The influence of body mass index, basal FSH and age on the response to gonadotrophin stimulation in non-polycystic ovarian syndrome patients

Transcription

1 Human Reproduction Vol.17, No.5 pp , 2002 The influence of body mass index, basal FSH and age on the response to gonadotrophin stimulation in non-polycystic ovarian syndrome patients Sheila Loh 1,2, Jim X.Wang 1,3 and Colin D.Matthews 1 1 Reproductive Medicine Unit, Department of Obstetrics and Gynaecology, Adelaide University, The Queen Elizabeth Hospital, 28 Woodville Road, Woodville, SA 5011, Australia 2 Present address: Reproductive Medicine Unit, KK Women s and Children s Hospital, 100 Bukit Timah Road, Singapore To whom correspondence should be addressed. jim.wang@adelaide.edu.au BACKGROUND: Adequate ovarian response to exogenous gonadotrophins is important for both ovulation induction (OI) and controlled ovarian stimulation (COS). The objective of this study was to analyse the effect of a number of clinical factors that influence ovarian response in non-polycystic ovarian syndrome (non-pcos) patients. METHODS: A total of 140 OI cycles (52 subjects), where each subject had a single abnormality (elevated FSH, abnormal body mass index (BMI) or 40 years of age), were compared with 54 cycles (15 subjects) where the patients displayed none of these abnormal features (the normal group). Similarly, 275 COS cycles (135 subjects), where each subject displayed a single abnormality, were compared with 79 cycles (40 subjects) in the normal group. RESULTS: For OI, subjects with a high basal FSH generally had an inadequate response with a poor chance of conception. Subjects with an abnormal BMI commonly required dosage adjustment so were more difficult to manage. Their potential for conception was normal. Older women seemed to respond normally with a normal expectation of conception. In the COS group, subjects with a moderately high basal FSH responded and conceived normally. Subjects with an abnormal BMI had an increased risk of an inadequate response leading to cancellation but if the response was adequate then the outlook was good. Older women required more gonadotrophin with a poor response and a low chance of conception. CONCLUSION: The results have better defined the anticipated responses of non-pcos patients to gonadotrophin stimulation in both OI and COS. Key words: age/bmi/controlled ovarian stimulation/fsh/ovulation induction Introduction The use of gonadotrophins for ovarian stimulation is common for infertility management. Indications include ovulation induction (OI) in anovular subjects or for patients receiving intrauterine insemination (IUI) and controlled ovarian stimulation (COS) for IVF. Near simulation of the natural cycle with a single dominant follicle is important to avoid multiple pregnancy in programmes of OI while a satisfactory ovarian response is critical to achieve an acceptable pregnancy rate without an excessive ovarian response for COS (Roest et al., 1996). Certain clinical factors including age, body weight and basal level of FSH are recognized as influencing the ovarian response (Cahill et al., 1994). Polycystic ovarian syndrome (PCOS) subjects, who have been well studied, may behave differently to non-pcos subjects due to their high sensitivity to gonadotrophin stimulation (Franks and Gilling-Smith, 1994). Variability in the response to gonadotrophin stimulation is common and can plague clinical management and it is important to improve pre-treatment counselling and the planning of management. Thus this study has focused on non-pcos patients and aimed to assess the influence of age, body mass index (BMI) and basal FSH on their response following gonadotrophin stimulation. Materials and methods The patients in this retrospective cohort study were managed between January 1995 and December All of them used human gonadotrophin purified from urinary sources (Pergonal ; Serono, Sydney, Australia and Humegon ; Organon, Sydney, Australia). These medications contained similar proportions of FSH and LH and were therefore designated generically as HMG or simply gonadotrophin. All patients had previously failed to respond adequately to clomiphene citrate up to a maximum dose of 150 mg for 5 days (Clomid ; Merrell Dow Ltd, Sydney, NSW, Australia). PCOS patients, as determined by the standard criteria (Norman et al., 1995) and male factor patients, based on World Health Organization standards (World Health Organization, 1992), were excluded in order to eliminate these confounding effects. Anovulation was proven by the low concentration of serum progesterone in well-timed and repetitively performed samples. BMI was measured at first attendance. European Society of Human Reproduction and Embryology 1207

