals. On the other hand we detected no prolactinoma as described by other authors (Asscheman et al., 1988, 1989; Kovacs et al., 1994; Gooren et al., 1980). The portion of patients developing galactorrhea (5/40 [congruent] 13%) was lower in our study than the corresponding one found by Futterweit (1980). None of our patients developed deep vein thrombosis or embolism during cross-gender hormone therapy performed in our clinic. These side elects are the most severe ones during estrogen therapy and have been seen in many patients (Fortin et al., 1984; Lehrman, 1976). One of our M-to-F patients suffered from lilac vein thrombosis following surgery, another from deep vein thrombosis before starting therapy. Here it was not clear if the patient self-administered high doses of estrogens without medical control. One patient suffered from lung embolism before starting therapy. Here self-administration of estrogens was also difficult to evaluate. One patient suffered from severe varicosis. Therefore in his case we administered very low doses of estrogens under permanent control of blood-clotting parameters. This therapeutic compromise was performed as an exception, because the patient, who proved to be a serious and reliable partner for hormone therapy, insisted on estrogen administration. A closer examination of blood-clotting parameters as a risk assessment for thromboembolic complications (data not shown) revealed for 11 patients 1 single case of pathologically altered parameters under high-dose estrogen therapy. For this purpose we analyzed the prothrombin time, partial thromboplastin time (PTT), antithrombin III, protein C antigen, functional protein C, protein S antigen and APC-resistance.

Interestingly cross-gender hormone therapy had an influence on erythropoesis. In 15 examined estrogen-treated M-to-F transsexuals we found a decrease of hemoglobin compared to levels of women. In 5, F-to-M transexual patients testosterone treatment induced an increase of hemoglobin to levels normally seen in biological men . These changes correspond to the effects of androgens on erythropoesis as described by Kennedy and Gilbertsen (1957). This finding has been tried with varying success as a therapeutic approach for different diseases (Fried et al., 1973, Alexanian, 1969; de Gowin et al., 1970). In view of the erythropoetic effect of androgens, chronic respiratory disorders like emphysema and bronchial asthma are relative contraindications to cross-gender hormone therapy, particularly in heavy smokers.

In the group of F-to-M transsexuals, 3 patients suffered from persistent bleeding which ceased after the application of a high-dose gestagen between the testosterone applications. Three patients also reported concentration and/or sleep problems. Development of acne was seen in 4 patients. In F-to-M transsexuals, acne is one of the most common side effects observed, which often has to be treated with antibiotics (Schlatterer et al., 1996). In the group of M-to-F transsexuals, 2 patients suffered from severe headaches accompanying estrogen medication. For both patient groups (M-to-F and F-to-M transsexuals) a transient increase of liver enzymes was observed, similar to that described by (Meyer et al., 1986; Asscheman et al., 1989). After further diagnostic procedures (screening for hepatitis B and C antigens, ultrasonography of the liver), followed by adjusting and reducing sex hormone doses, these changes in transaminases were no longer detected.

Long-term follow-up studies of cross-gender hormone-treated transsexual patients have not been performed. Interesting objectives would be a systematic evaluation of prevalence of neoplasia and the investigation of an influence of cross-gender hormones on the cardiovascular system. Single case reports of breast cancer in M-to-F transsexuals have been reported previously (Pritchard et al., 1988; Symmers, 1968). Effects of estrogens as part of oral contraceptives on the cardiovascular system are already known (Stadel, 1981; Hannaford et al., 1994; Glashan and Robinson, 1981; Biller and Saver, 1995; Goh et al., 1995; Damewood et al., 1989; de Marinis and Arnett, 1978).


In the last 5 years we have established a cross-gender hormone substitution model for our endocrinological outpatient clinic, embedded in a multistep treatment concept for the transsexual patient (Schlatterer et al., 1996). Here we present data, summarizing our experiences with this therapy. We are by no means certain that our sample of transsexual patients is complete and representative enough to carry out reliable epidemiological calculations. Our findings therefore should be regarded as estimates for an overview of the patient's personal background, endocrinological findings, and the outcome of sex reassignment.

