When should a patient who menstruates normally be diagnosed as having polycystic ovary syndrome, and what clinical differences exist between these somewhat atypical women and those with classic PCOS?
When should a patient who menstruates normally be diagnosed as having polycystic ovary syndrome, and what clinical differences exist between these somewhat atypical women and those with classic PCOS?
Normal menstruation is common in women who have hyperandrogenism, hirsutism, or both. In fact, in several studies that looked at different populations, 50% to 70% of hirsute women had normal menses.1-4 In our experience, for instance, 50.8% of 588 hirsute women had normal menstrual cycles.4
In the past, these menstruating women would have been diagnosed with idiopathic hirsutism, but there's now general agreement that this label should be reserved for hirsute women with normal menses and normal circulating androgen levels.5 When carefully evaluated, however, as many as 90% of hirsute women with normal menses have elevated androgen levels and therefore cannot be included in the group of women with idiopathic hirsutism.3,4
Some of these women may actually have polycystic ovary syndrome (PCOS). Of course, the main problem is that experts don't completely agree on the clinical or biological characteristics needed to diagnose PCOS. In the past, most clinicians diagnosed PCOS on the basis of hyperandrogenism and chronic anovulation, provided that some uncommon, but clearly defined forms of hyperandrogenism (tumors, Cushing's syndrome, and adrenal enzymatic deficiencies) had been excluded.6,7 If we use this definition, the syndrome should only be diagnosed in hyperandrogenic womenwith or without hirsutismwho have normal menses but are anovulatory. It is not uncommonand several studies have reportedthat anovulation may be present in women with normal menses.3,4 And more recently, we and others have provided evidence that the diagnosis of PCOS can be reached in hyperandrogenic ovulatory women.8-10
It's likely then that normal menses in PCOS is more common than we once believed. In this review, we'll discuss when to diagnose PCOS in women with normal menses and what differences there are between these patients and those with classic PCOS, who have irregular menses.
In our experience, 15% to 21% of hyperandrogenic women with normal menses are anovulatory and have to be considered as affected by PCOS.4,8 In 1963, after reviewing 1,079 patients, Goldzieher and colleagues reported that 12% of women with PCOS menstruate normally.11 Similar data from other large studies are reported in Table 1.
The diagnosis of PCOS in these women with normal menses is easy. It is sufficient to show the presence of anovulation (by low-serum progesterone) and hyperandrogenism (increased serum levels of testosterone or DHEAS). Of course, a single anovulatory cycle isn't enough to make the diagnosis; chronic anovulation should be ascertained. In clinical practice, we measure serum progesterone during day 22 to 23 of the cycle in all hyperandrogenic women with normal menses. If progesterone levels are below 3 ng/mL, one more cycle is studied and the diagnosis of PCOS is made if this cycle is also anovulatory. Most of these patients present polycystic ovaries at ultrasound as well, but pelvic sonography is not required for the diagnosis. Serum levels of 17-hydroxyprogesterone (17-OHP), measured during the follicular phase, should also be evaluated to exclude nonclassic 21-OH deficiency.
It is not clear why some women with PCOS also have normal menses, in spite of being anovulatory. In our study, hyperandrogenic anovulatory women with normal menses, compared to classic PCOS with irregular menses, had similar polycystic ovarian morphology and similarly increased response to a GnRH agonist acute test. Serum androgen levels were elevated and no different from those found in patients with irregular menses.8 Interestingly, patients with anovulatory PCOS and normal menses seemed to be leaner and have lower insulin and gonadotropin levels than those with classic PCOS. Therefore, differences in gonadotropin and/or insulin secretion could be responsible for the varying length of the ovarian cycle in anovulatory women. However, a large study comparing anovulatory patients with normal and irregular menses is yet to be done, and, in our experience, there is a large overlap between gonadotropin and insulin levels in individual PCOS women with normal and irregular menses.
Treatment of infertility and hirsutism in these women is no different from that of women with classic PCOS, and of course, clinicians should take steps to reduce cardiovascular risks in all patients with insulin resistance and altered serum lipids.