2 S.Loh, J.X.Wang and C.D.Matthews The key elements of the protocol for OI were as follows: day 2 or 3 (either a natural or artificial menstrual cycle): baseline serum FSH; day 5: 75 IU HMG given daily i.m.; after 5 days treatment: serum estradiol (E 2 ) measurement and vaginal ovarian ultrasound scan to assess the response in order to decide whether to vary the dosage or to cancel the treatment cycle; day 10 onwards: if no more than three large follicles (defined as 16 mm in diameter) were present, then 5000 IU HCG was administered i.m. and natural intercourse h later advised or insemination carried out. If more than three large follicles were present, HCG was withheld and the couples were advised against intercourse or not inseminated. The clinical management, for example the starting dose of HMG or the adjustment of dose, was subject to some minor variation based on clinical judgement in individual subjects although overall the treatment programmes were performed within tight guidelines of the above protocol. The protocols for COS associated with IVF have been previously described (Sathanandan et al., 1989). Briefly, all patients included in the analysis utilized a gonadotrophin agonist (Leuprolide Acetate ; Abbott Pharmaceuticals, Sydney, Australia) either as a pituitary luteal phase desensitizing protocol or occasionally a flare protocol. Gonadotrophins were introduced on day 2 of the induced menstrual bleed in the luteal desensitizing protocol. The ovarian response was monitored in a similar fashion to OI but the criteria for the administration of HCG was determined by the presence of at least two follicles 17 mm in diameter. The cancellation criterion was less than two follicles being present on ultrasound scan on day 9 of the treatment cycle. HCG administration was only withheld if there was a serious clinical potential for ovarian hyperstimulation syndrome (OHSS) as indicated by a 50 nmol/l serum E 2 level at day of HCG. Five groups of women who had undertaken ovulation induction were identified based on the following features from the database: (i) those subjects with moderately elevated basal FSH levels (10 20 IU/l); (ii) those subjects with grossly elevated FSH levels ( 20 IU/l); (iii) those with a BMI 20 kg/m 2 (thin); (iv) those with a BMI 30 kg/m 2 (obese) and (v) those of 40 years of age (aged group). Each subject was selected with only one feature outside the range generally considered normal. For example, women aged 40 years or more would have a FSH level 10 IU/l and BMI between kg/m 2. Similar groups were extracted from the IVF population though there was no patient with FSH 20 IU/l owing to its rarity. This enabled the influence of the selected factors, age, BMI or basal FSH level, to be studied independently. Subjects who had undergone OI or COS during the same period but did not exhibit any of the specified features were used for comparison and designated the normal group. The main outcome variables analysed included: the starting dose of HMG, the total HMG dose administered and the duration (days), the response to HMG in terms of the maximum serum E 2 reached, the number of oocytes recovered and the outcome of the cycles including cancellation due to an inadequate or excessive response, and the occurrence of clinical pregnancy, as defined by the detection of fetal heart by an ultrasound scan on week 6 7 of gestation. Statistical analysis was performed by analysis of variance (ANOVA) or χ 2 -test to test the differences between the groups. Intra-group comparisons were made using Tukey s test. Where the data were not normally distributed, Kruskal Wallis ANOVA on ranks was carried out. All statistical analysis was done with SPSS (SPSS Inc., Chicago, MI, USA). Results Ovulation induction (Table I) There were 168 cycles from 67 subjects, average 2.9 cycles per subject. The small number in the high FSH group (eight cycles, four subjects) has limited the statistical power of the analysis concerning this group. The mean age of the women in the aged group was 42.6 years, significantly greater (P 0.05) than the normal group. The thin women had a mean BMI of 18.0 kg/m 2 while the obese women had a mean BMI of 36.4 kg/m 2, both significantly different (P 0.05) from the other groups. Finally the two groups of women with moderately high and high FSH had mean values of FSH 12.6 IU/l and 66.6 IU/l respectively, significantly higher (P 0.01) than the normal group. Starting HMG dose, duration of treatment and total dosage of HMG The starting dose of gonadotrophin was generally 75 IU but those women with high FSH or 40 years were commenced on 112 IU daily. The duration of HMG stimulation was longest in the high FSH group (14 days), significantly longer than the normal group (9.1 days, P 0.01). Both thin (11.5 days, P 0.02) and obese women (11.7 days, NS) also required a longer treatment than the normal group. An average