Comparing our findings with others published so far, we first evaluated the psychosocial background (age distribution, marital status, number of children, occupational status, nicotine and alcohol consumption, family background) of our patients as well as the patient's anamnesis. For the most part psychosocial variables of the two groups of our transsexual patients did not differ significantly (see Table I-III – not shown), but these results compared to those obtained by other studies reveal some discrepancies. With regard to marital status (Table I – not shown) the two groups did not show homogeneity in the frequency distribution ([[Chi].sup.2]-test, p [less than] 0.05). Considerably more M-to-F than F-to-M transsexuals live in marriage, but here also the rate of divorce is higher. Data to date present controversial findings for this feature. Our findings confirm the reports of Hoenig and Kenna (1973) and Kockott and Fahrner (1988). Concerning occupational status, the F-to-M patients show employment patterns similar to M-to-F transsexuals. There are some differences in the frequencies of the single occupational status levels, but their values did not reach statistical significance: 8.7% of our M-to-F patients have already retired. The unemployment rate for F-to-M transsexual patients in our study is higher than that for M-to-F transsexuals. This is in contrast to the findings of Tsoi (1992). In the group of F-to-M transsexuals more patients were still in school, apprenticeship, or visited university than in the group of M-to-F transsexuals. Tsoi (1990) has described for Singapore a lower incidence for M-to-F transsexuals to be in higher occupational classes than F-to-M, due to the fact that many M-to-F transsexuals take up service and entertainment jobs which can be graded as skilled or semiskilled.

One parameter that differs significantly from the studies published is the number of siblings. In our sample as many transsexuals have siblings as those having none, in contrast to Dixen et al. (1984), who published an incidence of being the only child of approximately 12%. In the transsexual patients' parents a high history of psychiatric disorders has been described by Dixen et al., which could not be confirmed by us. We found a weak occurrence of endocrinopathies, cardiovascular problems, neoplastic, and psychiatric-neurological disorders in the parents. Approximately 50% of the M-to-F transsexuals and 30 % of the F-to-M transsexuals show further disorders (see Table II – not shown). Endocrinopathies and psychiatric problems are the most frequent disorders, followed by diseases affecting the cardiovascular system, dermatological disorders, and chronic infectious diseases. The relatively high incidence of psychiatric history is consistent with the literature and can be interpreted as evidence of an extreme dissatisfaction that the patients experience in their current, unaccepted gender (Fleming et al., 1981; Pauly 1974).

Previous studies showed a significant number (50%) of associated endocrinopathies in F-to-M transsexuals (Futterweit, 1980). The incidence of chronic infectious diseases like hepatitis B and C as well as HIV infection might possibly be related to the sexual behavior these patients show before their disorder is accepted and treated (same-sex partners, sometimes in a homosexual environment with an increased incidence of sexually transmittable infectious diseases). To a minor extent other disorders have been observed. Approximately 18% of our patients, independent of gender, regularly took other drugs beside cross-gender hormones, analgesics, psychopharmaceutics, and endocrinological agents are the most frequent. Slightly fewer F-to-M than M-to-F transexuals smoke, whereas the alcohol consumption in F-to-M transsexuals is significantly higher.

The two transsexual groups showed significant discrepancies both in the age of diagnosis and in the age at beginning cross-gender hormone therapy ([[Chi].sup.2]-test, p [less than] 0.05). Most of the patients are diagnosed as transsexual between the age of 21 and 30 years, independently of the biological gender. About 16% of M-to-F transsexuals are diagnosed older than 41 years (Table III – not shown). Many of the patients were referred to our clinic by the psychiatrist, followed by the neurologist, the general practitioner, and the internist. As many patients are member of patients' organizations as not.

