- Email: [email protected]
Idiopathic hirsutism: an uncommon cause of hirsutism in Alabama
FERTILITY AND STERILITYt VOL. 70, NO. 2, AUGUST 1998 Copyright ©1998 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A.
Idiopathic hirsutism: an uncommon cause of hirsutism in Alabama Ricardo Azziz, M.D., M.P.H.,*† Wende T. Waggoner, B.S.,* Tatiana Ochoa, M.D.,*‡ Eric S. Knochenhauer, M.D.,* and Larry R. Boots, Ph.D.* University of Alabama at Birmingham, Birmingham, Alabama
Objective: To determine the prevalence of idiopathic hirsutism among a population of consecutive hirsute patients. Design: Prospective cohort study. Setting: University-based clinic. Patient(s): Premenopausal women with a complaint of hirsutism who were not receiving hormonal therapy. Intervention(s): Evaluations for total and free testosterone, (T), 17-hydroxyprogesterone (17-HP), and DHEAS serum levels. Main Outcome Measure(s): Ovulatory function in women with cycles of #35 days in length was assessed with a basal body temperature (BBT) calendar and day 22–24 progesterone levels. Result(s): Of 132 consecutive hirsute women studied, 68 had cycles of .35 days in length. Of the remaining 64 patients, 25 also had oligo/anovulation by BBT and day 22–24 progesterone level. Of the 39 patients with hirsutism and regular ovulatory function, 22 had total and free T and DHEAS levels within normal limits.
Received December 30, 1997; revised and accepted March 5, 1998. Supported in part by grant RO1-HD29364 from the National Institutes of Health, Bethesda, Maryland (R.A.). Presented in part at the 53rd Annual Meeting of the American Society for Reproductive Medicine, Cincinnati, Ohio, October 18 –22, 1997. Reprint requests: Ricardo Azziz, M.D., M.P.H., The University of Alabama at Birmingham, Department of Obstetrics and Gynecology, 618 South 20th Street, OHB 549, Birmingham, Alabama 35233-7333 (FAX: 205-975-5732; E-mail: [email protected]). * Department of Obstetrics and Gynecology. † Department of Medicine. ‡ Present address: Ministerio de Salud Publica del Ecuador, Quito, Ecuador. 0015-0282/98/$19.00 PII S0015-0282(98)00141-1
274
Conclusion(s): If idiopathic hirsutism is defined by the presence of hirsutism, regular ovulation, and normal androgen levels, only 17% of consecutive hirsute patients can be diagnosed with the disorder. Alternatively, if idiopathic hirsutism is based solely on the presence of hirsutism and regular ovulation, regardless of androgen levels, then 29% of the total hirsute population can be considered as having idiopathic hirsutism. Importantly, 40% of hirsute patients with a history of “regular cycles” were actually oligo/anovulatory, indicating the need to objectively assess ovulatory function in such patients. (Fertil Sterilt 1998;70:274 – 8. ©1998 by American Society for Reproductive Medicine.) Key Words: Hirsutism, androgens, polycystic ovary syndrome, ovulatory dysfunction
Some investigators have defined idiopathic hirsutism by the presence of hirsutism and regular menstrual cycles only, regardless of circulating androgen levels (26 –28); hence, it has been suggested that idiopathic hirsutism is one of the most frequent causes of hirsutism because between 50% and 70% of all hirsute women claim to have regular menses (1–3). However, a normal menstrual history does not exclude the presence of polycystic ovaries determined either pathologically (4) or sonographically (2, 5), elevated circulating androgen levels (1), or the presence of nonclassic adrenal hyperplasia (6). In response, other investigators define idiopathic hirsutism more strictly, as presenting with hirsutism, regular ovulation, and normal androgen levels (7–10). Although the presence of eumenorrheic hirsutism may be sufficient to establish the diagnosis of idiopathic hirsutism, it is also possible that many of these women actually suffer from
the polycystic ovary syndrome (PCOS). Because the diagnosis of PCOS hinges on the finding of ovulatory dysfunction (11), then regular ovulatory function should be confirmed in eumenorrheic hirsute women suspected of suffering from idiopathic hirsutism. To determine the extent to which ovulatory dysfunction affects hirsute eumenorrheic women—and hence establish the prevalence of idiopathic hirsutism among all hirsute women—we prospectively studied 133 consecutive hirsute women.
MATERIALS AND METHODS Protocol Between June 1994 and December 1996, all women who were not receiving hormonal treatment and who presented with hirsutism (as defined below) were included in this study. Serum was obtained and measured for total and free T, 17-hydroxyprogesterone (17-HP), and
DHEAS serum levels. Hyperandrogenemia was defined as a total or free T, and/or DHEAS level above the 95th percentile of normal. Based on 44 consecutive healthy eumenorrheic nonhirsute women aged 18 – 40 years, the values used were as follows: total T .89 ng/dL (3.08 nmol/L), free T .0.66 ng/dL (0.023 nmol/L), and DHEAS .2,750 ng/mL (7.46 mmol/L) (Azziz, unpublished data). Patients with a basal 17-HP level of .2 ng/mL underwent an adrenocorticotropic hormone stimulation test to exclude the possibility of 21-hydroxylase deficient nonclassic adrenal hyperplasia (NCAH), defined by a stimulated 17-HP level of .10 ng/ mL, as previously described (12). Hirsutism was defined by a score of $6, determined with a modification of the Ferriman-Gallwey (F-G) method, as previously described (13). This hirsutism score is obtained for only 5%– 6% of women in our general population (14). Women with cycles lengths of .35 days were classified as oligoovulatory without further testing. Women with cycles of #35 days in length were asked to maintain a daily basal body temperature (BBT) calendar and to obtain a day 22–24 progesterone level during one menstrual cycle. In these women oligoovulation was established if the day 22–24 progesterone level was ,4 ng/mL (with levels .4 ng/mL indicating prior ovulation and corpus luteum formation). The BBT was used to confirm that the progesterone level was actually obtained between days 22 and 24 of the menstrual cycle. The study was approved by the institutional review board of the University of Alabama at Birmingham.