total of 720 IU HMG was required for OI in the normal group. The women with a moderately high FSH required similar amounts to the controls, as did the older women. However, the obese women (1197 IU) and women with high FSH (1650 IU) required statistically more compared with the normal group (P 0.05). Response to HMG: serum E 2, follicle numbers and cycle cancellation The women with high FSH had only 0.5 large follicle on average, which was significantly fewer than the normal group. This group also had a much lower E 2 concentration (0.5 nmol/l, P 0.001) compared with that (1.6 nmol/l) in the normal group. Lower E 2 concentrations were also observed in three other groups (ranging from nmol/l). Cancellation due to excess stimulation, i.e. more than three large follicles, occurred in 19% of the normal group. Thin women also displayed a similar propensity (11%) while hyperstimulation was rare in other groups. Conversely obese women had a significantly higher risk of hypostimulation (14%, P 0.05), compared with the normal group. Clinical pregnancy Pregnancy rate was 11% in the normal group and 30% in thin subjects but this difference was not significant due to the small sample size. Obese and aged groups also had similar pregnancy rates to the normal group while women with elevated FSH had very poor pregnancy rates (0% for high and 5% for moderately high FSH). Controlled ovarian stimulation (Table II) There were 354 cycles (79 in the normal group) derived from 175 subjects. The BMI of the thin subjects was 18.7 kg/m 2 and the obese group 33.6 kg/m 2, significantly different from the normal group. The moderately high FSH group had a mean FSH of 13.9 IU/l, which was statistically higher (P 0.01) than the normal group. The mean age of the older women was 42.1 years, again significantly greater than the normal group. Starting HMG dose, duration of stimulation and total dosage All patients commenced on a starting dose of 225 IU except for the older group (300 IU) and the moderately high FSH 1208

3 Gonadotrophin stimulation in non-pcos patients Table I. Demographic information and outcomes of gonadotrophin stimulation for ovulation induction in six groups of infertile women (mean and SD within brackets) Group n Age BMI FSH Start Total Days E 2 Large Stimulation Pregnant (%) (years) (kg/m 2 ) (IU/l) dose (IU) HMG (IU) HMG (nmol/l) follicles (no.) Hyper Hypo Moderate * %* 0% 5%* FSH (4.4) (4.4) (1.2) (287) (2.1) (0.6) (0.9) High * * 14.0* 0.5* 0.5* 0%* 13% 0%* FSH (3.6) (4.3) (33.2) (1060) (2.8) (0.6) (0.7) Thin * * % 4% 30% (3.4) (1.2) (1.6) (490) (4.6) (1.0) (0.8) Obese * * 11.7* %* 14%* 17% (3.5) (4.6) (1.8) (819) (7.0) (0.4) (0.6) Aged * %* 2% 12% ( 40 years) (1.6) (3.1) (2.9) (618) (2.4) (0.5) (0.9) Normal % 0% 11% group (3.0) (1.9) (1.4) (233) (1.6) (1.6) (1.0) P or NS NS *indicate statistically different (P 0.05) results from the normal group. n number of cycles; E 2 estradiol; NS not significant; BMI body mass index. Table II. Demographic information and outcomes of high dose gonadotrophin stimulation for IVF in five groups of infertile women (mean and SD within brackets) Group n Age BMI FSH (IU/l) Start dose Total Days E 2 Oocytes Embryos Stimulation Pregnant OHSS (years) (kg/m 2 ) (IU) per cycle HMG HMG (nmol/l) (no.) (no.) % % (IU) Hyper Hypo Moderate * * % 16% 23% 15% FSH (4.0) (3.9) (3.6) (1238) (1.6) (5.9) (7.0) (5.4) Thin * % 16% 9%* 15% (4.3) (1.0) (2.5) (1820) (2.0) (4.2) (6.5) (3.9) Obese * * * 1% 9% 23% 5% (3.6) (3.6) (3.0) (1072) (1.9) (3.5) (5.9) (3.4) Aged * * * 8.5* 5.0* 0% 15% 11%* 4% ( 40 years) (1.2) (5.1) (3.0) (652) (2.0) (3.4) (7.5) (4.1) Normal % 10% 26% 4% group (4.1) (2.6) (2.5) (973) (1.6) (3.5) (6.5) (4.9) P or NS NS 0.01 NS *indicate statistically different (P 0.05) results from the normal group. n number of cycles; OHSS ovarian hyperstimulation syndrome; NS not significant; BMI body mass index. group (262 IU). Only a small percentage of patients needed to have their dose decreased (5 7%) or increased (4%) to achieve a satisfactory response. The mean length of stimulation in the normal group was 11.5 days, obese women required significantly fewer days (9.9 days, P 0.01) while the remaining groups had between 10.2 and 10.8 days of stimulation, not different from the normal group. The total dose of gonadotrophin in the normal group was 2238 IU and was significantly less than in the moderately high FSH group (2931 IU, P 0.001) and the aged group (3128 IU, P 0.001) but was the same as the thin (2628 IU) or the obese groups (2187 IU). Thin women appeared to use more gonadotrophin than obese women (2628 versus 2187 IU) but this was not statistically significant. Response to HMG: serum E 2, number of oocytes and embryos and cancellation rate The normal group had a maximum serum E 2 of 6.7 nmol/l. The only group that failed to respond