Hormone replacement therapy was started in our clinic with the same age distribution. The kind of cross-gender hormone therapy was adjusted according to the side effects observed. The incidence of hyperprolactinemia we found in estrogen-treated F-to-M transsexuals (Table IV – not shown) lies in the range of studies published before (Asscheman et al., 1988, 1989), whereas the number of patients developing galactorrhea was significantly lower in our patients. None of our patients suffered from a prolactin-producing pituitary adenoma. The incidence of thromboembolic events during cross-gender hormone treatment in our patients was zero. Changes in haematological parameters were observed under cross-gender hormone therapy. Transient rises in transaminases occurred at a similar frequency as described by other authors (Asscheman et al., 1989; Meyer et al., 1986).

The follow-up of these patients for completed sex reassignment surgery revealed an incidence of problems due to surgery of approximately 20%. Wound healing and urological problems proved to be the most frequent. M-to-F transsexuals were affected slightly more than F-to-M patients, despite better surgical chances for this group. Less than 5% of our transsexual patients refused any surgical intervention. These numbers were independent of biological sex.

With this follow-up study we have been able to demonstrate a low incidence of severe complications occurring due to a specific cross-gender hormone replacement therapy in 88 transsexual patients. Long-term follow-up studies have to be carried out to evaluate further risks of cross-gender hormone replacement therapy like the possible development of neoplasia or long-term effects leading to cardiovascular diseases. Cases of ischemic cerebrovascular diseases accompanying infertility therapy or cross-gender hormone replacement therapy, as performed in transsexual patients, have been reported (Biller and Saver, 1995). Influences of estrogen and testosterone therapy on lipid/lipoprotein profiles are also described (Goh et al., 1995; Damewood et al., 1989). The design and realization of such studies could help to further improve therapy strategies for transsexuals.


