Abstract
Objective:
Obesity is common in women with polycystic ovary syndrome (PCOS). PCOS and obesity are associated with reduced fertility. Metabolic syndrome’s (MetS) effect on infertility treatment success and pregnancy outcomes in women with PCOS undergoing ovulation induction (OI) has not been investigated. The objectives of this study were to determine the associations of MetS with OI live birth rate and pregnancy complications in obese women with PCOS, and to see if these outcomes differ by specific agent used for OI.
Study Design:
This prospective cohort analysis was from data collected from participants in the Pregnancy in Polycystic Ovary Syndrome II (PPCOSII) clinical trial conducted by the Reproductive Medicine Network. In PPCOSII, 750 women with PCOS and infertility were randomized to either clomiphene citrate or letrozole for OI for 1-5 cycles or until pregnancy occurred. Cox regression and modified Poisson regression, Chi-square and Student’s t or Wilcoxon tests were applied. Outcomes of interest were live birth and clinical pregnancy rates, and pregnancy complications. Having MetS was defined by the presence of at least 3 out of 5 cardiometabolic risk factors (waist circumference > 88cm, low high-density lipoprotein cholesterol < 50mg/dL, triglycerides ≥ 150mg/dL, systolic BP ≥ 130 or diastolic BP ≥ 85 mmHg, and fasting glucose > 100mg/dl). We also used a continuous MetS Z-score. BMI categories were defined as normal (BMI < 25 kg/m2), high (25 to 35 kg/m2), and very high (> 35 kg/m2).
Results:
Fewer women achieved a clinical pregnancy (20.5% vs. 29.7%, p=.007) or had a live birth (16.5% vs. 27%, p=.001) in the presence of MetS. Early pregnancy losses were not different by MetS status. At least one pregnancy complication occurred more often with MetS, 61.9% (26/42) compared to 44.4% (59/133) (p=.05) without it. Gestational diabetes (35.7% vs. 18.2 %, p=.02) and macrosomia (21.4% vs. 8.3%, p=.02) were more common in the presence of MetS. After adjustment for other potential confounders, the live birth rate ratio for a one-unit change in the MetS z-score was 0.89 (95% CI 0.79, 1.00, p=.04) for those whose BMI was 25-35 kg/m2 . For the very high BMI subgroup (>35 kg/m2) the independent effects of MetS from obesity were harder to discern. The live birth rate was higher with letrozole, although MetS had a different detrimental effect by medication given. The overall incidence of pregnancy complications was high (≃ 49%) in PPCOSII, and similar with either agent. Letrozole was associated with more obstetrical complications in the presence of MetS and clomiphene was associated with a lower live birth rate when MetS was present.
Conclusions:
MetS is a risk factor that lowers the live birth rate with OI for women with PCOS independent of obesity, and it is particularly associated with a lower live birth rate for clomiphene compared to letrozole users. It is also a risk factor for pregnancy complications for obese letrozole users. Having the MetS is a risk factor for gestational diabetes and macrosomia.
Keywords: Metabolic syndrome (MetS), polycystic ovary syndrome (PCOS), ovulation induction (OI), pregnancy complications
Condensation:
Metabolic syndrome affects live birth and pregnancy complication rates differently for women with polycystic ovary syndrome after letrozole or clomiphene citrate treatment for ovulation induction.
