Association between polycystic ovary syndrome and female-to-male transsexuality

Transcription

1 Human Reproduction Vol.22, No.4 pp , 2007 Advance Access publication December 13, 2006 doi: /humrep/del474 Association between polycystic ovary syndrome and female-to-male transsexuality Tsuyoshi Baba 1,5, Toshiaki Endo 1, Hiroyuki Honnma 1, Yoshimitsu Kitajima 1, Takuhiro Hayashi 1, Hiroshi Ikeda 2, Naoya Masumori 3, Hirofumi Kamiya 4, Osamu Moriwaka 4 and Tsuyoshi Saito 1 1 Department of Obstetrics and Gynecology, 2 Department of Psychiatry, 3 Department of Urology, Sapporo Medical University, Sapporo, Hokkaido, Japan and 4 Kamiya Ladies Clinic, Sapporo, Hokkaido, Japan 5 To whom correspondence should be addressed at: Department of Obstetrics and Gynecology, Sapporo Medical University, South 1 West 16, Chu-o-ku, Sapporo, Hokkaido , Japan. tbaba@sapmed.ac.jp BACKGROUND: The aim of this study is to understand the relationship between polycystic ovary syndrome (PCOS), altered hormonal characteristics and insulin resistance in female-to-male (FTM) transsexual patients. METHODS: We studied 69 Japanese FTM cases, aged years, who were seen in the Gender Identity Disorder Clinic of Sapporo Medical University Hospital between December 2003 and May The subjects had never received hormonal treatment or sex re-assignment surgery. Prior to treatment, they received physical examinations entailing measurement of anthropometric, metabolic and endocrine parameters, after which we compared the values obtained according to the presence or absence of PCOS and/or obesity. Insulin resistance was determined using the homeostasis model assessment of insulin resistance (HOMA-IR). RESULTS: Of the 69 participating FTM cases, 40 (58.0%) were found to have PCOS. Of the 49 for whom HOMA-IR was calculated, 15 (30.6%) also showed insulin resistance, whereas of the 59 for whom adiponectin was measured, 18 (30.5%) showed hypoadiponectinaemia. Of 69 for whom androgens were measured, 29 (39.1%) showed hyperandrogenaemia. Insulin resistance was associated with obesity but not with PCOS. In contrast, hyperandrogenaemia was associated with both PCOS and obesity. CONCLUSIONS: FTM transsexual patients have a high prevalence of PCOS and hyperandrogenaemia. Key words: Adiponectin/female-to-male transsexuality/hyperandrogenaemia/insulin resistance/polycystic ovary syndrome Introduction Gender identity disorder (GID) is a disagreement between biological sexual differentiation and self-declared gender identity. It is characterized by (i) a strong and persistent cross-gender identification and (ii) persistent discomfort with the biological sexual or gender role behaviours associated with one s sex. The aetiology of GID remains unclear, although endocrinological, neuroanatomical and psychosocial factors are all thought to be causally involved. For example, Collaer and Hines (1995) showed that gonadal hormones affect sexual differentiation. In addition, examination of females with congenital adrenal hyperplasia due to 21-hydroxylase deficiency, which results in excessive prenatal androgen exposure, revealed that early androgen exposure has a modest effect on gender identification (Berenbaum and Bailey, 2003). We established a GID clinic in 2003, and since then, more than 90 female-to-male (FTM) transsexuals have been seen there. Prior to treatment, most of these individuals exhibited polycystic ovarian morphology on ultrasonography as well as symptoms of hyperandrogenism (e.g. hyperandrogenaemia and hirsutism). Polycystic ovary syndrome (PCOS) is characterized by chronic anovulation, polycystic ovarian morphology, and biochemical and/or biological signs of hyperandrogenism (The Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group, 2004). Most women with PCOS also exhibit insulin resistance and hyperinsulinaemia, which is independent of obesity (Dunaif et al., 1989). Insulin resistance is defined as a diminished ability of cells to respond to the action of insulin and predisposes one to type 2 diabetes mellitus. In addition, insulin resistance is associated with hypertension, dyslipidaemia, obesity and atherosclerotic cardiovascular disease (DeFronzo and Ferrannini, 1991). Numerous hormones and cytokines are known to regulate insulin action. Adipocytokines, such as adiponectin, leptin, resistin and tumour necrosis factor-a, are all involved in mediating insulin resistance. Among them, adiponectin is a circulating protein secreted by small adipocytes (Maeda et al., 1996). Notably, serum adiponectin levels are lower in patients with type 2 diabetes, obesity and coronary heart disease than in healthy subjects, and a number of studies have suggested that # The Author Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org 1011

