Obesity is a risk factor for early pregnancy loss after IVF or ICSI

Transcription

1 Acta Obstet Gynecol Scand 2000; 79: Copyright C Acta Obstet Gynecol Scand 2000 Printed in Denmark All rights reserved Acta Obstetricia et Gynecologica Scandinavica ISSN ORIGINAL ARTICLE Obesity is a risk factor for early pregnancy loss after IVF or ICSI PÉTER FEDORCSÁK 1,2,RITSA STORENG 1,PER OLAV DALE 1,TOM TANBO 1 AND THOMAS ÅBYHOLM 1 From the 1 Department of Obstetrics and Gynecology, National Hospital, University of Oslo, Oslo, Norway and the 2 First Department of Obstetrics and Gynecology, Semmelweis University Medical School, Budapest, Hungary Acta Obstet Gynecol Scand 2000; 79: C Acta Obstet Gynecol Scand 2000 Background. Experience with polycystic ovary syndrome shows that insulin resistance is related to early pregnancy loss. This association was examined by comparing pregnancy outcome in obese and lean women. Methods. A cohort of 383 patients conceiving after in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI) was studied. Ovarian stimulation was achieved by GnRHa and FSH or hmg (nω362), by FSH or hmg alone (nω16), or by clomiphene citrate and FSH or hmg (nω5). Luteal phase was supported with progesterone. Pregnancies were defined by ±10 IU/l plasma ß-hCG on day 14. Ultrasound scan on week 6 and week 12 confirmed fetal viability. Results. Lean group (body mass index [BMI] 25 kg/m 2 ; nω304) and obese group (BMI Ø25 kg/m 2 ; nω79) were established. Obese patients had fewer oocytes collected (median: 8 vs 10 pω0.03), they had higher abortion rate during the first 6 weeks (22% vs 12%; pω0.03) and lower live-birth rate (63% vs 75%; pω0.04). The relative risk of abortion before week 6 was 1.77 (95% CI: 1.05 to 2.97). Multivariate logistic regression analysis revealed that obesity and low oocyte count were independently associated with spontaneous abortion. In the obese group, low oocyte number was associated with a more profound increase in the risk of abortion than among lean patients. The effect of age, history of past pregnancies, or infertility diagnosis on the probability of miscarriage were not significant. Conclusions. Obesity is an independent risk factor for early pregnancy loss. This risk is, in part, related to the lower number of collected oocytes in obese women. Key words: intracytoplasmic sperm injection; in vitro fertilization; obesity; spontaneous abortion Submitted 12 April, 1999 Accepted 5 July, 1999 Obesity, especially abdominal obesity, impairs fecundity and reduces conception rate during infertility treatment (1 4). Furthermore, studies on obese, infertile polycystic ovary syndrome (PCOS) Abbreviations: BMI: body mass index; CC: clomiphene citrate; CI: confidence interval; FSH: follicle stimulating hormone; GnRHa: gonadotropin releasing hormone analog; hcg: human chorionic gonadotropin; hmg: human menopausal gonadotropin, ICSI: intracytoplasmic sperm injection; IM: intra muscular; IVF: in vitro fertilization LH: luteinizing hormone; n.s.: not significant; OR: odds ratio; p: probability; PCO: polycystic ovaries; PCOS: polycystic ovary syndrome. patients suggest that obesity is related to miscarriage, as well (5, 6). PCOS, which is characterized by hyperandrogenism, chronic anovulation, hypersecretion of luteinizing hormone (LH) and insulin resistance, is itself associated with early pregnancy loss (7). Long-term pituitary desensitization with gonadotropin releasing hormone analog (GnRHa) may decrease miscarriage rate in PCOS patients (8), therefore, LH hypersecretion has been postulated as a cause of miscarriage in PCOS (9). PCOS patients, however, can be divided into two subgroups: lean, non-insulin resistant patients with high LH levels and obese, insulin resistant patients