Introduction
Polycystic ovary syndrome (PCOS) is the most common endocrine-metabolic disorder in women of reproductive age, and the most common cause of anovulatory infertility. Up to 15% of women are affected. 1 Obesity is prevalent in women with PCOS, particularly in the U.S., although prevalence varies widely depending on ethnicity and region. Insulin resistance (IR) and dyslipidemia are common with PCOS, even in women of normal weight.2–4 Obesity, IR and dyslipidemia all increase the likelihood of a diagnosis of metabolic syndrome (MetS).5 MetS components include abdominal obesity, dyslipidemia (elevated triglycerides and lowered high-density lipoprotein cholesterol (HDL-C), elevated glucose and high blood pressure.6 The age-adjusted prevalence from the general population is increasing for women (25.3% to 34.2% from 1988 through 2012 - NHANES).7
In PCOS, oligo-ovulation and obesity both reduce fertility. Added obesity contributes to a lower success rate with in vitro fertilization (IVF).8, 9 Folliculogenesis is altered by various pathways (e.g. abnormal signaling of follicles by regulators such as Anti Mullerian Hormone, insulin-like growth factors and sex steroids) leading to an arrest of antral follicles.10, 11 Poor oocyte quality, smaller size of mature oocytes, more abnormal spindles, and chromosome misalignment are found with obesity.12 Poorer IVF performance (higher dose, longer duration of gonadotropins, fewer retrieved oocytes, and fewer available embryos) and a decreased cumulative live birth rate is found in patients with PCOS with MetS compared to those without MetS. 13
Ovulation induction (OI) with letrozole is now the first choice treatment for anovulation in PCOS-related infertility.14 Older age, a higher BMI, and a longer duration of infertility are associated with poorer success with OI.15 Interventions targeting insulin resistance (metformin, rosiglitazone, pioglitazone and d-chiro-inositol) and lifestyle changes (diet and exercise) to improve fertility outcomes in PCOS are less successful than OI.16–18 Targeting insulin resistance may improve ovulation and IR but does not necessarily improve fertility outcomes. 19 Our understanding of how MetS affects treatment success with OI for women with PCOS is limited.
In general, women with PCOS are at increased risk of pregnancy related hypertensive diseases, gestational diabetes mellitus, preterm delivery, intrauterine growth retardation (IUGR), and/or fetal macrosomia.1, 20, 21 Metabolic abnormalities in PCOS, such as hyperandrogenism, obesity and IR may contribute to more pregnancy complications.1
The impact of MetS on fertility treatment success and subsequent adverse pregnancy outcomes in women with PCOS undergoing OI has not been investigated. We hypothesized that MetS in women with PCOS adversely affects conception rates, live birth rates and pregnancy complications independent of body mass index (BMI). We wished to determine if results differ by which medication is used for OI. Data collected during the multicenter, randomized controlled clinical trial, Pregnancy in Polycystic Ovary Syndrome II (PPCOSII) trial 14, 22 offered an opportunity to assess our objectives.
Materials and Methods
Study Design
The PPCOS II trial was a multicenter, randomized controlled, double-blind clinical trial conducted at 11 clinical sites across the United States to study OI live birth rates after clomiphene citrate or letrozole treatment for anovulatory infertility due to PCOS. Although 750 infertile women underwent OI for up to five treatment cycles (1:1 ratio of letrozole and CC), 158 withdrew after a varying number (1-4) of unsuccessful attempts.
Participants were 18 to 40 years of age, without significant comorbidities. They had to have had at least one patent fallopian tube, a normal uterine cavity, and the male partner had to have a sperm concentration of at least 14 million per milliliter with documented motility according to World Health Organization cutoff points 23 in at least one ejaculate in the past year. The partners agreed to have regular intercourse with the intent of conception during the study. The design, baseline characteristics, and outcomes of the original study are published.14, 22, 24, 6 Institutional Review Board approval was obtained at each study site. The Reproductive Medicine Network conducted the trial (ClinicalTrials.gov number: NCT00719186).
Participant characteristics and relevant laboratory analysis
Participants in the PPCOSII trial fulfilled modified Rotterdam criteria for PCOS. 24 Other disorders that mimic PCOS, including thyroid disease and hyperprolactinemia were excluded.25 All measurements in this analysis were collected at screening prior to OI. A standardized protocol for all measurements was in place. A transvaginal ultrasound evaluated the ovaries for polycystic ovarian morphology.
