The prevalence of polycystic ovaries in women with infertility

E. Kousta1,3, D.M. White2, E. Cela2, M.I. McCarthy1 and S. Franks2

1 Section of Endocrinology and Metabolic Medicine and 2 Department of Reproductive Science and Medicine, Imperial College School of Medicine, St Mary's Hospital, London W2 1PG, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Polycystic ovaries (PCO) are highly prevalent in women presenting with hirsutism or recurrent miscarriage but the functional significance of PCO in ovulatory women presenting with infertility remains unclear. We examined the prevalence of PCO, on ultrasonography, among women presenting with infertility. Among 289 couples classified in four main diagnostic categories, PCO were found in 81 (83%) of 98 anovulatory patients, 40 (53%) of 76 patients whose partners had sperm dysfunction, 26 (50%) of 52 patients with tubal disease and in 28 (44%) of 63 patients with unexplained infertility. By comparison, in a control group of 67 parous volunteers, 19 (28%) were found to have PCO. PCO patients with unexplained infertility had higher midfollicular luteinizing hormone and testosterone compared with the group with normal ovaries. The prevalence of PCO was significantly higher in each of the infertility groups than in controls, and a similar tendency (not significant) was observed among women with unexplained infertility. Ovulatory PCO women with infertility had higher testosterone concentrations in comparison with PCO controls. In summary, the prevalence of PCO among ovulatory women with infertility is higher than that in the normal population, suggesting that PCO may, perhaps by virtue of an effect of hyperandrogenaemia, contribute to the causes of subfertility in women with regular menses.

Key words: infertility/ovulation/polycystic ovaries


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
It is estimated that subfertility affects up to 14% of couples in Western society (Templeton, 1992Go). Precise determination of the full range of factors contributing to infertility in any given clinical situation requires investigation of both partners and is only possible in specialist clinics. Patterns of causes will be affected by the couples seeking referral, by referral policy, resources and specialist interests of the clinic and, finally, the diagnostic criteria employed (Hull, 1992Go). The prevalence of unexplained infertility among the couples with subfertility who have been investigated in developed countries varies between 8 and 31% (Hull, 1992Go).

Polycystic ovary (PCO) syndrome is a very common endocrine disorder. Among women who present with oligomenorrhoea ~87% have PCO, and, of women with regular menstrual cycles who present with hirsutism, 92% have PCO (Adams et al., 1986Go). Over 50% of patients who present with recurrent miscarriage have PCO (Sagle et al., 1988Go; Regan et al., 1990Go). It has been shown that 21–23% of the normal female population have PCO on ultrasound scan (Polson et al., 1988Go; Clayton et al., 1992Go; Farquhar et al., 1994Go). Despite the high prevalence in the normal population, PCO appear to represent a distinct morphological entity (Adams et al., 1986Go). There is a wide spectrum of clinical and biochemical features associated with PCO. There is increasing evidence that PCO is an oligogenic disorder, with the interaction of a small number of key genes with environmental factors determining the clinical and biochemical manifestation (Franks et al., 1997Go). The functional significance of PCO in women with ovulatory cycles remains to be clarified.

The aim of this study was to define the prevalence of PCO in women with infertility and to compare the endocrine profiles in women with and without PCO within infertility groups and with control subjects with and without PCO.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
During the 2 year period from January 1993 to December 1994, 577 couples presented to the Infertility Clinic at St Mary's Hospital, London. Among them 289 couples were classified into four diagnostic categories (anovulatory, sperm dysfunction, tubal disease and unexplained infertility). In the remaining 288 couples, investigations were not completed to allow assignment to a diagnostic category (n = 249), or ovarian morphology was uncertain (n = 9), or had multiple or other factors contributing to their subfertility and could not be assigned to one of the four diagnostic categories (n = 30).