2 44 P. Fedorcsák et al. with normal LH secretion (10). It has recently been shown that spontaneous abortion is more frequent among insulin resistant than non-insulin resistant patients with PCOS undergoing ovarian stimulation with follicle stimulating hormone (FSH) (5). Thus, insulin resistance may be associated with early pregnancy loss independently. This led us to investigate the pregnancy outcome in obesity, a condition characterized by insulin resistance (11). Patients and methods Records of all 395 patients who conceived once or more with IVF or intracytoplasmic sperm injection (ICSI) between August 1996 and January 1998 were analyzed. Twelve patients (3%) were excluded because of incomplete data. Only the first pregnancy for each couple was included in this analysis. Thus, 383 pregnancies of 383 couples were studied. Patients received stimulation with FSH or human menopausal gonadotropin (hmg) after longterm GnRHa down-regulation (nω362; 94.5%), with FSH or hmg only (nω16; 4.2%), or with combination of clomiphene citrate (CC) and FSH or hmg (nω5; 1.3%). Before gonadotropin treatment, a baseline blood sample was taken, and was assayed for LH, FSH, estradiol, and prolactin. During GnRHa/FSH/hMG treatment, 600 mg/day nafarelin (Synarela, Searle, England) or buserelin (Suprefact, Hoechst, Germany) was given from the mid-luteal phase of the preceding cycle. Desensitization was considered complete after 2 4 weeks if no follicles ±10 mm were seen on ultrasound scan and serum estradiol was 0.2 nmol/l. Then, follicular development was stimulated with FSH (Gonal F, Serono, Switzerland) or hmg (Pergonal, Serono), with starting dose IU/day. In cycles where only FSH or hmg was given, gonadotropin treatment started on cycle day 3, with 300 IU/day for two days followed by 150 IU daily. CC/ FSH/hMG treatment comprised administration of 100 mg/day clomiphene citrate (Pergotime, Serono) from cycle day 3 to day 7 and FSH or hmg treatment (225 IU followed by 150 IU daily) from day 7. With each regimen, IU hcg (Profasi, Serono) was injected when three or more ±17 mm follicles were seen on ultrasound scan. Follicles were aspirated transvaginally under ultrasound guidance within h. Oocytes were fertilized by IVF or ICSI (12, 13). One to three embryos were transferred on day 2, 3, or 4 after oocyte collection. Based on embryonic morphology, embryo quality was graded from Grade 1 (equal or unequal size blastomeres, no fragmentation) to Grade 4 (±50% fragmentation) (14). The quality of the best transferred embryo is reported, since this shows the strongest correlation with pregnancy outcome (our unpublished observation). In each case, luteal phase was supported with progesterone (25 mg/day, IM injection). Pregnancies were defined by ±10 IU/I plasma ß-hCG level on day 14 after follicle puncture. Absence of gestational sac on ultrasound scan at sixth week confirmed early abortion. Further ultrasound scan was performed at 12 weeks gestation to assess fetal viability. Results are presented as median and interquartile range. Data were compared by c 2 test or by Student s t test. Data with skewed distribution were log-transformed for analysis. Since 97% of the original cohort was assessed, relative risk and its 95% confidence intervals (CI) were calculated. Univariate and multivariate logistic regression was used to analyze the relation between probability of early pregnancy loss and possible explanatory variables using the Statistical Package for Social Sciences programme (SPSS; Version 8). The following variables were considered in the model: presence of obesity, quality of the best transferred embryo, history of spontaneous abortion, presence of polycystic ovaries, presence of endometriosis, and the ovarian