The fasting blood samples were analyzed in batch, including glucose, lipid profile, and other biomarkers, at the core lab at the University of Virginia (Ligand Assay and Analysis Core of the Specialized Cooperative Centers Program in Reproduction and Infertility Research).22 The intra and inter-assay coefficients of variation (CV) were below 10% for each assay. Lipid profiles, including total cholesterol (TC), high-density lipoprotein cholesterol (HDL), and triglycerides (TG), were measured using the Abbott Architect c16000 automated analyzer. Low-density lipoprotein (LDL) cholesterol was estimated with the Friedewald equation.22, 24
Outcome variables and MetS
MetS was defined categorically (binary yes or no) as the presence or absence of at least 3 of 5 risk factors:1 1. Waist circumference > 88cm; 2. low high-density lipoprotein cholesterol (HDL-C < 50mg/dL); 3. hypertriglyceridemia (TG > 150mg/dL); 4. hypertension (Systolic BP ≥ 130 or diastolic BP ≥ 85 mmHg); and 5. fasting hyperglycemia (fasting glucose > 100mg/dl).6 A Z- score was calculated by first standardizing the individual MetS variables (e.g., waist circumference, systolic BP, glucose, HDL-C, and log of TG by subtracting the study population mean from the study participant’s value and dividing by the study population standard deviation. Since the standardized HDL-C is inversely related to metabolic risk, it was multiplied by −1. The standardized values of each MetS component are summed to create the Z-score.26 BMI was categorized as normal (< 25 kg/m2), high (25–35 kg/m2), or very high (>35 kg/m2). A clinical pregnancy was defined as documentation of a fetal heart beat by vaginal ultrasound. A live birth was defined as the delivery of a viable infant. Pregnancy complications included preeclampsia, gestational diabetes, preterm delivery, IUGR, macrosomia or any complication (at least one). Pregnancy outcomes were tracked through review of maternal and infant medical records.
Statistical Analysis
Comparisons of means or medians of the continuous variables were by Student’s t test or Mann-Whitney U test based on the normalcy of the distributions. Chi-Square or Fisher’s Exact tests were used to compare frequencies of patient characteristics and pregnancy outcomes by presence or absence of MetS. The cumulative incidence of pregnancy success was compared using the log rank test. Cox regression was used to compare the live birth rates for all the patients who entered the PPCOSII trial and had complete data available for MetS (n= 746). Participants who did not conceive and did not complete all treatment cycles were censored at the time of last patient contact. MetS z-score was assessed as a continuous variable within BMI strata. Unadjusted and BMI-adjusted rate ratios and 95% confidence intervals (CI) are reported for a one-unit change in MetS Z-score stratified for high (25–35 kg/m2) and very high (>35 kg/m2) BMI categories. Rate ratios were adjusted for BMI, age and smoking status. We also applied Poisson regression models with a robust error variance, 27 to examine the consistency of conclusions when using an alternative method to estimate risk ratios for live birth for those who continued in the study to completion. We calculated risk ratios and 95% CIs for those with complete follow-up data (n=592). Because patients were randomly assigned to clomiphene citrate or letrozole treatment for OI in the original clinical trial, we further examined how live birth rates differed by MetS and BMI when stratified by these two OI agents. Due to sparse numbers across strata, this assessment is limited to unadjusted comparisons of proportions using Chi-square and Fisher’s exact tests. A p-value of < 0.05 was considered statistically significant. All analyses were performed by using the Statistical Analysis System V9.4 (SAS Institute, Cary, NC) and NCSS Salt Lake City, Utah.
Results
Table 1 illustrates the demographic characteristics of the study population by presence of MetS. The majority were obese 84% (627/746). MetS was present in 34% (254/746). Those with MetS were slightly older and heavier. Only one patient in the BMI category <25 kg/m2 met criteria for MetS. The fasting glucose, blood pressure and lipid profiles reflect their MetS status. Those who had MetS were more likely to be current smokers. They gained less weight during pregnancy and there was no difference in alcohol consumption, income, education level or ovulated cycles (data not shown). Those with MetS had a longer duration of infertility. There was no difference in the percent who received clomiphene or letrozole.
Table 1.