As we mentioned previously, for many years most authors have been convinced that chronic anovulation is needed for the diagnosis of PCOS. However, many ovulatory hyperandrogenic women present with typical PCOS features.8,10 In fact, 45% to 50% of ovulatory hyperandrogenic women have polycystic ovaries on ultrasound, increased ovarian 17-OHP, and androstenedione response to a GnRH-agonist, suggesting ovarian hyperandrogenism. Moreover, in our studies we found that the same patients had mild insulin resistance and altered lipid profiles, similar to patients with classic (anovulatory) PCOS.7,10 So evidently some ovulatory hyperandrogenic women have the same clinical and endocrine characteristics as women with classic (anovulatory) PCOS and there is no reason to make the diagnosis of PCOS only in patients with chronic anovulation.
If we believe that the main characteristics of PCOS are hyperandrogenism and insulin resistance, both elements are not necessarily linked to anovulation. In fact, hyperandrogenism is often associated with normal ovulatory cycles. In our study, 40% of hirsute women had hyperandrogenism but were still ovulating.4 Similar data have been presented by other investigators.1 Others have reported that ovulatory hyperandrogenic women have similar androgen levels as anovulatory hyperandrogenic patients.14 While the prevalence of ovulation in different hyperandrogenic populations varies, it is clear that hyperandrogenism per se does not automatically cause anovulation.
Also keep in mind that hyperinsulinemia is not necessarily associated with anovulation. For example, while insulin resistance and hyperinsulinemia are well-known components of obesity, most obese women have normal ovulatory cycles and normal fertility.15,16 Moreover, insulin resistance is also a common finding in lean apparently normal women and in insulin resistance syndrome or syndrome X.17,18
It's possible that hyperinsulinemia induces anovulation only when hyperandrogenism is present. This hypothesis is supported by the finding that obese women who have altered menstrual patterns also have higher androgen levels than eumenorrheic obese women.19 However, our data have shown that in this situation, anovulation is not a necessary consequence. In fact, our hyperandrogenic ovulatory patients with polycystic ovaries had insulin resistance as well.10 This suggests that the association between insulin resistance and hyperandrogenism does not necessarily precipitate anovulation.
Patients with ovulatory PCOS present clinicians with a few unique problems:
How does one make the diagnosis? You should base the diagnosis on the finding of ovulatory cycles in hyperandrogenic women who have polycystic ovaries. Therefore, in ovulatory hyperandrogenic women, it is particularly important to evaluate ovarian morphology with pelvic sonography. The existence of polycystic ovaries is sufficient for the diagnosis of PCOS in these women because we have previously shown that in ovulatory hyperandrogenic women, there is a correlation between ovarian response to a GnRH agonist test and ovarian morphology by pelvic sonography.10 The sonographic diagnosis should be based on the classic criteria of Adams and associates.20 Figure 1 provides a scheme for diagnosing PCOS in hyperandrogenic women with normal menses.
What other tests should clinicians perform to arrive at the diagnosis of ovulatory PCOS? An evaluation of insulin sensitivity by a simple mathematical method based on insulin and glucose basal levels such as HOMA-IR or QUICKI should be performed.21,22 Moreover, cardiovascular risk should be assessed. Therefore glucose tolerance (by oral glucose tolerance test) and serum lipids (total cholesterol, HDL and LDL cholesterol, and triglycerides) should be measured.
If any of these values are altered, the patient should be carefully treated with nutritional therapy and, if needed, with insulin sensitizing agents. Consider metformin if diet fails to resolve insulin resistance or to improve lipids. If all values are normal, on the other hand, the patient should not be treated but insulin sensitivity, glucose tolerance, and serum lipids should be measured every 2 years.7
Is fertility normal in patients with ovulatory PCOS? We can't offer a definitive answer to that question at the moment. We have regarded these patients as fertile but recent studies have shown that women with ovulatory PCOS have some alterations in their early luteal phase.9 It's unclear if this might impair fertility.