stimulation regimen as categorical variables, while number of collected oocytes and age as continuous variables. An interaction term between obesity and number of collected oocytes was also included in the model, because preliminary analysis suggested that oocyte count modifies the effect of obesity. To minimize the number of explanatory variables and to improve the fit of the logistic model, multivariate analysis was performed by backward stepwise regression. The model s goodness-of-fit was assessed by the Hosmer and Lemeshow test, which did not reject the model (pω0.41). Indeed, the multivariate model correctly predicted the presence or absence of spontaneous abortion in 85.6% of the cases. Results of logistic regression are given as odds ratios (OR) and 95% CI. Regression analyses were based on 368 cases because of missing data. p 0.05 was considered statistically significant. Results The median age of the patients was 33 years (range 23 to 42) and the median body mass index (BMI) was 22.2 kg/m 2 (range 15.9 to 38.6). On the basis of body mass index two subgroups were established: lean group comprised 304 (79.4%) patients with BMI 25 kg/m 2, obese group comprised 79 (20.6%) patients with BMI Ø25 kg/m 2. Patient characteristics are shown in Table I. Groups were similar in age, in duration of infer-

3 Early pregnancy loss in obesity 45 Table I. Characteristics of lean and obese patients conceiving after IVF or ICSI Lean group Obese group (nω304) (nω79) p value Body mass index (kg/m 2 ) 21.5 ( ) 27.5 ( ) Age (years) 33 (30 35) 32 (29 36) N.S. Duration of infertility (years, mean) 6 (4 8) 6 (4 7.8) N.S. Baseline hormone levels* LH (IU/l) 1.7 ( ) 1.65 ( ) N.S. FSH (IU/I) 3.3 ( ) 3.5 ( ) N.S. Estradiol (nmol/l) 0.11 ( ) 0.13 ( ) N.S. Prolactin (miu/l) 204 ( ) 232 ( ) N.S. History of past pregnancies No. with primary infertility 177 (58.2%) 36 (45.6%) 0.04 No. with spontaneous abortion 54 (17.8%) 19 (24.0%) N.S. No. with induced abortion 28 (9.2%) 6 (7.7%) N.S. No. with ectopic pregnancy 39 (12.8%) 10 (12.7%) N.S. No. with previous live birth 55 (18.1%) 18 (22.8%) N.S. Causes of infertility No. with tubal disease 136 (44.7% 30 (37.9%) N.S. No. with male factor 78 (25.7%) 19 (24.1%) N.S. No. with endometriosis 62 (20.4%) 9 (11.4%) N.S. No. with polycystic ovaries 20 (6.6%) 16 (20.3%) No. with idiopathic infertility 44 (14.5%) 13 (16.5%) N.S. No. with other causes 3 (1.0%) 0 N.S. Results are median (interquartile range) or n (%) values as appropriate. N.S., not significant. * Before gonadotropin administration in cycles with GnRHa/FSH/hMG treatment (nω362). Individual cases with uterine myomata, habitual abortion and uterus bicornis. Table II. Ovarian stimulation, assisted conception and pregnancy outcome in lean and obese patients conceiving after IVF or ICSI Lean group Obese group (nω304) (nω79) p value Ovarian stimulation GnRHa/FSH/hMG 288 (94.7%) 74 (93.7%) N.S. FSH/hMG 13 (4.3%) 3 (3.8%) N.S. CC/FSH/hMG 3 (1.0%) 2 (2.5%) N.S. Amount of FSH/hMG (IU)* 1875 ( ) 2025 ( ) N.S. No. of oocytes recovered* 10 (6 14) 8 (5 13) 0.03 Fertilization and embryo transfer IVF 241 (79.3%) 65 (82.3%) N.S. ICSI 63 (20.7%) 14 (17.7%) N.S. No. of embryos transferred 2 (2 2) 2 (2 2) N.S. Quality of the best transferred embryo Grade 1 52 (17.9%) 10 (12.7%) Grade (71.0%) 57 (72.2%) Grade (11.1%) 12 (15.1%) Pregnancy outcome Live birth 228 (75.0%) 50 (63.3%) 0.04 Abortion before week 6 37 (12.2%) 17 (21.5%) 0.03 Abortion between week (8.9%) 10 (12 6%) N.S. Abortion after week 12 4 (1.3%) 1 (1.3%) N.S. Other 8 (2.6%) 1 (1.3%) N.S. Results are median (interquartile range) or n (%) values as appropriate. N.S., not significant. * In cycles with GnRHa/FSH/hMG treatment (nω362). Ectopic pregnancies and intrauterine deaths. N.S.