Demographic Characteristics of PPCOSII Cohort by Metabolic Syndrome Status
| Mean (SD) Median(95%CI) Chi-square or Fisher’s Student’s t Mann-Whitney U or Wilcoxon sum Rank-Sum |
Metabolic Syndrome N=254 |
No Metabolic Syndrome N= 492 |
p-value |
| Age (years) | 29.6 (4.1) | 28.5 (4.3) | <0.01 |
| Weight (kg) | 104.3 (102, 108)) | 86.1 (80.4, 89.1) | <.01 |
| Height (cm) | 164.2 (6.1) | 164.0 (6.6) | 0.60 |
| Waist circumference (cm) | 117 (114, 119) | 100 (97, 102) | <.01 |
| Systolic blood pressure (mmHg) | 126.3 (13.7) | 116.2 (11.0) | <.01 |
| Diastolic blood pressure (mmHg) | 82.1 (9.2) | 74.5 (8.4) | <.01 |
| BMI at Screen ( kg/m2) | 38.8 (38, 41) | 32.9 (30, 33) | <.01 |
| BMI Category | <.01 | ||
| <25 kg/m2 | 0.39% | 24% | |
| 25-35 kg/m2 | 30% | 36% | |
| >35 kg/m2 | 69% | 40.5% | |
| Race | |||
| Hispanic | 21.3% | 14.8% | 0.03 |
| White | 35.6% | 64.4% | 0.85 |
| Black | 9.5% | 15.5% | 0.02 |
| Asian | 2.97% | 3.9% | 0.19 |
| American Indian | 1.2% | 0.8% | 0.62 |
| Native Hawaiian | 0.39% | 0.0% | 0.34 |
| Mixed Race | 4.7% | 3.1% | 0.25 |
| Fasting Glucose mg/dL | 90.5 (89, 92) | 84.3 (83, 86) | <.01 |
| HDL Cholesterol mg/dL | 34 (32, 35) | 39 (38, 40) | <.01 |
| Triglycerides mg/dL | 167 (157, 173) | 87 (82, 91) | <.01 |
| Hs-CRP mg/L | 6.4 (5, 7) | 3.1 (3, 4) | <.01 |
| Current Smokers | 18.5% | 12.8% | 0.04 |
| Current alcohol users | 63.8% | 62.6% | 0.75 |
| Education | 0.31 | ||
| High school or less | 22.9% | 23.4% | |
| Some College | 69.3% | 63.2% | |
| Graduate school | 7.9% | 13.4% | |
| Prior pregnancy loss | 24.4% | 22/3% | |
| History of pregnancy | 40.2% | 34.6% | 0.13 |
| Prior pregnancy loss | 24.4% | 22.6% | 0.57 |
| Duration of infertility | 36 (30, 48) | 24 (24, 24) | <.01 |
| Treatment | 0.76 | ||
| Clomiphene | 50.8% | 49.6% | |
| Letrozole | 49.2% | 50.4% | |
Figure 1 illustrates the cumulative incidence of live births for the whole cohort (A); for those whose BMI was 25-35 kg/m2 (B); and for those whose BMI was > 35 kg/m2 (C). MetS was associated with a lower cumulative live birthrate when the BMI was 25-35 kg/m2.
Figure 1.
ABC illustrates the cumulative incidence of live births for the whole cohort (A); for those whose BMI was 25-35 kg/m2 (B); and for those whose BMI was > 35 kg/m2 (C). Metabolic syndrome was associated with a lower cumulative live birth rate when the BMI was 25-35 kg/m2 (p = 0.014).
With MetS fewer women achieved a clinical pregnancy 20.5% (52/254) vs. 29.7% (146/492), p=.007 or had a live birth [16.5% (42/254) vs. 27% (133/492), p=.001; with MetS vs. without, respectively. The incidences of gestational diabetes 35.6% (16/45) vs. 17.9% (24/134), p=.014 and macrosomia 21.4% (9/42) vs. 8.3% (11/132), p=.019 were higher with MetS (both obesity strata combined).
Table 2 demonstrates the pregnancy outcomes and complication rates contrasting those with and without MetS stratified within the three BMI categories. Fewer women with MetS achieved a pregnancy or had a singleton live birth when their BMI was 25-35 kg/m2 compared to those who did not have MetS. When their BMI was >35 kg/m2 there was no statistical difference. For those who had a BMI of 25-35 kg/m2, MetS was associated with more gestational diabetes and macrosomia but not statistically different within BMI stratum. Comparisons are hampered by small numbers within strata.
Table 2.