We have known for several years that many apparently normal, fertile women have polycystic ovaries.23-25 Up to 20% of normal women from different countries have polycystic ovaries on ultrasound. We don't know what causes the altered ovarian morphology in many of these apparently normal women, but their fertility is normal. In fact, studying women who were ovum donors and had proven fertility, we found that some of these women have polycystic ovaries.26 Most of these women have normal serum androgens as well, although we have shown that they may have a slightly increased ovarian response to a GnRH agonist acute test.27,28 These apparently normal women also have some signs of very mild insulin resistance.27,28
This is probably a third group of women who have a very mild defect of insulin sensitivity. Because of their normal fertility and the absence of basal hyperandrogenism, however, we don't believe these women should be labeled as having PCOS.
The most diffuse theory on the pathogenesis of PCOS is based on the presence of two different defectshyperandrogenism and insulin resistanceboth of which are probably inherited.7,29,30 While both are needed for the syndrome to come about, differing severities of these two defects may be present in individual patients.
These underlying defects are probably responsible for a wide spectrum of conditions that are all part of what we call PCOS, but each of which may have very different clinical and endocrine presentations. In this spectrum, ovulatory PCOS represents the milder form of PCOS but is still able to induce sub-fertility, hirsutism, and increased cardiovascular risk. The differences between these milder forms of PCOS and the classical disorder are summed up in Table 2.
1. Mehta A, Matwijiw I, Taylor PJ, et al. Should androgen levels be measured in hirsute women with normal menstrual cycles? Int J Fertil. 1992;37:354-357.
2. Jahanfar S, Eden JA. Idiopathic hirsutism or polycystic ovary syndrome? Austr N Z J Obstet Gynaecol. 1993;33:414-416.
3. Azziz R, Waggoner WT, Ochoa T, et al. Idiopathic hirsutism: an uncommon cause of hirsutism in Alabama. Fertil Steril. 1998;70:274-278.
4. Carmina E. Prevalence of idiopathic hirsutism. Eur J Endocrinol. 1998;139:421-423.
5. Azziz R, Carmina E, Sawaya ME. Idiopathic hirsutism. Endocr Rev. 2000;21:347-362.
6. Zawadzki JK, Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach. In: Dunaif A, Givens JR, Haseltine FP, et al., eds. Polycystic Ovary Syndrome. Boston, Mass: Blackwell Scientific Publications; 1992:377-384.
7. Lobo RA, Carmina E. The importance of diagnosing the polycystic ovary syndrome. Ann Intern Med. 2000; 132:989-993.
8. Carmina E, Lobo RA. Do hyperandrogenic women with normal menses have polycystic ovary syndrome? Fertil Steril. 1999;71:319-322.
9. Joseph-Horne R, Mason H, Batty S, et al. Luteal phase progesterone excretion in ovulatory women with polycystic ovaries. Hum Reprod. 2002;17:1459-1463.
10. Carmina E, Lobo RA. Polycystic ovaries in hirsute women with normal menses. Am J Med. 2001;111:602-606.
11. Goldzieher JW, Axelrod LR. Clinical and biochemical features of polycystic ovarian disease. Fertil Steril. 1963;14:631-653.
12. Lobo RA, Carmina E. Polycystic Ovary Syndrome. In: Lobo RA, Mishell DR Jr, Paulson RJ, et al., eds. Mishell's Textbook of Infertility, Contraception, and Reproductive Endocrinology. Boston, Mass: Blackwell Scientific Publications; 1997;363-383.
13. Balen AH, Conway GS, Kaltsas G, et al. Polycystic ovary syndrome: the spectrum of the disorder in 1741 patients. Hum Reprod. 1995;10:2107-2111.
14. Franks S. Polycystic ovary syndrome: a changing perspective. Clin Endocrinol. 1989;31:87-120.
15. Caro JF. Clinical review 26: insulin resistance in obese and nonobese man. J Clin Endocrinol Metab. 1991;73:691-695.
16. Glass AR. Endocrine aspects of obesity. Med Clin North Am. 1989;73:139-160.
17. Reaven GM. Role of insulin resistance in human disease. Diabetes. 1988;37:1595-1607.
18. Ruderman NB. The "metabolically-obese" normal-weight individual. Am J Clin Nutr. 1981;34:1617-1621.
19. Zhang YW, Stern B, Rebar RW. Endocrine comparison of obese menstruating and amenorrheic women. J Clin Endocrinol Metab. 1984;58:1077-1083.