well was the older group, which had a mean E 2 of just 4.6 nmol/l, lower than the normal group (P 0.001). All other groups reached similar levels as the normal group ( nmol/l). The mean number of oocytes recovered in the normal group was 10.9, significantly more than the older group (8.5, P 0.05) but similar to the other groups (range ). The number of embryos formed from the IVF procedures was also significantly lower in older (5.0) and obese women (5.1) compared with the normal group (6.6, P 0.05), while thin women also appeared to have fewer embryos (5.2), but not statistically significantly so. Interestingly, women with a moderately high FSH had similar numbers of embryos (6.7) formed as the normal group. Cancellation due to possible hyperstimulation was very low ( 1%) in all groups. By contrast cancellation due to poor response was common, 15 16% in women with moderately high FSH, the thin and the older groups, significantly higher than the 10% cancellation rate in the normal group (P 0.05). Only 9% of the cycles among the obese subjects were cancelled. 1209

4 S.Loh, J.X.Wang and C.D.Matthews Clinical pregnancy The pregnancy rate per completed cycle was 26% in the normal group, similar to the moderately high FSH group (23%, 13 pregnancy cycles) and the obese patients (23%, 15 pregnancy cycles). However the pregnancy rate was significantly lower in the older (11%, six pregnancy cycles, P 0.01) and the thin groups (9%, five pregnancy cycles, P 0.01) compared with the normal group. Discussion Ovulation induction The usual objective of OI is to stimulate the ovary to develop one to two large follicles. However, it is clinically well appreciated that certain factors can greatly influence the response to HMG. Studies in PCOS patients have shown that basal FSH level, BMI and age are important influencing factors. The basal FSH level reasonably accurately reflects ovarian reserve. Subjects with high FSH levels commonly had a reduced fecundity with ovulation induction (Buyalos et al., 1997). Older women tend to have higher FSH levels (Fitzgerald et al., 1994). Not well documented are the effects of these factors in non-pcos women. In the present study of non-pcos women, those with a high basal FSH level also had a poor response to stimulation and no pregnancy even after an extended stimulation with high dose of HMG (mean 1650 IU). By contrast those women with only moderately raised FSH levels generally responded satisfactorily to an average dosage regime though their pregnancy rate was low (5%). This may suggest that the remaining complement of oocytes have poor pregnancy potential even though the subjects were young. Studies of PCOS women have found little influence of BMI on the response to ovulation induction (McClure et al., 1992). Dale et al. concluded that increased BMI in PCOS women was the only factor that they investigated that was associated with a decreased response to clomiphene administration (Dale et al., 1998). In the present study about one quarter of the obese women and half of the thin women required an increase in HMG dosage to effect a satisfactory response. Clearly the extremes of BMI make the choice of the optimal dosage regime more difficult than in patients with a normal BMI. Normal BMI increases the success rate in OI while weight disturbances, especially overweight, have a deleterious effect (Chong et al., 1986; Kably-Ambe et al., 1999). Both thin and obese women utilized more HMG than others and required a longer duration of treatment but the eventual response was reasonable and the expectation of pregnancy was also satisfactory. The older women in this study with normal FSH levels and BMI responded well to gonadotrophin stimulation and the pregnancy rate was also as good as the normal group. Age per se (within the range in this study) did not seem to have an obvious adverse effect. This is in some contrast to COS where age was an important factor with excessive gonadotrophin dosage only resulting in a poor ovarian response and a reduced pregnancy potential (see below). The cancellation rate was 19% in the normal group, essentially to avoid multiple pregnancy rather than to 1210 avoid OHSS, which did not occur in the study population and has been rare in our OI programme. Cancellation for an excessive response was less likely in thin or obese women (11 and 3% respectively). Controlled ovarian stimulation The objective of gonadotrophin stimulation for COS has been to achieve an optimal ovarian response which is crucial to the efficiency of IVF procedures. The clinical features commonly associated with a poor or hyper response to stimulation, including age, BMI and moderately high FSH level, were studied in non-pcos patients. Non-PCOS women constituted more than half the patients receiving COS in our programme. The older women