All couples were investigated using the established clinic protocol. The female partner underwent cycle monitoring. This included a series of ultrasound scans and timed hormone measurements to assess ovarian morphology and confirm ovulation. The first ultrasound scan was performed on day 8 of the cycle and serum luteinizing hormone (LH), follicle stimulating hormone (FSH) and testosterone concentrations were measured. Ovarian morphology was assessed and PCO were diagnosed using established criteria (Adams et al., 1986Go), by an experienced team of ultrasound operators. The reproducibility of ultrasound diagnosis had been previously assessed for every operator. Repeat ultrasound scans were undertaken at 3–4 day intervals to assess follicular growth and estimate the day of ovulation. Ovulation was confirmed by a serum progesterone measurement 7 days post ovulation. Serum LH, FSH and testosterone concentrations were measured at this time. In a subsequent cycle, mid-luteal serum progesterone, gonadotrophins and testosterone measurements were repeated. Ovulation was considered satisfactory when serum progesterone measurements were >30 nmol/l in two successive cycles.

Tubal patency was confirmed by laparoscopy and/or hysterosalpingography (HSG). The majority of patients underwent laparoscopy as the initial investigation of tubal function, an HSG being undertaken when additional information was required. Tubal function was considered normal when the patency of both tubes had been confirmed. Only four patients from our unexplained infertility group had an HSG only, because of specific clinical indications relevant only to them (i.e. previous abdominal surgery with adhesions). Of the rest of the patients, three had mild endometriosis on laparoscopy. We thought that the mild endometriosis in these few subjects was not an adequate explanation for unexplained infertility.

The male partner underwent a semen analysis. This included a full separation test as well as screening for antisperm antibodies, white cells and sperm precursors. The semen analysis was considered normal if the sperm concentration was >40x106/ml, motility >50%, and <10% abnormal forms (World Health Organization, 1987Go). In the separation test, the normal range established by our laboratory as suitable for ovulation induction/intrauterine insemination (IUI) therapy was used. The criteria were that in a volume of 1 ml, there were >4x106 spermatozoa with a progressive motility of >70%.

We included in the `unexplained infertility' group those couples in whom all investigations were normal, so that they fulfilled the following criteria: (i) primary or secondary infertility for longer than 1 year, (ii) regular cycles in the female partner with ovulation confirmed in two successive cycles, (iii) a normal laparoscopy or HSG in the female partner, (iv) a normal semen analysis and separation test in the male partner.

We recruited 67 women, as a control group, from manual or computerized antenatal databases of St Mary's and Hammersmith & Queen Charlotte's hospitals (London, UK). They were sent a letter and were invited to attend the hospital for an ultrasound scan and a single blood test. These women considered themselves to be normal, conceived spontaneously and had not found it necessary to consult a doctor for menstrual disturbance, infertility or hirsutism. The history was checked and all women were examined for assessment of hirsutism. An ultrasound scan was performed on day 8 of the cycle to assess ovarian morphology and serum LH, FSH and testosterone concentrations were measured. Those women who were on the oral contraceptive (n = 25) were scanned during the `pill-free' interval of oral contraceptive use and no hormone assessment was performed. Mean (SD) age was 35.7 (5.2) years and body mass index (BMI) was 24.5 (4.3) kg/m2.

Results are expressed as mean (SD). Statistical analyses were performed in SPSS 7.5 for Windows using Mann–Whitney and {chi}2 tests, as appropriate.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Among 289 couples entered into the study and classified according to the reason for their infertility, 98 (34%) were diagnosed to be anovulatory, 76 (26%) had sperm dysfunction, 52 (18%) had tubal disease, and 63 (22%) had unexplained infertility. The prevalence of PCO was assessed in these patients.

Among the 98 anovulatory patients 17 (17%) had normal ovaries and 81 (83%) had PCO (Table IGo). In this group, there was no difference in mean BMI, day 8 FSH, testosterone and duration of infertility between women with normal ovaries and those with PCO. Mean age was significantly higher in the group with normal ovaries and day 8 LH significantly higher in the PCO group (Table IIGo). In the 17 patients with anovulation and normal ovarian morphology we could detect no definite endocrine abnormality. Hyperprolactinaemia or raised FSH concentrations were excluded. So these were patients with normal ovarian morphology but unsatisfactory quality of ovulation and therefore they required induction of ovulation.