Statistical Analysis Continuous variables between two groups were compared with use of the two-tailed unpaired t-test if values were normally distributed, or by the Mann-Whitney U test if values were parametrically distributed (i.e., F-G scores). Fractions (i.e., percent nonwhite) were compared by the x 2 test.
RESULTS Our study algorithm is depicted in Figure 1. Of 133 consecutive hirsute women initially included, one did not complete the study and was then excluded. Of the remaining 132 patients studied, 68 (52% of total) had a history of menstrual cycles of .35 days in length. Of the remaining 64 patients claiming to have regular menses at ,35-day intervals, 25 (39% or 19% of total) actually had oligo/anovulation by BBT and day 22–24 progesterone level. Finally, of the 39 patients with hirsutism and regular ovulatory function (i.e., 61% or 29% of the total), 22 (56% or 17% of the total) had total and free T, and DHEAS levels within normal limits; and the remaining 17 women had at least one elevated androgen level. One of these hirsute women with regular ovulation, but increased androgen levels, was diagnosed as having NCAH. Women with hirsutism and regular menses (n 5 64) had FERTILITY & STERILITYt
a lower mean body mass index (BMI), waist-hip ratio (WHR), and total and free T levels, but similar mean DHEAS levels, age, F-G score, or percent nonwhite, compared with those with irregular cycles (n 5 68) (Table 1). The only difference between hirsute women with “regular cycles” and oligo/anovulation (n 5 25), and those with normal ovulatory function (n 5 39), was a higher total T level in the former, although free T also tended to be higher (P 5 .078) (Table 2). Other than the expected differences in androgen levels, there were no statistically significant differences between ovulatory euandrogenic and hyperandrogenic hirsute patients (Table 3).
DISCUSSION The definition of idiopathic hirsutism has been a dynamic process, evolving in conjunction with our understanding of the various androgen excess disorders and the phenotypic heterogeneity of these patients. Early reports defined idiopathic hirsutism as “hirsutism of unknown cause,” regardless of whether they had irregular menstruation and ovulation and/or elevated circulating androgen levels (15–20). Subsequently, other investigators defined idiopathic hirsutism as hirsutism in the presence of “regular” menses (2, 21, 22). However, approximately 50%–75% of eumenorrheic hirsute women also demonstrate polycystic appearing ovaries on sonography (2, 3) and/or elevated circulating levels of either DHEAS or total and/or free T (1), suggestive of a more general endocrine abnormality. Furthermore, a history of “regular cycles” appears to be a poor predictor of regular ovulation, because 40% of our eumenorrheic hirsute patients had oligo/anovulation on closer investigation. Consequently, 40% of patients previously presumed to have idiopathic hirsutism can be considered as having PCOS, as previously defined (11). Whether the presence of hyperandrogenemia excludes the diagnosis of idiopathic hirsutism remains less clear, with some investigators diagnosing the disorder regardless of circulating androgen levels (2, 21, 22). Nonetheless, we do not agree and argue that the hirsutism cannot be considered “idiopathic” when patients have clear-cut hyperandrogenemia. Although the source of the elevated androgens may remain a mystery, the physical finding of hirsutism is directly explainable by the increased circulating androgen levels. In fact, .40% of our hirsute regularly ovulating patients had evidence of hyperandrogenemia. In agreement, other investigators reported that approximately one half of regularly cycling hirsute women had elevated levels of one or more androgens (1). Consequently, and in agreement with other investigators (7–10), we believe that idiopathic hirsutism should be defined only in those women with [1] hirsutism [2] normal ovulatory function, and [3] a normal androgen profile. Because the definition of idiopathic hirsutism has varied significantly over the past three decades, a precise estimate 275
FIGURE 1 Algorithm of 132 consecutive patients seen prospectively to determine the prevalence of idiopathic hirsutism. Patients with the polycystic ovary syndrome (oligoovulation, and hirsutism and/or hyperandrogenemia) and women with hirsutism and hyperandrogenemia but apparently normal ovulatory function, were considered as having functional androgen excess. If the diagnosis of idiopathic hirsutism is based solely on the presence of hirsutism and regular ovulation, regardless of androgen levels, then 29% of our total untreated hirsute population could be considered affected. However, only 17% of all hirsute patients can be diagnosed as having the disorder, if idiopathic hirsutism is defined as hirsutism, regular ovulation, and normal androgen levels. BBT 5 basal body temperature chart; P4 5 progesterone level.
of its prevalence as a cause of hirsutism has been difficult to establish. Consistent with other reports (1–3), approximately 50% of our consecutive hirsute women claimed to have “regular menses,” and it may then be supposed that idio276
Azziz et al.
Prevalence of idiopathic hirsutism
pathic hirsutism is a common cause of hirsutism. However, as noted above, 40% of these women had ovulatory dysfunction and would be considered as having PCOS. In our study, if idiopathic hirsutism, was defined as hirsutism in the presVol. 70, No. 2, August 1998
TABLE 1
TABLE 3
Comparison of hirsute patients with and without a history of regular menstrual cycles.
Comparison of hirsute ovulatory patients with and without elevated androgen levels.