20. Adams J, Polson DW, Franks S. Prevalence of polycystic ovaries in women with anovulation and idiopathic hirsutism. Br Med J (Clin Res Ed). 1986;293:355-359.
21. Haffner SM, Miettinen H, Stern MP. The homeostasis model in the San Antonio Heart Study. Diabetes Care. 1996;19:1138-1141.
22. Katz A, Nambi SS, Mather K, et al. Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab. 2000;85:2402-2410.
23. Polson DW, Adams J, Wadswoth J, et al. Polycystic ovariesa common finding in normal women. Lancet. 1988;1:870-872.
24. Clayton RN, Ogden V, Hodgkinson J, et al. How common are polycystic ovaries in normal women and what is their significance for the fertility of the population? Clin Endocrinol (Oxf). 1992;37:127-134.
25. Abdel Gadir A, Khatim MS, Mowafi RS, et al. Implications of ultrasonically diagnosed polycystic ovaries. I. Correlations with basal hormonal profiles. Hum Reprod. 1992;7:453-457.
26. Wong LI, Morris RS, Lobo RA, et al. Isolated polycystic morphology in ovum donors predicts response to ovarian stimulation. Hum Reprod. 1995;10:524-528.
27. Carmina E, Wong L, Chang L, et al. Endocrine abnormalities in ovulatory women with polycystic ovaries on ultrasound. Hum Reprod. 1997;12:905-909.
28. Chang PL, Lindheim SR, Lowre C, et al. Normal ovulatory women with polycystic ovaries have hyperandrogenic pituitary-ovarian responses to gonadotropin-releasing hormone-agonist testing. J Clin Endocrinol Metab. 2000;85:995-1000.
29. Carmina E, Lobo RA. Polycystic ovary syndrome (PCOS): arguably the most common endocrinopathy is associated with significant morbidity in women. J Clin Endocrinol Metab. 1999;84:1897-1899.
30. Legro RS. Polycystic ovary syndrome: the new millennium. Mol Cell Endocrinol. 2001;184:87-93.
Dr. Carmina presents a provocative and important argument on the relationship between PCOS and normal menses. While significant strides have been made in elucidating many of the associated features of polycystic ovary syndrome, much remains unclear.
Dr. Carmina alludes to an important fact that "normal menses" doesn't exclude the presence of oligo-ovulation, particularly in the presence of other hyperandrogenic features like hirsutism and hyperandrogenemia. Up to 40% of hirsute women who menstruate regularly are anovulatory when evaluated more closely by luteal phase progesterone levels and basal body temperature charts.1 Hence, it is important for the reader to always distinguish between "normal menses" and "normal ovulatory function". As the author points out, we still don't know why some women continue to have vaginal bleeding at regular intervals despite being anovulatory.
It is likely, however, that the presence of vaginal bleeding relates primarily to differences in endometrial factors, either intrinsic or in response to other extrauterine circulating factors. In the study by Carmina and Lobo, for instance, women who were having regular vaginal bleeding had lower circulating insulin levels than those who did not.2 It is possible that insulin, directly or indirectly, alters endometrial integrity and growth in these women. Further studies are needed in this area, particularly in regards to the clinical implications of regular menstrual bleeding in the face of anovulation in women who otherwise have PCOS.
The question of whether ovulatory dysfunction is required for the diagnosis of PCOS is an important one. The informal proceedings of a 1990 NICHD-sponsored conference on the subject noted that most participants in the conference felt that PCOS should be diagnosed by the presence of: (a) oligo-anovulation, (b) clinical and/or biochemical hyperandrogenism (e.g., hirsutism and/or hyperandrogenemia), after (c) the exclusion of related disorders, such as nonclassic adrenal hyperplasia, Cushing's syndrome, and androgen-secreting neoplasms.3
However, it is also very clear that this disorder is quite heterogeneous. For example, many women will have varying degrees of hirsutism and other dermatologic signs of androgen excess, despite having relatively similar circulating androgen levels. Likewise, many patients may not have "polycystic ovaries" on ultrasonography, nor will they have gross evidence of gonadotropic abnormalities on evaluation.