View this table:
[in this window]
[in a new window]
 
Table I. Prevalence of polycystic ovaries (PCO) in the four diagnostic infertility groups and in the control group
 

View this table:
[in this window]
[in a new window]
 
Table II. Clinical and hormone data in the four diagnostic categories of infertility and in the control group. Comparison is made (i) between the normal ovaries group and the polycystic ovary (PCO) group within each diagnostic category and (ii) between each of the four PCO infertility groups and the PCO control group
 
Among the 76 patients, whose partners had an abnormal semen analysis, 36 (47%) had normal ovaries and 40 (53%) had PCO. In this group, there was no difference in BMI, day 8 LH, FSH, testosterone and duration of infertility between women with normal ovaries and those with PCO. Mean age was significantly higher in the group with normal ovaries (Table IIGo). Of the 76 men with sperm dysfunction, 36 had a total count of <0.5x106 in the separation test, but only six had a separation count of zero.

Among 52 patients with tubal disease, 26 (50%) had normal ovaries and 26 (50%) had PCO. Of these 52 patients, 25 had bilateral tubal damage and 27 had significant tubal damage but there was patency in one tube. There was no difference in BMI, day 8 LH and FSH between women with normal ovaries and women with PCO. Mean day 8 testosterone was significantly higher in the PCO group, and mean age and duration of infertility were significantly higher in the group with normal ovaries (Table IIGo).

Among the 63 women with unexplained infertility, 35 (56%) had normal ovaries and 28 (44%) had PCO. Of these 63 women, 46 (73%) were European in origin, five (8%) Afro-Caribbean, 12 (19%) of other ethnic origin (nine Middle-eastern, two Indian and one Chinese). Among the 35 women with normal ovaries 28 (80%) were European, two (6%) Afro-Caribbean and five (14%) of other ethnic origin (four Middle-eastern and one Chinese). Among the 28 women with PCO, 18 (64%) were European, three (11%) Afro-Caribbean and seven (25%) other (five Middle-eastern and two Indians). There was no difference in the prevalence of PCO among the different ethnic groups ({chi}2 = 1.95, P = 0.377). Among the 28 women with PCO, six (21%) were hirsute (Ferriman–Gallwey score >5) (Ferriman and Gallwey, 1961Go). None of the women with normal ovaries was hirsute. In the same group of women with unexplained infertility 34 had primary and 29 had secondary infertility. Of the 34 women with primary infertility, 20 had normal ovaries and 14 had PCO. Among the 29 women with secondary infertility, 15 had normal ovaries and 14 had PCO. There was no difference in BMI, day 8 FSH and duration of infertility between women with normal ovaries and women with PCO. Mean age was significantly higher in the group with normal ovaries and day 8 LH and testosterone significantly higher in the PCO group (Table IIGo). Within the infertile group as a whole, there was no difference in midluteal LH, FSH, testosterone and progesterone between women with normal ovaries and women with PCO [mean (SD): 3.9 (2.6) versus 5.2 (3.6) IU/l, 3.8 (1.3) versus 4.2 (1.6) IU/l, 2.2 (0.7) versus 2.5 (0.8) nmol/l, 44.2 (11.7) versus 45.4 (16.0) nmol/l respectively].

Among the 67 women of the control group, 48 (72%) women had normal ovaries and 19 (28%) had PCO. Among the women with normal ovaries, three (6%) had an irregular cycle and two (4%) were hirsute. Among the women with PCO, five (26%) had an irregular cycle and one (5%) was hirsute. There was no difference in BMI, day 8 FSH and testosterone between the group of women with normal ovaries and those with PCO. Mean age was significantly lower and day 8 LH significantly higher in the PCO group (Table IIGo). The women on the oral contraceptive did not have hormone-measurement. Among the women with normal ovaries, 35 (73%) were European in origin, 11 (23%) Afro-Caribbean and two (4%) of other ethnic origin (one Chinese and one Colombian). Among the women with PCO, 17 (89%) were European in origin and two (11%) Afro-Caribbean. There was no difference in the prevalence of PCO among the different ethnic groups (Fisher's exact test P = 0.44).