2 T.Baba et al. adiponectin exerts an insulin sensitizing effect (Hotta et al., 2000; Weyer et al., 2001; Matsubara et al., 2002; Stefan and Stumvoll, 2002). Despite reports of an association between PCOS and adiponectin (Sieminska et al., 2004; Ardawi and Rouzi, 2005; Carmina et al., 2005), there have been no studies examining adiponectin levels in, or the prevalence of PCOS and insulin resistance, among FTM transsexuals. The aim of the present study, therefore, is to investigate the possible association between PCOS, insulin resistance and FTM transsexuality. Materials and methods Subjects The study population consisted of 69 Japanese FTM transsexual patients, years of age, who were seen in the GID clinic of Sapporo Medical University Hospital between December 2003 and May None of these women had ever received hormone treatment or sex re-assignment surgery. Psychological diagnoses were made by a specialized psychiatrist at the clinic according to the guidelines of our institution, the Japanese Society of Psychiatry and Neurology, ICD-10 and DSM-IV (World Health Organization, 1992; American Psychiatric Association, 1994). Patients who were found to have other psychiatric abnormalities such as schizophrenia or virilizing endocrine disorders were excluded. All patients were chromosomal and phenotypic females. Physical and biological study Prior to treatment, the patients received a physical examination. Transvaginal or transrectal ultrasonographic scans were performed using a 5 MHz probe, and serum endocrine analysis was performed during the follicular phase of their menstrual cycle. The measured hormones included luteinizing hormone (LH), follicle stimulating hormone (FSH), estradiol, prolactin, total testosterone, free testosterone, androstenedione and dehydroepiandrosterone sulphate (DHEAS). Elevated serum androgen concentrations were defined as: total testosterone 0.7 ng ml 21, free testosterone 1.0 pg ml 21, androstenedione 3.5 ng ml 21 and DHEAS 3500 pg ml 21. Adiponectin, fasting plasma glucose and fasting insulin levels were also measured, and insulin resistance was determined using the homeostasis model assessment of insulin resistance (HOMA-IR) (Matthews et al., 1985). HOMA-IR was calculated using the formula: FPG (mg dl 21 )- IRI (mu ml 21 )/405, where FPG is the fasting plasma glucose level and IRI is the fasting insulin level. Patients with an HOMA-IR. 2.0 were considered to be insulin resistant. Adiponectin levels,8.0 mgml 21 were considered abnormal. PCOS criteria The diagnosis of PCOS was based on the Rotterdam 2003 criteria, which include oligo- and/or anovulation, clinical and/or biochemical signs of hyperandrogenism and polycystic ovarian morphology. Women who met at least two of the above-mentioned criteria were diagnosed as having PCOS. Results On the basis of the Rotterdam 2003 criteria, 40 of the 69 FTM transsexual patients (58.0%) participating in this study were diagnosed as having PCOS. Of those, 11 were obese (BMI 25 kg m 22 ) and 29 were lean (BMI,25 kg m 22 ). Among the 29 patients without PCOS, only five were obese. Tables I and II summarize the anthropometric, endocrine and metabolic characteristics of the study population. There were statistically significant differences between the PCOS and non-pcos groups with respect to LH (P, 0.001) and androstenedione (P, 0.05) levels. Although the difference was not statistically significant, the PCOS group also had slightly higher total testosterone levels than the non-pcos group ( versus , P ¼ ). When compared with the lean group, the obese group had significantly higher fasting insulin (P, 0.05), higher HOMA-IR (P, 0.05), lower adiponectin (P ¼ ) and higher total testosterone (P, 0.05) (Table I). Patients in the obese PCOS group had significantly higher LH (P, 0.05), higher FSH (P, 