Finally, while up to 70% of women with PCOS demonstrate some degree of insulin resistance compared to weight-matched controls, up to a third of these patients may not have any evidence of insulin resistance whatsoever.4,5 Thus, it is conceivable that there is also considerable variation in the degree of hypothalamic-pituitary-ovarian axis abnormality. We have observed that approximately 16% of our patients with functional androgen excess have a phenotype similar to that described by Dr. Carmina; that is, the presence of hirsutism, hyperandrogenemia, and normal menstrual and ovulatory function (at least in the cycles evaluated).
A greater problem arises when we begin to consider hyperandrogenic ovulatory women with polycystic ovaries as having PCOS. Carmina and Lobo indicate that on average their hyperandrogenic ovulatory patients with polycystic ovaries (labeled "ovulatory PCOS") had higher circulating insulin levels, lower glucose-to-insulin ratios, and higher body mass compared to hirsute ovulatory women who had normal androgens; that is, 'idiopathic hirsutism', although ovarian morphology by ultrasound was not reported.6 In a previous study these investigators reported that women with "ovulatory PCOS'' generally had lesser degrees of hyperinsulinism than patients with frank PCOS.7 Thus, it is unclear whether patients who have hyperandrogenism (i.e., hirsutism and/or hyperandrogenemia), polycystic ovaries, but regular ovulation and menstruation actually have PCOS. Do they represent an intermediate stable phenotype between idiopathic hirsutism and PCOS, or do they represent an early form of PCOS that will progress over time, or is this an entirely different disorder? These questions remain unanswered and need to be investigated by performing long-term follow-up and genetic or familial studies.
Overall, it's premature to begin to define a new disorder, namely "ovulatory PCOS". While Dr. Carmina goes to great lengths to explain how this disorder may be diagnosed, it's still unclear whether this is actually a disorder of similar concern as frank PCOS. Thus, we should be somewhat cautious before labeling these patients as having PCOS, particularly since the diagnosis has significant medical and reproductive implications.
Nonetheless, it is clear that this is an area that requires continued and aggressive study to elucidate the full spectrum of the disorder, its clinical implications, and potential preventive strategies. Dr. Carmina's presentation provides significant "food for thought" in this regard. We should not forget that PCOS is the single most common endocrine disorder of reproductive-aged women, with significant reproductive and metabolic morbidity.
1. Azziz R, Waggoner WT, Ochoa T, et al. Idiopathic hirsutism: an uncommon cause of hirsutism in Alabama. Fertil Steril. 1998;70:274-278.
2. Carmina E, Lobo RA. Do hyperandrogenic women with normal menses have polycystic ovary syndrome? Fertil. Steril. 1999;71:319-322.
3. Zawdaki JK, Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rationale approach. In: Dunaif A, Givens JR, Haseltine F, et al, eds. Polycystic Ovary Syndrome. Boston, Mass: Blackwell Scientific Publications; 1992:377-384.
4. Dunaif A, Segal KR, Futterweit W, et al. Profound peripheral insulin resistance, independent of obesity, in polycystic ovary syndrome. Diabetes. 1989;38:1165-1174.
5. Legro RS, Finegood D, Dunaif A. A fasting glucose to insulin ratio is a useful measure of insulin sensitivity in women with polycystic ovary syndrome. J Clin Endocrinol Metab. 1998;83:2694-2698.
6. Carmina E, Lobo RA. Polycystic ovaries in Hirsute women with normal menses. Am J Med. 2001;111:602-606.
7. Carmina E, Wong L, Chang J, et al. Endocrine abnormalities in ovulatory women with polycystic ovaries on ultrasound. Hum Reprod. 1997;12:905-909.
Enrico Carmina. Diagnosing PCOS in women who menstruate regularly.
Jul. 1, 2003;48:53-64.