4 46 P. Fedorcsák et al. Table III. Risk of spontaneous abortion before week 6 of gestation as modelled by univariate and multivariate logistic regression analysis Univariate analysis Multivariate analysis OR 95% CI p OR 95% CI p Presence of obesity to to No. of oocytes recovered to to Interaction term between obesity and no. of recovered oocytes to 1.11 N.S to Age to 1.10 N.S. History of spontaneous abortion to 3.24 N.S. Presence of PCO to 2.47 N.S. Presence of endometriosis to 1.72 N.S. Ovarian stimulation N.S. GnRHa/FSH/hMG 1.0 reference FSH/hMG to 5.96 CC/FSH/hMG to 4* Quality of the best transferred embryo N.S. Grade reference Grade to 4.19 Grade to 6.29 tility, and in baseline levels of LH, FSH, estradiol, and prolactin. Lean patients were more likely to present with primary infertility (58% vs 46%, pω 0.04), but the history of spontaneous and induced abortions, ectopic pregnancy, and live birth were similar. More obese than lean patients had polycystic ovaries (20% vs 7%, p 0.001). Endometriosis tended to be less prevalent in the obese group. Other causes of infertility were similar in the two groups. The proportion of patients receiving different stimulation regimens and fertilization method was similar between groups (Table II). Obese patients tended to require more gonadotropin in GnRHa/ FSH/hMG cycles, still they had significantly fewer oocytes collected. The number and the quality of transferred embryos were similar. Groups differed in pregnancy outcome. Obese patients had significantly fewer live births because of increase in abortion frequency during the first 6 weeks: the relative risk of abortion before week 6 was 1.77 (95% CI: 1.05 to 2.97). Factors, which are related to both obesity and spontaneous abortion may confound or modify the relationship of obesity and abortion. To assess the independent role of obesity, the probability of spontaneous abortion before 6 weeks gestation was modelled by multivariate logistic regression (Table III). Multivariate analysis retained three significant explanatory variables in the model: the presence of obesity, the number of collected oocytes, and the interaction term between obesity and oocyte count (Table III). To show that these findings were independent of the definition of obesity as BMI Ø25 kg/m 2, other multivariate models where obesity was defined as BMI Ø26, 27, or 28 kg/m 2 were also tested. These models gave similar results as above (not shown). Thus, obesity and low oocyte count were independently associated with spontaneous abortion, but the effect of oocyte count was different among obese and lean women. To illustrate the interaction between obesity and the number of oocytes, the predicted probability of early pregnancy loss was Fig.1. The relationship of the number of collected oocytes and the predicted probability of spontaneous abortion before 6 weeks gestation in obese and lean women (circles; left Y axis; the size of the circles is proportional to the number of cases). The distribution of the total number of patients (open bars; right Y axis), and the total incidence of early pregnancy loss (shaded bars; right Y axis).

5 Early pregnancy loss in obesity 47 plotted against the number of collected oocytes (Fig. 1). In the obese group, decrease in oocyte number was associated with a more profound increase in probability of abortion than among lean patients. Note that among patients with ±15 collected oocytes the number of cases and the observed incidence of spontaneous abortion is too low to make inferences from the predicted probability (Fig. 1). The number of collected oocytes and the quality of transferred embryos were positively correlated: the proportion of cases with Ø10 oocytes decreased from 62% to 48% and to 37% as the embryo quality declined from Grade 1 to 2 and to 3 4(c 2 trendω6.35; d.f.ω1; pω0.01). Discussion To demonstrate the association of obesity and miscarriage, we examined the outcome of IVF and ICSI pregnancies in obese and lean women. Obese patients had lower live-birth rate and higher rate of abortion before the sixth week of gestation. We also showed that the relationship of obesity and miscarriage is independent from age, history of spontaneous abortion, endometriosis, and ovarian stimulation regimen, which are all related to spontaneous abortion and its recurrence (15 17). Interestingly, the association of obesity and miscarriage was also independent from the presence of polycystic ovaries, although in other studies PCOS was associated with early pregnancy loss (7). Furthermore, all our patients received