0.05) and higher androstenedione (P, 0.005) than those in the obese non-pcos group. Although the differences were not statistically significant, possibly due to the lack of power, the obese PCOS group also had higher total testosterone Table I. The characteristics of study subjects Polycystic ovary syndrome (PCOS) (n ¼ 40) Non-PCOS (n ¼ 29) Obese (n ¼ 16) Lean (n ¼ 53) Age BMI * LH ** FSH Estradiol Prolactin Fasting plasma glucose Fasting insulin *** Homeostasis model assessment of insulin resistance *** Adiponectin **** Total testosterone *** Androstenedione ***** Dehydro epiandrosterone sulphate Data are expressed as mean + SD. Statistical analyses were performed using the Mann Whitney U-test. *P, versus lean group, **P, versus non-pcos group, ***P, 0.05, ****P ¼ , *****P,

3 PCOS and female-to-male transsexuality Table II. The characteristics of obese and lean subjects Obese polycystic ovary syndrome (PCOS) (n ¼ 11) Obese non-pcos (n ¼ 5) Lean PCOS (n ¼ 29) Lean non-pcos (n ¼ 24) Age BMI LH * ** FSH * Estradiol Prolactin Fasting plasma glucose Fasting insulin Homeostasis model assessment of insulin resistance Adiponectin Total testosterone Androstenedione *** Dehydro epiandrosterone sulphate Data are expressed as mean + SD. Statistical analyses were performed using the Mann Whitney U-test. *P, 0.05, **P, versus lean non-pcos group, ***P, versus obese non-pcos group. ( versus , P ¼ ) and DHEAS ( versus , P ¼0.1125) levels than the obese non-pcos group. There were no statistically significant differences between the lean PCOS group and the lean non-pcos group, except with respect to LH (P, ) (Table II). Of the 49 patients for whom HOMA-IRs were calculated, 15 (30.6%) showed insulin resistance. These insulin resistant cases were 10 out of 32 (31.3%) patients in the PCOS group and 5 out of 17 (29.4%) patients in the non-pcos group. Within the various groups, the percentages of patients exhibiting abnormal HOMA-IR were 72.7% (obese), 17.1% (lean), 62.5% (obese PCOS), 100% (obese non-pcos), 19.0% (lean PCOS) and 14.3% (lean non-pcos). A statistically significant difference was seen only between the obese and lean groups (P, 0.005, OR ¼ 12.89, 95% CI: ) (Figure 1). Of the 59 patients in whom serum adiponectin levels were measured, 18 (30.5%) showed hypoadiponectinaemia. Within the groups, the percentages of patients exhibiting hypoadiponectinaemia were 33.3% (PCOS), 26.9% (non-pcos), 71.4% (obese), 17.8% (lean), 77.8% (obese PCOS), 60.0% (obese non-pcos), 16.7% (lean PCOS) and 19.0% (lean non-pcos). A statistically significant difference was seen only between the obese and lean groups (P, , OR ¼ 11.56, 95% CI: ) (Figure 2). Elevated levels of at least one of the four measured androgens (hyperandrogenaemia) were seen in 27 of the 69 study subjects (39.1%). Within the groups, the percentages of patients exhibiting hyperandrogenaemia were 57.5% (PCOS) 13.8% (non-pcos), 62.5% (obese), 32.1% (lean), 81.8% (obese PCOS), 20.0% (obese non-pcos), 48.3% (lean PCOS), and 12.5% (lean non-pcos) (Figures 3 and 4). In most cases, high levels of androstenedione and DHEA coexisted with high testosterone levels. There were statistically significant differences between the PCOS and non-pcos groups (P, , OR ¼ 8.46, 95% CI: ), the obese and lean groups (P, 0.05, OR ¼ 3.53, 95% CI: ), the obese PCOS and obese non-pcos groups (P, 0.05, OR ¼ 18.00, 95% CI: ) and the lean PCOS and lean non-pcos groups (P, 0.01, OR ¼ 6.53, 95% CI: ). There were no statistically significant differences between the obese PCOS and lean PCOS (P ¼ ) groups or between the obese non-pcos and lean non-pcos groups (P ¼ ). Figure 1. Percentage of women with insulin resistance based on homeostasis model assessment of insulin resistance (HOMA-IR). HOMA-IR 2.0 was considered insulin resistant. 1013