with mean age of 42 years and the women with moderately high FSH (up to twice normal in the present study) were not considered post-menopausal in the clinical sense so both groups were not in extreme clinical situations. High FSH lead to a poor response to gonadotrophins in COS (Ebrahim et al., 1993) as reflected by the high cancellation rate due to hypostimulation. This rate was also influenced by a government policy of enforcing a stop or go decision at day 9 for financial reimbursement purposes, nevertheless the clinical decision-making was similar for all the groups investigated. However, due to the large amount of HMG used, there was the potential for a hyper response as the high risk of OHSS suggested. Excessive stimulation in COS can lead to life-threatening clinical situations and therefore monitoring of the ovarian response is essential. Increasing age is an important adverse factor and our policy of limiting access to IVF for women 43 years of age meant that there were fewer cycles in this group and limited statistical power to draw conclusions. Nonetheless, this group started with a higher gonadotrophin dose and used a greater total amount of HMG. Even so, this group still had 15% of the cycles cancelled due to poor response and among those responding, the response was poor in terms of maximum E 2 and number of oocytes recovered. Since the mean FSH levels were higher in this group (6.7 nmol/l) than other groups except the moderately high FSH group, it could be that some form of mild incipient ovarian failure is present with no obvious clinical manifestation (Cahill et al., 1994). Roest et al. also found an increased rate of poor ovarian response in women 40 years of age but a proportion with a good response and with a satisfactory outlook for conception (Roest et al., 1996). Overall there was no particularly consistent effect of BMI on the response to COS, a finding similar to that reported by Lewis et al. (Lewis et al., 1990) and Lashen et al. (Lashen et al., 1999). This result was contrary to that reported by Crosignani et al. (Crosignani et al., 1994). Somewhat surprisingly, thin patients required more gonadotrophin (2628 IU) than obese women (2187 IU). The starting dose of gonadotrophin was similar for them. It should be noted that the response of thin patients was a little in excess of the obese women in terms of E 2 levels (6.2 versus 5.6 nmol/l) and number of oocytes recovered (10.6 versus 9.1). The explanation for the need for more total gonadotrophin in thin women is not clear, since the best indicator of the need for gonadotrophin, namely the basal FSH level, was similar in both obese and thin women.

5 Gonadotrophin stimulation in non-pcos patients Weight loss, even limited, has been shown to be effective in causing the recovery of ovulation or increasing response to gonadotrophin stimulation (Clark et al., 1995). Comparing the responses of the various groups with both OI and COS, some differences were apparent. In ovulation induction, patients with moderately high basal FSH responded well to gonadotrophin stimulation but with a poor chance of conception. Similar subjects under COS had a satisfactory (23%) pregnancy rate; perhaps this group would benefit from a slightly more aggressive gonadotrophin regime. For the thin group under COS, the amount of gonadotrophin used was not excessive and the response was also normal with adequate serum E 2 levels, numbers of oocytes and embryos created although 15% of cycles were cancelled because of an excessive response. This was in contrast to thin women undertaking ovulation induction where no cycle was cancelled for a poor ovarian response. The obese subjects undertaking OI tended to respond poorly while similar patients under COS responded, by and large, normally and had a similar expectation of pregnancy as non-obese women. Generally BMI was not a significant factor in the response to either OI or COS but clinical management during the treatment cycle was clearly more demanding in both thin and obese subjects. The influence of age was also interesting. Older normogonadotrophic women of normal BMI undertaking OI appeared to present no particular special issue and responded normally. However similar women undertaking COS required excessive gonadotrophin (3128 IU) to reach an acceptable, albeit lower, response level. Since there was no overlapping of these two groups of patients this may indicate that the older ovary has less reserve to respond to the call for multiple follicles even though basal FSH levels are normal. Although miscarriage rate and multiple pregnancy rate are important parameters in infertility treatment, due to the small sample size, they were not assessed here. This limitation in statistical power also meant that the interaction between these important clinical factors