History of menstrual cycle
Characteristic Age (y) BMI (kg/m2) Waist-hip ratio Total T (ng/dL) Free T (ng/dL) DHEAS (ng/mL) SHBG (nmol/L) 17-HP (ng/mL) F-G score No. of nonwhite subjects with indicated cycles (%)
Hirsute ovulatory patient group
“Regular cycles”
“Irregular cycles”
P value
30.1 6 9.8 29.1 6 8.3 0.79 6 0.10 71.2 6 16.2 0.74 6 0.24 2,815 6 1,336 184 6 88 1.14 6 1.11 9.8 6 3.2 8/65 (32)
27.9 6 7.4 32.2 6 7.6 0.83 6 0.07 80.1 6 20.0 0.88 6 0.32 2,514 6 1,426 183 6 52 2.52 6 7.67 9.6 6 4.0 11/68 (16)
NS .027 .018 .006 .006 NS NS NS NS NS
Characteristic Age (y) BMI (kg/m2) Waist-hip ratio T (ng/dL) Free T (ng/dL) DHEAS (ng/mL) SHBG (nmol/L) 17-HP (ng/mL) F-G score No. of nonwhite subjects with indicated cycles (%)
Elevated androgens
Normal androgens
P value
29.4 6 10.1 30.4 6 7.6 0.80 6 0.09 72.5 6 12.4 0.84 6 0.23 3,362 6 1,428 165 6 53 1.0 6 1.2 10.3 6 3.4 2/22 (9%)
34.1 6 9.9 27.1 6 6.5 0.81 6 0.15 60.1 6 13.4 0.51 6 0.08 1,950 6 627 219 6 54 0.8 6 0.8 8.8 6 3.1 3/17 (18%)
NS NS NS ,.005 ,.001 ,.001 ,.004 NS NS NS
Note: All values are means 6 SD unless otherwise indicated. BMI 5 body mass index; F-G score 5 hirsutism scored by a modified Ferriman-Gallwey method; 17-HP 5 17-hydroxyprogesterone; NS 5 not significant; SHBG 5 sex hormone-binding globulin.
Note: All values are means 6 1 SD unless otherwise indicated. BMI 5 body mass index; F-G score 5 hirsutism scored by a modified Ferriman-Gallwey method; 17-HP 5 17-hydroxyprogesterone; NS 5 not significant; SHBG 5 sex hormone-binding globulin.
ence of regular ovulation, regardless of androgen levels, then 29% (39 of 132) would be considered to have the disorder.
agreement, Carmina (23) noted that of 588 consecutive Italian patients complaining of hirsutism, 51% claimed regular cycles, but only 6% had regular ovulation, normal total and free T, and DHEAS levels, and no evidence of polycystic ovaries on ultrasonography.
Alternatively, if the more strict criteria were used whereby idiopathic hirsutism was considered only if patients had hirsutism, regular ovulation, and normal androgen levels, then only 17% would be considered as being affected. In
TABLE 2 Comparison of hirsute patients with a history of “regular menses” who do or do not have regular ovulation. History of “regular” menstrual cycle
Characteristic
Hirsute ovulatory
Hirsute oligo/ anovulatory
Age (y) 31.5 6 10.2 28.0 6 9.2 BMI (kg/m2) 29.0 6 7.2 29.1 6 10.0 Waist-hip ratio 0.80 6 0.11 0.78 6 0.07 Total T (ng/dL) 67.1 6 14.1 76.4 6 17.4 Free T (ng/dL) 0.70 6 0.24 0.80 6 0.22 DHEAS (ng/mL) 2,747 6 1,340 2,960 6 1,361 SHBG (nmol/L) 188 6 59 174 6 38 17-HP (ng/mL) 1.0 6 1.0 1.4 6 1.3 F-G score 9.6 6 3.3 10.1 6 3.1 No. of nonwhite subjects with indicated cycles (%) 5/39 (13%) 2/25 (8%)
P value NS NS NS ,.025 NS (.078) NS NS NS NS NS
Note: All values are means 6 1 SD unless otherwise indicated. BMI 5 body mass index; F-G score 5 hirsutism scored by a modified Ferriman-Gallwey method; 17-HP 5 17-hydroxyprogesterone; NS 5 not significant, SHBG 5 sex hormone-binding globulin.