The prevalence of PCO was significantly higher in the group of anovulatory women ({chi}2 = 49.1, P < 0.0001), in the group with sperm dysfunction ({chi}2 = 8.7, P = 0.003) and in the group with tubal disease ({chi}2 = 5.8, P = 0.016) versus the control group. In the group of women with unexplained infertility, the prevalence of PCO appeared to be higher versus control women, but the difference was not statistically significant. There was no difference in the prevalence of PCO among ethnic groups when the unexplained infertility and the control groups were compared together (Mantel–Haenszel test, P = 0.111).

Finally, women with PCO of all infertility groups were compared with PCO women of the control group. Day 8 testosterone was significantly higher in the PCO anovulatory group, sperm dysfunction group, tubal disease group and unexplained infertility group versus the PCO control group. Mean age was significantly lower in the anovulatory and tubal disease PCO group versus the control group (Table IIGo).

It should be noted that our infertility clinic is a predominantly secondary referral centre from two local health authorities (Kensington and Chelsea; Westminster, Brent and Harrow) and is the main provider of infertility services in this area. Therefore, most of our patients are secondary referrals with only a few tertiary referrals, and our clinic data reflect this referral pattern.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Firstly, we observed a high prevalence of PCO in patients with tubal disease (50%), sperm dysfunction (53%) and, as expected, in women with anovulatory infertility (83%) (Adams et al., 1986Go; Hull, 1987Go). The prevalence of PCO was significantly higher in the tubal disease group and in the sperm dysfunction group compared with our control group (28%). In women with unexplained infertility the prevalence of PCO (44%) appeared to be higher than that of our control population, but the difference was not statistically significant. The prevalence of PCO in our control population is similar to that previously published (Polson et al., 1988Go; Clayton et al., 1992Go; Farquhar et al., 1994Go) 23, 22 and 21% respectively. All PCO women included in the tubal disease, sperm dysfunction and unexplained infertility groups were, by definition, ovulatory and had regular cycles. It is possible that, although these PCO women were ovulatory, there was some other unknown factor associated with PCO that contributed to, but was not the major cause of, subfertility. Among patients with tubal disease not all had completely blocked tubes; approximately half of them had some tubal patency. Similarly, among the couples with sperm dysfunction most of them had suboptimal sperm count and only six (8%) had zero sperm count in the separation test. These observations suggest that couples with a combination of sperm dysfunction or tubal disease and PCO are more likely to suffer from a conception delay and present to infertility clinics.

Secondly, we found that PCO women in all four diagnostic categories (anovulatory, sperm dysfunction, tubal disease and unexplained infertility) had higher testosterone concentrations in comparison with PCO women of the control group. It is not known whether hyperandogenaemia contributes to subfertility in normally ovulating PCO women, but this is a possibility perhaps by exerting an adverse effect in the endometrium.

Thirdly, in the comparison between normal ovaries versus PCO within infertility groups, younger age was a consistent finding in the PCO women of all diagnostic categories. This suggests that PCO are not only overrepresented in every diagnostic category (our first observation), but such patients may also present earlier to the infertility clinic. The PCO women with unexplained infertility have the characteristic biochemical finding of hyperandrogenaemia in comparison with those with normal ovaries. Midfollicular phase LH was significantly higher in the PCO group with anovulation versus the normal ovaries group, as expected, but also in the unexplained infertility group, although still within the normal range. Elevated values of midfollicular LH (>10 IU/l on day 8 of a regular menstrual cycle) have been associated with a significant impairment of fertility in women with regular spontaneous menstrual cycles (Regan et al., 1990Go). Neither higher testosterone nor higher midfollicular LH was found in women with PCO whose partners had sperm dysfunction versus those with normal ovaries, and higher testosterone was only documented in women with unexplained infertility or tubal disease and PCO.