  • Alexanian, R. (1969). Erythropoietin and erythropoiesis in anemic man following androgens. Blood 33: 564-572.
  • Asscheman, H., Gooren, L. J. G., Assies, J., Smits, J. P. H., and de Slegte, R. (1988). Prolactin levels and pituitary enlargement in hormone treated male-to-female transsexuals. Clin Endocrinol 28: 583-588.
  • Asscheman, H., Gooren, L. J. G., and Eklund, P. L. E. (1989). Mortality and morbidity in transsexual patients with cross-gender hormone treatment. Metabolism 38: 869-873.
  • Biller, J., and Saver, J. (1995). Ischemic cerebrovascular disease and hormone therapy for infertility and transsexualism. Neurology 45: 1611-1613.
  • Damewood, M. D., Bellantoni, J. J., Bachorik, P. S., Kinball, A. W. Jr., and Rock, J. A. (1989). Exogenous estrogen effect on lipid/lipoprotein cholesterol in transsexual males. J. Endocrinol Invest. 12: 449-454.
  • de Gowin, R., Richard, L., Lavender, A. R., Forland, M., and Charleston, D. (1970). Erythropoietin and erythropoiesis in patients with chronic renal failure treated with hemodialysis and testosterone. Ann. Intern. Med. 72: 913-918.
  • de Marinis, M., and Arnett, E. N. (1978). Cerebrovascular occlusion in a transsexual man taking mestranol. Arch. Intern. Med. 138: 1732-1733.
  • Dixen, J. M., Maddever, H., van Maasdam, J., and Edwards, P. W. (1984). Psychosocial characteristics of applicants evaluated for surgical cross-gender reassignment. Arch. Sex Behav. 13: 269-276.
  • Fleming, M., Cohen, D., Salt, P., Jones, D., and Jenkins, S. (1981). A study of pre- and postsurgical transsexuals: MMPI characteristics. Arch. Sex Behav. 10: 161-170.
  • Fortin, C. L., Klein, T., Messmore, H. L., and O'Connell, J. B. (1984). Myocardial infarction and severe thromboembolic complications as seen in an estrogen-dependent transsexual. Arch. Intern. Med. 144: 1082-1083.
  • Fried, W., Jonasson, O., and Lang, G. (1973). The hematological effect of androgen in uremic patients. Ann. Intern. Med. 79: 823-827.
  • Futterweit, W. (1980). Endocrine management of the transsexual. N.Y. State J. Med. 80: 1260-1264.
  • Glashan, R. W., and Robinson, M. R. G. (1981). Cardiovascular complications in the treatment of prostatic carcinoma. Br. J. Urol. 53: 624-627.
  • Goh, H. H., Loke, D. F. M., and Ratnam, S. S. (1995). The impact of long-term testosterone replacement therapy on lipid and lipoprotein profiles in women. Maturitas 21: 65-70.
  • Goodwin, W. E., and Cummings, R. H. (1984). Squamous metaplasia of the verumontanum with obstruction due to hypertrophy: long-term effects of estrogen on the prostate in an aging male-to-female transsexual. J. Urol. 131: 553-554.
  • Gooren, L. (1990). The endocrinology of transsexualism: Review and commentary. Psychoneuroendocrinology 15: 3-14.
  • Gooren, L. J. G., van der Veen, E. A., and van Kessel, H. (1980). Modulation of prolactin secretion by gonadal steroids in men. In: MacLeod, R. M., and Scapagnini, U. (eds.), Central and Peripheral Regulation of Prolactin Function, Raven Press, New York, pp. 373-375.
  • Green, R. (1966). Mythological, historical, and cross-cultural aspects of transsexualism. In Benjamin H. The Transsexual Phenomenon. The Julian Press, New York.
  • Hannaford, P. C., Croft, P. R., and Kay, C. R. (1994). Oral contraception and stroke. Stroke 25: 935-942.
  • Hoenig, J., and Kenna, J. (1973). Epidemiological aspects of transsexualism. Psychiat. Clin. 6: 65-80.
  • Kennedy, B. J., and Gilbertsen, S. (1957). Increased erythropoesis induced by androgenic hormone therapy. New. Engl. J. Med. 256: 719-726.
  • Kockott, G., and Fahrner, E. M. (1988). Male-to-female and female-to-male transsexuals: A comparison. Arch. Sex Behav. 17: 539-546.
  • Kovacs, K., Stefaneanu, L., Ezzat, S., and Smyth, H. S. (1994). Prolactin-producing pituitary adenoma in a male-to-female transsexual patient with protracted estrogen administration. Arch. Pathol. Lab. Med. 118: 562-565.
  • Lehrman, K. L. (1976). Pulmonary embolism in a transsexual man taking diethlstilbestrol. J. Am Med. Assoc. 235: 532-533.
  • Meyer, W. J., Webb, A., and Stuart, C. A. (1986). Physical and hormonal evaluation of the transsexual patient. A longitudinal study. Arch. Sex. Behav. 15: 121-138.
  • Pauly, I. (1965). Male psychosexual inversion: Transsexualism. Arch. Gen. Psychiat. 13: 172-181.
  • Pauly, I. B. (1968). The current status of the sex change operation. J. Nerv. Ment. Dis. 147: 460-471.
  • Pauly, I. B. (1974). Female transsexualism I and II. Arch. Sex Behav. 3: 487-526.
  • Pritchard, T. J., Pankowsky, D. A., Crowe, J. P., and Abdul-Karim, F. W. (1988). Breast cancer in a male-to-female transsexual. A case report. J. Am. Med. Assoc. 259: 2278-2280.
  • Schlatterer, K., von Werder, K., and Stalla, G. K. (1996). Multistep treatment concept of transsexual patients. Exp. Clin. Endocrinol. (in press).
  • Stadel, B. V. (1981). Oral contraceptives and cardiovascular disease. New Engl. J. Med. 305: 612-618, 672-677.
  • Symmers, W. C. (1968). Carcinoma of the breast in trans-sexual individuals after surgical and hormonal interference with the primary and secondary sex characteristics. Br. Med. J. 2: 83-85.
  • Tsoi, W. F. (1990). Developmental profile of 200 male and 100 female transsexuals in Singapore. Arch. Sex. Behav. 19: 595-605.
  • Tsoi, W. F. (1992). Male and female transsexuals: A comparison. Singapore Med. J. 33: 182-185.


comments powered by Disqus