4 T.Baba et al. Figure 2. Percentage of women with hypoadiponectinaemia. Adiponectin levels,8.0 mgml 21 were considered abnormal. Discussion PCOS is a common hormonal disorder that, according to previously published diagnostic criteria (Futterweit, 1999; Ehrmann, 2005), affects 5 10% of women of reproductive age. In contrast, our study shows that as many as 58% of FTM transsexual patients exhibit PCOS prior to treatment. Using the Rotterdam criteria, patients with ovulatory menstrual cycles, polycystic ovaries and hyperandrogenaemia, and who had been excluded from the PCOS group on the basis of the earlier diagnostic criteria, were also diagnosed as having PCOS. Consequently, the prevalence of PCOS was estimated to be higher than previously thought. Unfortunately, there are no data available on the prevalence of PCOS diagnosed based on the Rotterdam criteria among healthy Japanese women. A population-based study may be needed to compare the prevalence of PCOS among FTM transsexual individuals and among women in the general population. Notably, PCOS often accompanies hyperandrogenism, insulin resistance and obesity, which suggests that FTM transsexual patients are likely to show excessive androgen levels. In adult humans, hypersecretion of androgens causes the ovary to have a polycystic appearance; moreover, it is known that ovaries removed from FTM transsexual patients after long-term androgen therapy are histopathologically polycystic. These findings thus suggest that FTM transsexualism is frequently associated with PCOS (Futterweit and Deligdisch, 1986; Pache et al., 1993). The aetiology of transsexualism remains unclear, although high androgen levels during critical periods of brain differentiation have been shown to cause male-like behaviour in female animals. In human studies, genetic females with congenital adrenal hyperplasia are more likely to show genderatypical play behaviour and gynephilic sexual orientation (Meyer-Bahlburg, 1979; Dittmann et al., 1992). In that regard, the prevalence of PCOS is significantly higher among lesbian women than among heterosexual women, and lesbian women with PCOS have more pronounced hyperandrogenism than heterosexual women with PCOS (Agrawal et al., 2004). In an earlier study, males with a 5a-reductase deficiency showed diminished in utero production of dihydrotestosterone (Imperato-McGinley et al., 1979), which resulted in ambiguous genitalia in the affected male fetus. Indeed, affected infants were often considered to be girls at birth and raised as females. Virilization occurred during puberty, however, Figure 3. Percentage of women with hyperandrogenaemia. Elevated serum androgen levels were defined as follows: total testosterone, 0.7 ng ml 21 ; free testosterone, 1.0 pg ml 21 ; androstenedione, 3.5 ng ml 21 and dehydroepiandrosterone sulphate 3500 pg ml

5 PCOS and female-to-male transsexuality Figure 4. Percentage of women with hyperandrogenaemia. Elevated serum androgen levels were defined as follows: total testosterone, 0.7 ng ml 21 ; free testosterone, 1.0 pg ml 21 ; androstenedione, 3.5 ng ml 21 and dehydroepiandrosterone sulphate 3500 pg ml 21. because testicular biosynthesis of testosterone and its peripheral actions were normal. These children typically began to realize that they were different from other girls when they were between 7 and 12 years of age, with 17 of the 18 children studied changing their gender identity to male. Thus, androgen production during the pubertal period appears to make a strong contribution to formation of male gender identity. This idea was further confirmed in another study of complete androgen insensitivity syndrome, which results from androgen receptor dysfunction and is a useful model for studying the effects of androgens during psychosexual development. Patients with this disorder are 46, XY individuals with testes, normal or high levels of testosterone and normal female external genitalia. Wisniewski et al. (2000) reported that all 14 of the cases of complete androgen insensitivity syndrome they studied reported satisfaction with being a woman. Taken together, these studies indicate that androgen production during both the fetal and prepubertal periods is closely related to the formation of gender identity, although other factors may also contribute. Obesity is also a characteristic of at least 50 65% of women with PCOS (Futterweit, 1999). In our study, however, only 11 of 40 (27.5%) FTM cases with PCOS were obese and, overall, only 16 of 69 (23.2%) were obese. Obesity reduces the levels of circulating adiponectin, which is related to insulin resistance. It also worsens insulin resistance through dysregulated production of other cytokines and proteins, including tumour necrosis factor-a, leptin and resistin. In