luteal phase support with progesterone, making it unlikely that corpus luteum insufficiency could explain the risk associated with obesity (18). Our results, however, have some limitations: further studies are needed to clarify the role of fat distribution, since abdominal obesity may have more pronounced impact on fertility than peripheral fat distribution (1, 2, 19). The confounding effect of smoking, alcohol or caffeine consumption (20), must be examined by future studies, as well. Nevertheless, obesity may be an independent risk factor for early pregnancy loss in patients undergoing IVF or ICSI. Previous studies on PCOS patients and on the general population also support this conclusion (5, 6). Therefore, overweight patients enrolled in assisted reproduction program should be encouraged to lose weight, since it may increase their chances for conception and for live birth (21, 22). In this study low oocyte count was also associated with early pregnancy loss, and a positive correlation was found between oocyte count and embryo quality at transfer. Thus, few oocytes may result in fewer embryos for selection before transfer, or if the oocytes are abnormal, they give rise to abnormal em- bryos. In either case, the result is poor embryonic quality, which can be a reflection of genetic abnormalities (23) and, so, results in implantation failure and later, spontaneous abortion (14). Obese women had fewer oocytes collected, therefore we propose that low oocyte count contributes to miscarriage in obesity. Since low oocyte count had a more deleterious effect on pregnancy outcome among obese than lean women, it is indeed possible that oocyte quality is also altered by obesity. This hypothesis implies that obesity has a significant influence on follicular development. In fact, several such mechanisms exist. Normal follicle growth requires a fine balance of sex steroid, gonadotropin and paracrine growth factor synthesis and action (24). Obesity may interfere with these by altered metabolism of sex steroids in excess fatty tissue (11), by hyperinsulinemia inducing hyperandrogenemia and disturbance of local growth factors (25), or by hyperleptinemia (26), which can influence intrafollicular steroid synthesis, as well (27, 28). Thus, we suggest that increased abortion rate in obese patients is, at least in part, the result of altered follicular development. This idea is supported by Cano et al. (29) who described the association of obesity and hyperinsulinemia with inferior oocyte and embryo quality in PCOS. One should keep in mind, however, that we do not exclude contribution of factors other than follicular development in causing spontaneous abortion in obese patients. Indeed, obesity independently affected pregnancy outcome. A possible factor besides follicular development is suboptimal uterine environment. In fact, animal experiments show that gonadotropin therapy alters uterine receptivity (30), which may explain the impaired implantation and embryonic development after gonadotropin treatment in mice (31, 32), and the high incidence of spontaneous abortion in human beings (17). Since obesity alters gonadotropin pharmacokinetics, and obese patients need more gonadotropin for ovarian stimulation (33), the higher gonadotropin dose possibly contributes to their poor pregnancy outcome. In summary, we found that obesity is a risk factor for spontaneous abortion after IVF or ICSI. Results suggest that this association, in part, is related to low oocyte number in obese women. Acknowledgments We thank Thore Egeland for his advice on statistics. P.F. receives a grant from the Research Council of Norway. References 1. Zaadstra BM, Seidell JC, Van Noord PA, te Velde ER, Habbema JD, Vrieswijk B et al. Fat and female fecundity:

6 48 P. Fedorcsák et al. prospective study of effect of body fat distribution on conception rates. BMJ 1993; 306: Wass P, Waldenstrom U, Rossner S, Hellberg D. An android body fat distribution in females impairs the pregnancy rate of in-vitro fertilization-embryo transfer. Hum Reprod 1997; 12: Green BB, Weiss NS, Daling JR. Risk of ovulatory infertility in relation to body weight. Fertil Steril 1988; 50: Hartz AJ, Barboriak PN, Wong A, Katayama KP, Rimm AA, Kalkhoff R et al. The association of obesity with infertility and related menstrual abnormalities in women. A study of factors associated with the ability to maintain weight loss. Int J Obes 1979; 8: Dale PO, Tanbo T, Haug E, Åbyholm T. 