FERTILITY & STERILITYt
Overall, it appears that idiopathic hirsutism affects between 15% and 30% of hirsute patients seen, depending on whether normal androgen levels are required for the definition (Fig. 1). Because menstrual history does not accurately reflect ovulatory status in hirsute patients, it is imperative that ovulatory function be documented clearly, perhaps by using daily BBT and/or a luteal progesterone level. Although not all investigators agree, we believe that idiopathic hirsutism should be a diagnosis of exclusion in which ovulatory disorders, hyperandrogenemia, and other androgen excess disorders (e.g., NCAH) are ruled out. All other patients should be defined as having “functional androgen excess” (Fig. 1), and include both patients with PCOS and those with hirsutism and hyperandrogenemia, but apparently normal ovulatory function. Because the sole measurement of a total T serum level is grossly insensitive for the detection of hyperandrogenemia (24 –27), we suggest that only those patients who have demonstrated at a minimum normal DHEAS, and total and free T levels be included. Nonetheless, it remains important to better define idiopathic hirsutism, perhaps by determining the differences between patients with and without hyperandrogenemia, such as risk of developing frank ovulatory dysfunction (i.e., PCOS) and/or metabolic aberrations. 277
References 1. Mehta A, Matwijiw I, Taylor PJ, Salamon EA, Kredentser JV, Faiman C. Should androgen levels be measured in hirsute women with normal menstrual cycles? Int J Fertil 1992;37:354 –7. 2. Jahanfar S, Eden JA. Idiopathic hirsutism or polycystic ovary syndrome? Aust NZ J Obstet Gynaecol 1993;33:414 – 6. 3. O’Driscoll JB, Mamtora H, Higginson J, Pollock A, Kane J, Anderson DC. A prospective study of the prevalence of clear-cut endocrine disorders and polycystic ovaries in 350 patients presenting with hirsutism or androgenic alopecia. Clin Endocrinol (Oxf) 1994;41:231– 6. 4. Raj SG, Talbert LM. Polycystic ovarian disease. Obstet Gynecol Annu 1984;13:261–73. 5. Adams J, Polson DW, Franks S. Prevalence of polycystic ovaries in women with anovulation and idiopathic hirsutism. Br Med J 1986;293: 355–9. 6. Azziz R, Dewailly D, Owerbach D. Non-classic adrenal hyperplasia: current concepts. J Clin Endocrinol Metab 1994;78:810 –5. 7. Mowszowicz I, Melanitou E, Doukani A, Wright F, Kuttenn F, Mauvais-Jarvis P. Androgen binding capacity and 5a-reductase activity in pubic skin fibroblasts from hirsute patients. J Clin Endocrinol Metab 1983;56:1209 –13. 8. Serafini P, Lobo RA. Increased 5a reductase activity in idiopathic hirsutism. Fertil Steril 1985;43:74 – 8. 9. Couzinet B, Pholsena M, Young J, Schaison G. The impact of a pure anti-androgen (flutamide) on LH, FSH, androgens and clinical status in idiopathic hirsutism. Clin Endocrinol (Oxf) 1993;39:157– 62. 10. Tolino A, Petrone A, Sarnacchiaro F, Cirillo D, Ronsini S, Lombardi G, et al. Finasteride in the treatment of hirsutism: new therapeutic perspectives. Fertil Steril 1996;66:61–5. 11. Zawadzki JK, Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach. In: Dunaif A, Givens JR, Haseltine F, Merriam GR, editors. Polycystic ovary syndrome. Boston: Blackwell Scientific Publications, 1992:377– 84. 12. Azziz R, Zacur HA. 21-hydroxylase deficiency in female hyperandrogenism: screening and diagnosis. J Clin Endocrinol Metab 1989;69: 577– 84. 13. Hatch R, Rosenfield RL, Kim MH, Tredway D. Hirsutism: Implications, etiology, and management. Am J Obstet Gynecol 1981;140:815– 30. 14. Knochenhauer ES, Kahsar-Miller MD, Boots LR, Key TJ, Azziz R. Prevalence of hirsutism among unselected Black and Caucasian women in the southeastern United States (U.S.) [abstract P1-357]. Program and
278
Azziz et al.
Prevalence of idiopathic hirsutism
15. 16. 17. 18. 19. 20.
21. 22.
23.
24. 25. 26. 27.
abstracts of the 79th Annual Meeting of the Endocrine Society. Minneapolis, (MN): The Endocrine Society, 1997:224. Bardin WC, Lipsett MB. Testosterone and androstenedione blood production rates in normal women and women with idiopathic hirsutism or polycystic ovaries. J Clin Invest 1967;46:891–7. Casey JH, Nabarro JDN. Plasma testosterone in idiopathic hirsutism, and the changes produced by adrenal and ovarian stimulation and suppression. J Clin Endocrinol 1967;27:1431–5. Thomas JP, Oake RJ. Androgen metabolism in the skin of hirsute women. J Clin Endocrinol Metab 1974;38:19 –22. Kirschner M, Zucker R, Jespersen D. Idiopathic hirsutism—an ovarian abnormality. N Engl J Med 1976;294:637– 41. Ettinger B, Goldfield EB, Burrill KC, Von Werder K, Forsham PH. Plasma testosterone stimulation-suppression dynamics in hirsute women. Am J Med 1973;54:195–200. Pugeat M, Forest MG, Nisula BC, Corniau J, De Peretti E, Tourniaire J. Evidence of excessive androgen secretion by both the ovary and the adrenal in patients with idiopathic hirsutism. Obstet Gynecol 1982;59: 46 –51. McKenna TJ, Miller RB, Liddle GW. Plasma pregnenolone and 17-OH pregnenolone in patients with adrenal tumors, ACTH excess, or idiopathic hirsutism. J Clin Endocrinol Metab 1977;44:231– 6. Rittmaster RS, Thompson DL. Effect of leuprolide and dexamethasone on hair growth and hormone levels in hirsute women: the relative importance of the ovary and the adrenal in the pathogenesis of hirsutism. J Clin Endocrinol Metab 1990;70:1096 –1102. Carmina E. Prevalence of idiopathic hirsutism [abstract no. P-286]. Program and abstracts of the 53rd Annual Meeting of the American Society for Reproductive Medicine. Cincinnati (OH): American Society for Reproductive Medicine, 1997:S229. Boots LR, Potter S, Potter HD, Azziz R. Measurement of total serum testosterone level using commercial kits: high degree of variability and inaccuracy. Fertil Steril 1998;69:286 –92. Vermeulen A, Stoica T, Verdonck L. The apparent free testosterone concentration, an index of androgenicity. J Clin Endocrinol 1971;33: 759 – 67. Mathur RS, Moody LO, Landgrebe S, Williamson HO. Plasma androgens and sex hormone binding globulin in the evaluation of hirsute females. Fertil Steril 1981;35:29 –35. Carter GD, Holland SM, Alaghband-Zadeh J, Rayman G, DorringtonWard P, Wise PH. Investigation of hirsutism: testosterone is not enough. Ann Clin Biochem 1983;20:262–3.