the present study, the obese FTM group showed significantly higher incidence of high HOMA-IR and/or low adiponectin than the lean group. In contrast, there was little association between the occurrence of PCOS and high HOMA-IR or hypoadiponectinaemia. Thus, insulin resistance in FTM transsexual patients was associated with obesity, not with PCOS, itself. Administration of exogenous androgen, the treatment of choice for FTM transsexual patients, causes insulin resistance. Given that insulin resistance is closely related to type 2 diabetes, hypertension, dyslipidacmia and cardiovascular disease, complications associated with PCOS would be expected to represent a significant health risk for FTM transsexual cases. Hyperandrogenism is a common finding in PCOS and FTM transsexual patients (Bosinski et al., 1997). Notably, 39.1% of untreated FTM transsexual patients showed high levels of at least one androgen. In our experience, plasma androgen levels in PCOS patients without transsexualism do not exceed the following concentrations: total testosterone, 2.0 ng ml 21 ; free testosterone, 2.0 pg ml 21 ; androstenedione, 5.0 ng ml 21 and DHEAS, 5000 pg ml 21. It should be noted, however, that we experienced five cases of FTM transsexuality with extremely high androgen levels. In one case of extremely high testosterone levels, the total and free testosterone concentrations were 4.3 ng ml 21 and 4.32 pg ml 21, respectively. Among the other cases, one individual showed an androstenedione level of 5.3 ng ml 21 and another showed a DHEAS level of 5560 pg ml 21. All these cases were complicated by PCOS. The very high levels make one suspect that these individuals have been self-administering exogenous androgen, as self-medication by transsexual patients is frequently seen in Japan. For instance, in a study carried out at the Okayama University, Okayama, Japan, 127 of 312 (40.7%) FTM patients self-medicated. In our experience, however, almost all FTM patients taking exogenous androgen show haemoconcentration (haematocrit.45%), and the haematocrit levels of our five patients were,45%. Moreover, androgen therapy quickly induces amenorrhoea, irrespective of the state of the patient s menstrual cycle before treatment, and our patients were all cycling at the time of their first clinic visit. Finally, self-medication was favourable to their treatment, so it was not in their interest to deceive us. It thus seems highly unlikely to us that it is by chance that most cases of FTM transsexuality are complicated by PCOS and extremely high androgen levels. Among the subjects in the present study, those who exhibited PCOS were more likely to show hyperandrogenaemia than those who did not. In addition, patients in the obese group were more likely to show hyperandrogenaemia than those in 1015

6 T.Baba et al. the lean group. Given that PCOS and obesity are thought to be associated, we compared the obese PCOS group with obese non-pcos group and lean PCOS group with lean non-pcos group to eliminate the influence of the BMI. The results of these comparisons suggested that PCOS is associated with hyperandrogenaemia. Similarly, we compared the obese PCOS group with lean PCOS group and obese non-pcos group with lean non-pcos group to eliminate the influence of the PCOS. These findings showed that there is little relationship between obesity itself and hyperandrogenaemia. Our findings show that many cases of FTM transsexuality are associated with PCOS and hyperandrogenaemia, which suggests that they are important factors in the pathogenesis of FTM transsexualism. In addition, our findings also suggest that when administering androgen therapy to FTM transsexual patients, it is important that practitioners keep in mind that this treatment worsens insulin resistance. Acknowledgements This study was supported by a grant-in aid for scientific research from the Ministry of Education, Culture, Sports, Science and Technology of Japan. We thank Dr Kengo Manase for his help. References Agrawal R, Sharma S, Bekir J, Conway G, Bailey J, Balen AH and Prelevic G. (2004) Prevalence of polycystic ovaries and polycystic ovary syndrome in lesbian women compared with heterosexual women. Fertil Steril 82, American Psychiatric Association (1994) Diagnostic and Statistical