The role of higher midfollicular phase LH and testosterone in normally ovulating PCO women remains to be clarified. In a previous study, ovulatory women with PCO were found to have a longer follicular phase, larger follicles and higher median follicular phase LH, FSH and testosterone when compared with ovulatory women with normal ovaries, suggesting that these differences may contribute to subfertility (Eden et al., 1989Go). It has been suggested that a high LH in the mid-follicular phase directly and adversely affects the timing of maturation of the oocyte, resulting in release of an `aged' oocyte (Homburg et al., 1988Go). Morphologically normal but functionally defective oocytes failing to fertilize in vitro have been associated with unexplained infertility in a small subgroup of patients treated with in-vitro fertilization (IVF) (Ezra et al., 1992Go). It could be postulated that the higher midfollicular phase LH, although still within the normal range, seen in our group of PCO patients with unexplained infertility contributes to subfertility by affecting the `quality' of the oocyte.

In summary, we have demonstrated a higher prevalence of PCO in patients with tubal disease (50%) and sperm dysfunction (53%) in comparison with that in a control population (28%) and a similar but not significant tendency among women with unexplained infertility (44%). We assume that, although these PCO women were ovulatory, there is some other unknown factor connected with PCO which contributes to subfertility, possibly related to hyperandogenaemia.


    Acknowledgments
 
This study was partially funded by a project grant from the U.K. Medical Research Council and through support from the Joint Research Standing Committee at St Mary's Hospital.


    Notes
 
3 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Adams, J., Polson, D.W. and Franks, S. (1986) Prevalence of polycystic ovaries in women with anovulation and idiopathic hirsutism. Br. Med. J., 293, 355–359.[ISI][Medline]

Clayton, R.N, Ogden, V., Hodgkinson, J. et al. (1992) How common are the polycystic ovaries in normal women and what is their significance for the fertility of the population? Clin. Endocrinol., 37, 127–134.[ISI][Medline]

Eden, J.A., Place, G.D., Carter, G.D. et al. (1989) Is the polycystic ovary a cause of infertility in the ovulatory woman? Clin. Endocrinol., 30, 77–82.[ISI][Medline]

Ezra, Y., Simon, A. and Laufer, N. (1992) Defective oocytes: a new subgroup of unexplained infertility. Fertil. Steril., 58, 24–27.[ISI][Medline]

Farquhar, C.M., Birdsall, M., Manning, P. et al. (1994) The prevalence of polycystic ovaries on ultrasound scanning in a population of randomly selected women. Aust. NZ J. Obstet. Gynaecol., 34, 67–72.[ISI][Medline]

Ferriman, D. and Gallwey, S.D. (1961) Clinical assessment of body hair growth in women. J. Clin. Endocrinol. Metab., 21, 1440.[ISI]

Franks, S., Gharani, N., Waterworth, D. et al. (1997) The genetic basis of polycystic ovary syndrome. Hum. Reprod., 12, 2641–2648.[Abstract]

Homburg, R., Armar, N.A., Eshel. A. et al. (1988) Influence of serum luteinising hormone concentrations on ovulation, conception, and early pregnancy loss in polycystic ovary syndrome. Br. Med. J., 297, 1024–1027.[ISI][Medline]

Hull, M.G.R. (1987) Epidemiology of infertilily and polycystic ovarian disease: endocrinological and dermographical studies. Gynecol. Endocrinol., 1, 235–245[Medline]

Hull, M.G.R. (1992) The causes of infertility and relative effectiveness of treatment. In Templeton, A.A. and Drife, J.O. (eds), Infertility. Springer-Verlag, London, pp. 33–58.

Polson, D.W., Adams, J., Wadsworth, J. et al. (1988) Polycystic ovaries—a common finding in normal women. Lancet, i, 870–872.

Regan, L., Owen, E.J. and Jacobs, H.S. (1990) Hypersecretion of luteinising hormone, infertilty and miscarriage. Lancet, 336, 1141–1144.[ISI][Medline]

Sagle, M., Bishop, K., Ridley, N. et al. (1988) Recurrent early miscarriage and polycystic ovaries. Br. Med. J., 297, 1027–1028.[ISI][Medline]

Templeton, A. (1992) The epidemiology of infertility. In Templeton, A.A. and Drife, J.O. (eds), Infertility. Springer-Verlag, London, pp. 23–32.

World Health Organization (1987) WHO Laboratory Manual for the Examination of Human Semen and Sperm–Cervical Mucus Interaction. Cambridge University Press, Cambridge.

Submitted on May 18, 1999; accepted on July 22, 1999.