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 1998; 13: Hamilton-Fairley D, Kiddy D, Watson H, Paterson C, Franks S. Association of moderate obesity with a poor pregnancy outcome in women with polycystic ovary syndrome treated with low dose gonadotrophin. Br J Obstet Gynaecol 1992; 99: Homburg R, Armar NA, Eshel A, Adams J, Jacobs HS. Influence of serum luteinising hormone concentrations on ovulation, conception, and early pregnancy loss in polycystic ovary syndrome. BMJ 1988; 297: Homburg R, Levy T, Berkovitz D, Farchi J, Feldberg D, Ashkenazi J et al. Gonadotropin-releasing hormone agonist reduces the miscarriage rate for pregnancies achieved in women with polycystic ovarian syndrome. Fertil Steril 1993; 59: Tarlatzis BC, Grimbizis G. The significance of high follicular-phase luteinizing hormone levels in the treatment of women with polycystic ovarian syndrome by in vitro fertilization. J Assist Reprod Genet 1997; 14: Dale PO, Tanbo T, Vaaler S, Åbyholm T. Body weight, hyperinsulinemia, and gonadotropin levels in the polycystic ovarian syndrome: evidence of two distinct populations. Fertil Steril 1992; 58: Smith SR. The endocrinology of obesity. Endocrinol Metab Clin North Am 1996; 25: Tanbo T, Kjekshus E, Dale PO, Storeng R, Lunde O, Magnus Ø et al. Intracytoplasmatic spermieinjeksjon. [Intracytoplasmic sperm injection]. (in Norwegian with English abstract). Tidsskr Nor Laegeforen 1998; 118: Åbyholm T, Tanbo T, Dale PO, Kjekshus E, Magnus Ø. The first attempt at IVF treatment. Results and requirements for a satisfactory success rate. Eur J Obstet Gynecol Reprod Biol 1991; 38: Mills CL. Factors affecting embryological parameteres and embryo selection for IVF-ET. In: Brinsden PR, Rainsbury PA, eds. A textbook of in vitro fertilization and assisted reproduction. Carnforth: Parthenon, 1992; Risch HA, Weiss NS, Clarke EA, Miller AB. Risk factors for spontaneous abortion and its recurrence. Am J Epidemiol 1988; 128: Groll M. Endometriosis and spontaneous abortion. Fertil Steril 1984; 41: Jansen RP. Spontaneous abortion incidence in the treatment of infertility. Am J Obstet Gynecol 1982; 143: Fedele L, Bianchi S. Habitual abortion: endocrinological aspects. Curr Opin Obstet Gynecol 1995; 7: Holte J, Bergh T, Gennarelli G, Wide L. The independent effects of polycystic ovary syndrome and obesity on serum concentrations of gonadotrophins and sex steroids in premenopausal women. Clin Endocrinol (Oxf) 1994; 41: Zhang H, Bracken MB. Tree-based, two-stage risk factor analysis for spontaneous abortion. Am J Epidemiol 1996; 144: Hollmann M, Runnebaum B, Gerhard I. Effects of weight loss on the hormonal profile in obese, infertile women. Hum Reprod 1996; 11: Clark AM, Thornley B, Tomlinson L, Galletley C, Norman RJ. Weight loss in obese infertile women results in improvement in reproductive outcome for all forms of fertility treatment. Hum Reprod 1998; 13: Pellestor F, Girardet A, Andreo B, Arnal F, Humeau C. Relationship between morphology and chromosomal constitution in human preimplantation embryo. Mol Reprod Dev 1994; 39: Richards JS. Hormonal control of gene expression in the ovary. Endocr Rev 1994; 15: Poretsky L, Piper B. Insulin resistance, hypersecretion of LH, and a dual-defect hypothesis for the pathogenesis of polycystic ovary syndrome. Obstet Gynecol 1994; 84: Auwerx J, Staels B. Leptin. Lancet 1998; 351: Karlsson C, Lindell K, Svensson E, Bergh C, Lind P, Billig H et al. Expression of functional leptin receptors in the human ovary. J Clin Endocrinol Metab 1997; 82: Cioffi JA, Van Blerkom J, Antczak M, Shafer A, Wittmer S, Snodgrass HR. The expression of leptin and its receptors in pre-ovulatory human follicles. Mol Hum Reprod 1997; 3: Cano F, Garcia-Velasco JA, Millet A, Remohi J, Simon C, Pellicer A. Oocyte quality in polycystic ovaries revisited: identification of a particular subgroup of women. J Assist Reprod Genet 1997; 14: Katagiri S, Moon YS, Yuen BH. The role for the uterine insulin-like growth factor I in early embryonic loss after superovulation in the rat. Fertil Steril 1996; 65: Ertzeid G, Storeng R, Lyberg T. Treatment with gonadotropins impaired implantation and fetal development in mice. J Assist Reprod Genet 1993; 10: Ertzeid G, Storeng R. Adverse effects of gonadotrophin treatment on pre- and postimplantation development in mice. J Reprod Fertil 1992; 96: Fridström M, Sjöblom P, Pousette A, Hillensjö T. Serum FSH levels in women with polycystic ovary syndrome during ovulation induction using down-regulation and urofollitropin. Eur J Endocrinol 1997; 136: Address for correspondence: Ritsa Storeng, Ph.D. Department of Obstetrics and Gynecology, IVF Unit National Hospital Oslo 0027 Norway