Vol. 70, No. 2, August 1998
Idiopathic hirsutism: an uncommon cause of hirsutism in Alabama Ricardo Azziz, M.D., M.P.H.,*† Wende T. Waggoner, B.S.,* Tatiana Ochoa, M.D.,*‡ Eric S. Knochenhauer, M.D.,* and Larry R. Boots, Ph.D.* University of Alabama at Birmingham, Birmingham, Alabama
Objective: To determine the prevalence of idiopathic hirsutism among a population of consecutive hirsute patients. Design: Prospective cohort study. Setting: University-based clinic. Patient(s): Premenopausal women with a complaint of hirsutism who were not receiving hormonal therapy. Intervention(s): Evaluations for total and free testosterone, (T), 17-hydroxyprogesterone (17-HP), and DHEAS serum levels. Main Outcome Measure(s): Ovulatory function in women with cycles of #35 days in length was assessed with a basal body temperature (BBT) calendar and day 22–24 progesterone levels. Result(s): Of 132 consecutive hirsute women studied, 68 had cycles of .35 days in length. Of the remaining 64 patients, 25 also had oligo/anovulation by BBT and day 22–24 progesterone level. Of the 39 patients with hirsutism and regular ovulatory function, 22 had total and free T and DHEAS levels within normal limits.
Received December 30, 1997; revised and accepted March 5, 1998. Supported in part by grant RO1-HD29364 from the National Institutes of Health, Bethesda, Maryland (R.A.). Presented in part at the 53rd Annual Meeting of the American Society for Reproductive Medicine, Cincinnati, Ohio, October 18 –22, 1997. Reprint requests: Ricardo Azziz, M.D., M.P.H., The University of Alabama at Birmingham, Department of Obstetrics and Gynecology, 618 South 20th Street, OHB 549, Birmingham, Alabama 35233-7333 (FAX: 205-975-5732; E-mail: [email protected]). * Department of Obstetrics and Gynecology. † Department of Medicine. ‡ Present address: Ministerio de Salud Publica del Ecuador, Quito, Ecuador. 0015-0282/98/$19.00 PII S0015-0282(98)00141-1
274
Conclusion(s): If idiopathic hirsutism is defined by the presence of hirsutism, regular ovulation, and normal androgen levels, only 17% of consecutive hirsute patients can be diagnosed with the disorder. Alternatively, if idiopathic hirsutism is based solely on the presence of hirsutism and regular ovulation, regardless of androgen levels, then 29% of the total hirsute population can be considered as having idiopathic hirsutism. Importantly, 40% of hirsute patients with a history of “regular cycles” were actually oligo/anovulatory, indicating the need to objectively assess ovulatory function in such patients. (Fertil Sterilt 1998;70:274 – 8. ©1998 by American Society for Reproductive Medicine.) Key Words: Hirsutism, androgens, polycystic ovary syndrome, ovulatory dysfunction
Some investigators have defined idiopathic hirsutism by the presence of hirsutism and regular menstrual cycles only, regardless of circulating androgen levels (26 –28); hence, it has been suggested that idiopathic hirsutism is one of the most frequent causes of hirsutism because between 50% and 70% of all hirsute women claim to have regular menses (1–3). However, a normal menstrual history does not exclude the presence of polycystic ovaries determined either pathologically (4) or sonographically (2, 5), elevated circulating androgen levels (1), or the presence of nonclassic adrenal hyperplasia (6). In response, other investigators define idiopathic hirsutism more strictly, as presenting with hirsutism, regular ovulation, and normal androgen levels (7–10). Although the presence of eumenorrheic hirsutism may be sufficient to establish the diagnosis of idiopathic hirsutism, it is also possible that many of these women actually suffer from
the polycystic ovary syndrome (PCOS). Because the diagnosis of PCOS hinges on the finding of ovulatory dysfunction (11), then regular ovulatory function should be confirmed in eumenorrheic hirsute women suspected of suffering from idiopathic hirsutism. To determine the extent to which ovulatory dysfunction affects hirsute eumenorrheic women—and hence establish the prevalence of idiopathic hirsutism among all hirsute women—we prospectively studied 133 consecutive hirsute women.
MATERIALS AND METHODS Protocol Between June 1994 and December 1996, all women who were not receiving hormonal treatment and who presented with hirsutism (as defined below) were included in this study. Serum was obtained and measured for total and free T, 17-hydroxyprogesterone (17-HP), and
DHEAS serum levels. Hyperandrogenemia was defined as a total or free T, and/or DHEAS level above the 95th percentile of normal. Based on 44 consecutive healthy eumenorrheic nonhirsute women aged 18 – 40 years, the values used were as follows: total T .89 ng/dL (3.08 nmol/L), free T .0.66 ng/dL (0.023 nmol/L), and DHEAS .2,750 ng/mL (7.46 mmol/L) (Azziz, unpublished data). Patients with a basal 17-HP level of .2 ng/mL underwent an adrenocorticotropic hormone stimulation test to exclude the possibility of 21-hydroxylase deficient nonclassic adrenal hyperplasia (NCAH), defined by a stimulated 17-HP level of .10 ng/ mL, as previously described (12). Hirsutism was defined by a score of $6, determined with a modification of the Ferriman-Gallwey (F-G) method, as previously described (13). This hirsutism score is obtained for only 5%– 6% of women in our general population (14). Women with cycles lengths of .35 days were classified as oligoovulatory without further testing. Women with cycles of #35 days in length were asked to maintain a daily basal body temperature (BBT) calendar and to obtain a day 22–24 progesterone level during one menstrual cycle. In these women oligoovulation was established if the day 22–24 progesterone level was ,4 ng/mL (with levels .4 ng/mL indicating prior ovulation and corpus luteum formation). The BBT was used to confirm that the progesterone level was actually obtained between days 22 and 24 of the menstrual cycle. The study was approved by the institutional review board of the University of Alabama at Birmingham.