Manual of Mental Disorders, 4th edn (DSM-IV). American Psychiatric Press, Washington, DC, USA. Ardawi MS and Rouzi AA (2005) Plasma adiponectin and insulin resistance in women with polycystic ovary syndrome. Fertil Steril 83, Berenbaum SA and Bailey JM (2003) Effects on gender identity of prenatal androgens and genital appearance: evidence from girls with congenital adrenal hyperplasia. J Clin Endocrinol Metab 88, Bosinski HAG, Peter M, Bonatz G, Arndt R, Heidenreich M, Sippell WG and Wille R (1997) A higher rate of hyperandrogenic disorders in female-to-male transsexuals. Psychoneuroendocrinology 22, Carmina E, Orio F, Palomba S, Cascella T, Longo RA, Colao AM, Lombardi G and Lobo RA (2005) Evidence for altered adipocyte function in polycystic ovary syndrome. Eur J Endocrinol 152, Collaer ML and Hines M (1995) Human behavioral sex differences: a role for gonadal hormones during early development? Psychol Bull 118, DeFronzo RA and Ferrannini E (1991) Insulin resistance. A multifaceted syndrome responsible for NIDDM obesity hypertension dyslipidemia, and atherosclerotic cardiovascular disease. Diabetes Care 14, Dittmann RW, Kappes ME and Kappes MH (1992) Sexual behavior in adolescent and adult females with congenital adrenal hyperplasia. Psychoneuroendocrinology 17, Dunaif A, Segal KR, Futterweit W and Dobrjansky A (1989) Profound peripheral insulin resistance, independent of obesity, in polycystic ovary syndrome. Diabetes 38, Ehrmann DA (2005) Polycystic ovary syndrome. N Engl J Med 352, Futterweit W (1999) Polycystic ovary syndrome: clinical perspectives and management. Obstet Gynecol Surv 54, Futterweit W and Deligdisch L (1986) Histopathological effects of exogenously administered testosterone in 19 female to male transsexuals. J Clin Endocrinol Metab 62, Hotta K, Funahashi T, Arita Y, Takahashi M, Matsuda M, Okamoto Y, Iwahashi H, Kuriyama H, Ouchi N and Maeda K et al. (2000) Plasma concentration of a novel, adipose-specific protein, adiponectin, in type 2 diabetic patients. Arterioscler Thromb Vasc Biol 20, Imperato-McGinley J, Peterson RE, Gautier T and Sturla E (1979) Androgens and the evolution of male-gender identity among male pseudohermaphrodites with 5a-reductase deficiency. N Engl J Med 300, Maeda K, Okubo K, Shimomura I, Funahashi T, Matsuzawa Y and Matsubara K (1996) cdna cloning and expression of a novel adipose specific collagen-like factor, apm1. Biochem Biophys Res Commun 221, Matsubara M, Maruoka S and Katayose S (2002) Decreased plasma adiponectin concentrations in women with dyslipidemia. J Clin Endocrinol Metab 87, Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF and Turner RC (1985) Homeostasis model assessment: insulin resistance and b-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28, Meyer-Bahlburg HFL (1979) Sex hormones and female homosexuality: a critical examination. Arch Sex Behav 8, Pache TD, de Jong FH, Hop WC and Fauser BC (1993) Association between ovarian changes assessed by transvaginal sonography and clinical and endocrine signs of the polycystic ovary syndrome. Fertil Steril 59, Sieminska L, Marek B, Kos-Kudla B, Niedziolka D, Kajdaniuk D, Nowak M and Glogowska-Szelag J (2004) Serum adiponectin in women with polycystic ovarian syndrome and its relation to clinical, metabolic and endocrine parameters. J Endocrinol Invest 27, Stefan N and Stumvoll M (2002) Adiponectin its role in metabolism and beyond. Horm Metab Res 34, The Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group (2004) Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod 19, Weyer C, Funahashi T, Tanaka S, Hotta K, Matsuzawa Y, Pratley RE and Tataranni PA (2001) Hypoadiponectinemia in obesity and type 2 diabetes: close association with insulin resistance and hyperinsulinemia. J Clin Endocrinol Metab 86, Wisniewski AB, Migeon CJ, Meyer-Bahlburg HFL, Gearhart JP, Berkovitz GD, Brown TR and Money J (2000) Complete androgen insensitivity syndrome: long-term medical, surgical, and psychosexual outcome. J Clin Endcrinol Metab 85, World Health Organization (1992) The ICD-10 classification of mental and behavioral disorders. Clinical descriptions and diagnostic guidelines. World Health Organization, Geneva, Switzerland. Submitted on August 15, 2006; resubmitted on October 10, 2006; accepted on October 23,