could not be properly studied. Conclusions In non-pcos women undergoing OI, high basal FSH was associated with poor response to ovarian stimulation. Thin or obese patients were more difficult to manage clinically with varying responses to gonadotrophins, resulting in cancellation for either a poor or hyper response. Those who responded adequately, however, had an acceptable pregnancy rate. Older women who were normogonadotrophic with normal BMI responded normally and had a normal expectation of conception. Non-PCOS patients undergoing COS with moderately high basal FSH responded normally with a normal pregnancy rate. Thin or obese patients generally responded normally and had a normal chance of conception. Older women who were normogonadotrophic and of normal BMI required a larger amount of gonadotrophin and achieved a less optimal ovarian response with a low expectation of conception. Acknowledgements Our appreciation goes to the administrative and nursing staff of The Reproductive Medicine Unit of Adelaide University for assistance for facilitating data collection from clinical records of Repromed Pty Ltd. Dr Sheila Loh was supported by a clinical training grant from The Singapore Government. References Buyalos, R.P., Daneshmand, S. and Brzechffa, P.R. (1997) Basal estradiol and follicle-stimulating hormone predict fecundity in women of advanced reproductive age undergoing ovulation induction therapy. Fertil. Steril., 68, Cahill, D.J., Prosser, C.J., Wardle, P.G., Ford, W.C. and Hull, M.G. (1994) Relative influence of serum follicle stimulating hormone, age and other factors on ovarian response to gonadotrophin stimulation. Br. J. Obstet. Gynaecol., 101, Chong, A.P., Rafael, R.W. and Forte, C.C. (1986) Influence of weight in the induction of ovulation with human menopausal gonadotropin and human chorionic gonadotropin. Fertil. Steril., 46, Clark, A.M., Ledger, W., Galletly, C., Tomlinson, L., Blaney, F., Wang, X. and Norman, R.J. (1995) Weight loss results in significant improvement in pregnancy and ovulation rates in anovulatory obese women. Hum. Reprod., 10, Crosignani, P.G., Ragni, G., Parazzini, F., Wyssling, H., Lombroso, G. and Perotti, L. (1994) Anthropometric indicators and response to gonadotrophin for ovulation induction. Hum. Reprod., 9, Dale, P.O., Tanbo, T., Haug, E. and Abyholm, T. (1998) The impact of insulin resistance on the outcome of ovulation induction with low-dose follicle stimulating hormone in women with polycystic ovary syndrome. Hum. Reprod., 13, Ebrahim, A., Rienhardt, G., Morris, S., Kruger, T.F., Lombard, C.J. and Van der Merwe, J.P. (1993) Follicle stimulating hormone levels on cycle day 3 predict ovulation stimulation response. J. Assist. Reprod. Genet., 10, Fitzgerald, C.T., Seif, M.W., Killick, S.R. and Elstein, M. (1994) Age related changes in the female reproductive cycle [published erratum appears in Br. J. Obstet. Gynaecol. (1994) 101 (4): 360]. Br. J. Obstet. Gynaecol., 101, Franks, S. and Gilling-Smith, C. (1994) Advances in induction of ovulation. Curr. Opin. Obstet. Gynecol., 6, Kably-Ambe, A., Barron-Vallejo, J. and Limon-Luque, L.M. (1999) Effect of body mass index and follicular synchrony in the ovarian response to ovulation induction with menotropins. Int. J. Fertil. Women s Med., 44, Lashen, H., Ledger, W., Bernal, A.L. and Barlow, D. (1999) Extremes of body mass do not adversely affect the outcome of superovulation and invitro fertilization. Hum. Reprod., 14, Lewis, C.G., Warnes, G.M., Wang, X.J. and Matthews, C.D. (1990) Failure of body mass index or body weight to influence markedly the response to ovarian hyperstimulation in normal cycling women. Fertil. Steril., 53, McClure, N., McQuinn, B., McDonald, J., Kovacs, G.T., Healy, D.L. and Burger, H.G. (1992) Body weight, body mass index, and age: predictors of menotropin dose and cycle outcome in polycystic ovarian syndrome? Fertil. Steril., 58, Norman, R.J., Masters, S.C., Hague, W., Beng, C., Pannall, P. and Wang, J.X. (1995) Metabolic approaches to the subclassification of polycystic ovary syndrome. Fertil. Steril., 63, Roest, J., van Heusden, A.M., Mous, H., Zeilmaker, G.H. and Verhoeff, A. (1996) The ovarian response as a predictor for successful in vitro fertilization treatment after the age of 40 years. Fertil. Steril., 66, Sathanandan, M., Warnes, G.M., Kirby, C.A., Petrucco, O.M. and Matthews, C.D. (1989) Adjuvant leuprolide in normal, abnormal, and poor responders to controlled ovarian hyperstimulation for in vitro fertilization/gamete intrafallopian transfer. Fertil. Steril., 51, World Health Organization (1992) WHO laboratory manual for the examination of human sperm and sperm cervical mucus interaction, 3rd edn, Cambridge University Press, Cambridge, p. 44. Submitted on June 8, 2001; resubmitted on October 12, 2001; accepted on December 3,