Statistical Analysis Continuous variables between two groups were compared with use of the two-tailed unpaired t-test if values were normally distributed, or by the Mann-Whitney U test if values were parametrically distributed (i.e., F-G scores). Fractions (i.e., percent nonwhite) were compared by the x 2 test.
RESULTS Our study algorithm is depicted in Figure 1. Of 133 consecutive hirsute women initially included, one did not complete the study and was then excluded. Of the remaining 132 patients studied, 68 (52% of total) had a history of menstrual cycles of .35 days in length. Of the remaining 64 patients claiming to have regular menses at ,35-day intervals, 25 (39% or 19% of total) actually had oligo/anovulation by BBT and day 22–24 progesterone level. Finally, of the 39 patients with hirsutism and regular ovulatory function (i.e., 61% or 29% of the total), 22 (56% or 17% of the total) had total and free T, and DHEAS levels within normal limits; and the remaining 17 women had at least one elevated androgen level. One of these hirsute women with regular ovulation, but increased androgen levels, was diagnosed as having NCAH. Women with hirsutism and regular menses (n 5 64) had FERTILITY & STERILITYt
a lower mean body mass index (BMI), waist-hip ratio (WHR), and total and free T levels, but similar mean DHEAS levels, age, F-G score, or percent nonwhite, compared with those with irregular cycles (n 5 68) (Table 1). The only difference between hirsute women with “regular cycles” and oligo/anovulation (n 5 25), and those with normal ovulatory function (n 5 39), was a higher total T level in the former, although free T also tended to be higher (P 5 .078) (Table 2). Other than the expected differences in androgen levels, there were no statistically significant differences between ovulatory euandrogenic and hyperandrogenic hirsute patients (Table 3).
DISCUSSION The definition of idiopathic hirsutism has been a dynamic process, evolving in conjunction with our understanding of the various androgen excess disorders and the phenotypic heterogeneity of these patients. Early reports defined idiopathic hirsutism as “hirsutism of unknown cause,” regardless of whether they had irregular menstruation and ovulation and/or elevated circulating androgen levels (15–20). Subsequently, other investigators defined idiopathic hirsutism as hirsutism in the presence of “regular” menses (2, 21, 22). However, approximately 50%–75% of eumenorrheic hirsute women also demonstrate polycystic appearing ovaries on sonography (2, 3) and/or elevated circulating levels of either DHEAS or total and/or free T (1), suggestive of a more general endocrine abnormality. Furthermore, a history of “regular cycles” appears to be a poor predictor of regular ovulation, because 40% of our eumenorrheic hirsute patients had oligo/anovulation on closer investigation. Consequently, 40% of patients previously presumed to have idiopathic hirsutism can be considered as having PCOS, as previously defined (11). Whether the presence of hyperandrogenemia excludes the diagnosis of idiopathic hirsutism remains less clear, with some investigators diagnosing the disorder regardless of circulating androgen levels (2, 21, 22). Nonetheless, we do not agree and argue that the hirsutism cannot be considered “idiopathic” when patients have clear-cut hyperandrogenemia. Although the source of the elevated androgens may remain a mystery, the physical finding of hirsutism is directly explainable by the increased circulating androgen levels. In fact, .40% of our hirsute regularly ovulating patients had evidence of hyperandrogenemia. In agreement, other investigators reported that approximately one half of regularly cycling hirsute women had elevated levels of one or more androgens (1). Consequently, and in agreement with other investigators (7–10), we believe that idiopathic hirsutism should be defined only in those women with [1] hirsutism [2] normal ovulatory function, and [3] a normal androgen profile. Because the definition of idiopathic hirsutism has varied significantly over the past three decades, a precise estimate 275
FIGURE 1 Algorithm of 132 consecutive patients seen prospectively to determine the prevalence of idiopathic hirsutism. Patients with the polycystic ovary syndrome (oligoovulation, and hirsutism and/or hyperandrogenemia) and women with hirsutism and hyperandrogenemia but apparently normal ovulatory function, were considered as having functional androgen excess. If the diagnosis of idiopathic hirsutism is based solely on the presence of hirsutism and regular ovulation, regardless of androgen levels, then 29% of our total untreated hirsute population could be considered affected. However, only 17% of all hirsute patients can be diagnosed as having the disorder, if idiopathic hirsutism is defined as hirsutism, regular ovulation, and normal androgen levels. BBT 5 basal body temperature chart; P4 5 progesterone level.
of its prevalence as a cause of hirsutism has been difficult to establish. Consistent with other reports (1–3), approximately 50% of our consecutive hirsute women claimed to have “regular menses,” and it may then be supposed that idio276
Azziz et al.
Prevalence of idiopathic hirsutism
pathic hirsutism is a common cause of hirsutism. However, as noted above, 40% of these women had ovulatory dysfunction and would be considered as having PCOS. In our study, if idiopathic hirsutism, was defined as hirsutism in the presVol. 70, No. 2, August 1998
TABLE 1
TABLE 3
Comparison of hirsute patients with and without a history of regular menstrual cycles.
Comparison of hirsute ovulatory patients with and without elevated androgen levels.
History of menstrual cycle
Characteristic Age (y) BMI (kg/m2) Waist-hip ratio Total T (ng/dL) Free T (ng/dL) DHEAS (ng/mL) SHBG (nmol/L) 17-HP (ng/mL) F-G score No. of nonwhite subjects with indicated cycles (%)
Hirsute ovulatory patient group
“Regular cycles”
“Irregular cycles”
P value
30.1 6 9.8 29.1 6 8.3 0.79 6 0.10 71.2 6 16.2 0.74 6 0.24 2,815 6 1,336 184 6 88 1.14 6 1.11 9.8 6 3.2 8/65 (32)
27.9 6 7.4 32.2 6 7.6 0.83 6 0.07 80.1 6 20.0 0.88 6 0.32 2,514 6 1,426 183 6 52 2.52 6 7.67 9.6 6 4.0 11/68 (16)
NS .027 .018 .006 .006 NS NS NS NS NS
Characteristic Age (y) BMI (kg/m2) Waist-hip ratio T (ng/dL) Free T (ng/dL) DHEAS (ng/mL) SHBG (nmol/L) 17-HP (ng/mL) F-G score No. of nonwhite subjects with indicated cycles (%)
Elevated androgens
Normal androgens
P value
29.4 6 10.1 30.4 6 7.6 0.80 6 0.09 72.5 6 12.4 0.84 6 0.23 3,362 6 1,428 165 6 53 1.0 6 1.2 10.3 6 3.4 2/22 (9%)
34.1 6 9.9 27.1 6 6.5 0.81 6 0.15 60.1 6 13.4 0.51 6 0.08 1,950 6 627 219 6 54 0.8 6 0.8 8.8 6 3.1 3/17 (18%)
NS NS NS ,.005 ,.001 ,.001 ,.004 NS NS NS
Note: All values are means 6 SD unless otherwise indicated. BMI 5 body mass index; F-G score 5 hirsutism scored by a modified Ferriman-Gallwey method; 17-HP 5 17-hydroxyprogesterone; NS 5 not significant; SHBG 5 sex hormone-binding globulin.
Note: All values are means 6 1 SD unless otherwise indicated. BMI 5 body mass index; F-G score 5 hirsutism scored by a modified Ferriman-Gallwey method; 17-HP 5 17-hydroxyprogesterone; NS 5 not significant; SHBG 5 sex hormone-binding globulin.
ence of regular ovulation, regardless of androgen levels, then 29% (39 of 132) would be considered to have the disorder.
agreement, Carmina (23) noted that of 588 consecutive Italian patients complaining of hirsutism, 51% claimed regular cycles, but only 6% had regular ovulation, normal total and free T, and DHEAS levels, and no evidence of polycystic ovaries on ultrasonography.
Alternatively, if the more strict criteria were used whereby idiopathic hirsutism was considered only if patients had hirsutism, regular ovulation, and normal androgen levels, then only 17% would be considered as being affected. In
TABLE 2 Comparison of hirsute patients with a history of “regular menses” who do or do not have regular ovulation. History of “regular” menstrual cycle
Characteristic
Hirsute ovulatory
Hirsute oligo/ anovulatory
Age (y) 31.5 6 10.2 28.0 6 9.2 BMI (kg/m2) 29.0 6 7.2 29.1 6 10.0 Waist-hip ratio 0.80 6 0.11 0.78 6 0.07 Total T (ng/dL) 67.1 6 14.1 76.4 6 17.4 Free T (ng/dL) 0.70 6 0.24 0.80 6 0.22 DHEAS (ng/mL) 2,747 6 1,340 2,960 6 1,361 SHBG (nmol/L) 188 6 59 174 6 38 17-HP (ng/mL) 1.0 6 1.0 1.4 6 1.3 F-G score 9.6 6 3.3 10.1 6 3.1 No. of nonwhite subjects with indicated cycles (%) 5/39 (13%) 2/25 (8%)
P value NS NS NS ,.025 NS (.078) NS NS NS NS NS
Note: All values are means 6 1 SD unless otherwise indicated. BMI 5 body mass index; F-G score 5 hirsutism scored by a modified Ferriman-Gallwey method; 17-HP 5 17-hydroxyprogesterone; NS 5 not significant, SHBG 5 sex hormone-binding globulin.
FERTILITY & STERILITYt
Overall, it appears that idiopathic hirsutism affects between 15% and 30% of hirsute patients seen, depending on whether normal androgen levels are required for the definition (Fig. 1). Because menstrual history does not accurately reflect ovulatory status in hirsute patients, it is imperative that ovulatory function be documented clearly, perhaps by using daily BBT and/or a luteal progesterone level. Although not all investigators agree, we believe that idiopathic hirsutism should be a diagnosis of exclusion in which ovulatory disorders, hyperandrogenemia, and other androgen excess disorders (e.g., NCAH) are ruled out. All other patients should be defined as having “functional androgen excess” (Fig. 1), and include both patients with PCOS and those with hirsutism and hyperandrogenemia, but apparently normal ovulatory function. Because the sole measurement of a total T serum level is grossly insensitive for the detection of hyperandrogenemia (24 –27), we suggest that only those patients who have demonstrated at a minimum normal DHEAS, and total and free T levels be included. Nonetheless, it remains important to better define idiopathic hirsutism, perhaps by determining the differences between patients with and without hyperandrogenemia, such as risk of developing frank ovulatory dysfunction (i.e., PCOS) and/or metabolic aberrations. 277
References 1. Mehta A, Matwijiw I, Taylor PJ, Salamon EA, Kredentser JV, Faiman C. Should androgen levels be measured in hirsute women with normal menstrual cycles? Int J Fertil 1992;37:354 –7. 2. Jahanfar S, Eden JA. Idiopathic hirsutism or polycystic ovary syndrome? Aust NZ J Obstet Gynaecol 1993;33:414 – 6. 3. O’Driscoll JB, Mamtora H, Higginson J, Pollock A, Kane J, Anderson DC. A prospective study of the prevalence of clear-cut endocrine disorders and polycystic ovaries in 350 patients presenting with hirsutism or androgenic alopecia. Clin Endocrinol (Oxf) 1994;41:231– 6. 4. Raj SG, Talbert LM. Polycystic ovarian disease. Obstet Gynecol Annu 1984;13:261–73. 5. Adams J, Polson DW, Franks S. Prevalence of polycystic ovaries in women with anovulation and idiopathic hirsutism. Br Med J 1986;293: 355–9. 6. Azziz R, Dewailly D, Owerbach D. Non-classic adrenal hyperplasia: current concepts. J Clin Endocrinol Metab 1994;78:810 –5. 7. Mowszowicz I, Melanitou E, Doukani A, Wright F, Kuttenn F, Mauvais-Jarvis P. Androgen binding capacity and 5a-reductase activity in pubic skin fibroblasts from hirsute patients. J Clin Endocrinol Metab 1983;56:1209 –13. 8. Serafini P, Lobo RA. Increased 5a reductase activity in idiopathic hirsutism. Fertil Steril 1985;43:74 – 8. 9. Couzinet B, Pholsena M, Young J, Schaison G. The impact of a pure anti-androgen (flutamide) on LH, FSH, androgens and clinical status in idiopathic hirsutism. Clin Endocrinol (Oxf) 1993;39:157– 62. 10. Tolino A, Petrone A, Sarnacchiaro F, Cirillo D, Ronsini S, Lombardi G, et al. Finasteride in the treatment of hirsutism: new therapeutic perspectives. Fertil Steril 1996;66:61–5. 11. Zawadzki JK, Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach. In: Dunaif A, Givens JR, Haseltine F, Merriam GR, editors. Polycystic ovary syndrome. Boston: Blackwell Scientific Publications, 1992:377– 84. 12. Azziz R, Zacur HA. 21-hydroxylase deficiency in female hyperandrogenism: screening and diagnosis. J Clin Endocrinol Metab 1989;69: 577– 84. 13. Hatch R, Rosenfield RL, Kim MH, Tredway D. Hirsutism: Implications, etiology, and management. Am J Obstet Gynecol 1981;140:815– 30. 14. Knochenhauer ES, Kahsar-Miller MD, Boots LR, Key TJ, Azziz R. Prevalence of hirsutism among unselected Black and Caucasian women in the southeastern United States (U.S.) [abstract P1-357]. Program and
278
Azziz et al.
Prevalence of idiopathic hirsutism
15. 16. 17. 18. 19. 20.
21. 22.
23.
24. 25. 26. 27.
abstracts of the 79th Annual Meeting of the Endocrine Society. Minneapolis, (MN): The Endocrine Society, 1997:224. Bardin WC, Lipsett MB. Testosterone and androstenedione blood production rates in normal women and women with idiopathic hirsutism or polycystic ovaries. J Clin Invest 1967;46:891–7. Casey JH, Nabarro JDN. Plasma testosterone in idiopathic hirsutism, and the changes produced by adrenal and ovarian stimulation and suppression. J Clin Endocrinol 1967;27:1431–5. Thomas JP, Oake RJ. Androgen metabolism in the skin of hirsute women. J Clin Endocrinol Metab 1974;38:19 –22. Kirschner M, Zucker R, Jespersen D. Idiopathic hirsutism—an ovarian abnormality. N Engl J Med 1976;294:637– 41. Ettinger B, Goldfield EB, Burrill KC, Von Werder K, Forsham PH. Plasma testosterone stimulation-suppression dynamics in hirsute women. Am J Med 1973;54:195–200. Pugeat M, Forest MG, Nisula BC, Corniau J, De Peretti E, Tourniaire J. Evidence of excessive androgen secretion by both the ovary and the adrenal in patients with idiopathic hirsutism. Obstet Gynecol 1982;59: 46 –51. McKenna TJ, Miller RB, Liddle GW. Plasma pregnenolone and 17-OH pregnenolone in patients with adrenal tumors, ACTH excess, or idiopathic hirsutism. J Clin Endocrinol Metab 1977;44:231– 6. Rittmaster RS, Thompson DL. Effect of leuprolide and dexamethasone on hair growth and hormone levels in hirsute women: the relative importance of the ovary and the adrenal in the pathogenesis of hirsutism. J Clin Endocrinol Metab 1990;70:1096 –1102. Carmina E. Prevalence of idiopathic hirsutism [abstract no. P-286]. Program and abstracts of the 53rd Annual Meeting of the American Society for Reproductive Medicine. Cincinnati (OH): American Society for Reproductive Medicine, 1997:S229. Boots LR, Potter S, Potter HD, Azziz R. Measurement of total serum testosterone level using commercial kits: high degree of variability and inaccuracy. Fertil Steril 1998;69:286 –92. Vermeulen A, Stoica T, Verdonck L. The apparent free testosterone concentration, an index of androgenicity. J Clin Endocrinol 1971;33: 759 – 67. Mathur RS, Moody LO, Landgrebe S, Williamson HO. Plasma androgens and sex hormone binding globulin in the evaluation of hirsute females. Fertil Steril 1981;35:29 –35. Carter GD, Holland SM, Alaghband-Zadeh J, Rayman G, DorringtonWard P, Wise PH. Investigation of hirsutism: testosterone is not enough. Ann Clin Biochem 1983;20:262–3.
Vol. 70, No. 2, August 1998