Pregnancy Outcomes by Metabolic Syndrome and Obesity Status
| Outcome | BMI <25 kg/m2 | BMI 25-35 kg/m2 | BMI >35 kg/m2 | ||||
|---|---|---|---|---|---|---|---|
| MetS | No MetS | Mets | No MetS | MetS | No MetS | P Value | |
| Live Births | 1/1 (100%) |
42/118 (35.6%) |
11/77 (14.3%) |
52/175 (29.7%) |
30/176 (17.1%) |
39/199 (19.5%) |
0.04* 0.52** |
| Singleton | 1/1 (100%) |
39/118 (33.1%) |
11/77 (14.3%) |
50/175 (28.6%) |
28/176 (15.9%) |
37/199 (18.6%) |
0.03* 0.49** |
| Twins | 0/1 (0.0%) |
3/118 2.54% |
0/77 (0.00%) |
2/175 (1.1%) |
2/176 (1.1%) |
2/199 (1.0%) |
0.47* 0.90** |
| Achieved Pregnancy | 1/1 (100%) |
58/118 (49.2%) |
22/77 (28.6%) |
71/175 (40.6%) |
47/176 (26.7%) |
58/199 (29.2%) |
0.14* 0.60** |
| Clinical Pregnancy | 1/1 (100%) |
44/58 (75.9%) |
15/22 (68.2%) |
57/71 (80.3%) |
36/47 (76.6%) |
45/58 (77.6%) |
0.24* 0.90** |
| Loss in first trimester | 0/1 (0.0%) |
16/58 (27.6%) |
9/22 (40.9%) |
18/71 (25.4%) |
15/47 (31.9%) |
16/58 (27.6%) |
0.16* 0.63** |
| Biochemical | 0/1 (0.0%) |
7/118 (5.93%) |
2/77 (2.6%) |
5/175 (2.9%) |
6/176 (3.4%) |
6/199 (3.0%) |
0.82* 0.91** |
| Miscarriage | 0/1 (0.0%) |
7/58 (12.1%) |
6/22 (27.3%) |
11/67 (16.4%) |
9/47 (19.2%) |
11/57 (19.3%) |
0.26* 0.99** |
| Ectopic | 0/1 (0.0%) |
2/58 (3.5%) |
1/22 (4.6%) |
1/71 (1.4%) |
2/47 (4.3%) |
1/58 (1.7%) |
0.38* 0.44** |
| Loss in second or third trimester | 0/1 (0.0%) |
0/58 (0.0%) |
0/22 (0.0%) |
1/71 (1.4%) |
2/47 (4.3%) |
2/58 (3.5%) |
0.58* 0.83** |
| Preterm Labor | 0/1 (0.0%) |
7/42 (16.7%) |
1/11 (9.1) |
5/52 (9.6%) |
5/30 (16.7%) |
5/38 (13.2%) |
0.66* 0.69** |
| Gestational diabetes | 0/1 (0.0%) |
4/41 (9.8%) |
4/11 (36.4%) |
8/52 (16.4%) |
11/30 (36.7%) |
12/39 (30.8%) |
0.11* 0.61** |
| Preeclampsia | 0/1 (0.0%) |
4/42 (9.5%) |
0/11 (0.0%) |
7/52 (13.5%) |
8/30 (12.7%) |
12/39 (30.8%) |
0.20* 0.71** |
| IUGR | 0/1 (0.0%) |
1/42 (2.4%) |
0/10 (0.0%) |
2/52 (3.9%) |
0/30 (0.0%) |
3/39 (7.50%) |
0.53* 0.25** |
| Macrosomia | 0/1 (0.0%) |
0/42 (0.0%) |
2/11 (18.2%) |
7/52 (13.5%) |
7/29 (24.1%) |
4/39 (10.3%) |
0.68* 0.12** |
| Cesarean Section | 0/11 (0.0%) |
11/45 (24.4%) |
4/11 (36.4%) |
14/55 (25.5%) |
18/32 (56.3%) |
21/40 (52.5%) |
0.46* 0.75** |
| Any Pregnancy Complication | 0/1 (0.00%) |
14/42 (33.3%) |
5/11 (45.5%) |
22/52 (42.3%) |
21/30 (70.0%) |
23/39 (59.0%) |
0.89* 0.36** |
Chi-square or Fischer’s Exact test.
Metabolic syndrome effect within BMI category 25-35 25 kg/m2;
Metabolic syndrome effect within BMI category >35 25 kg/m2.
Only 1 person had metabolic syndrome in BMI category 25 kg/m2. Pregnancy losses before 20 weeks is per achieved pregnancy except biochemical pregnancies. Complications of pregnancy is per live birth.
Table 3 shows the regression results for live births for the entire cohort stratified by obesity BMI categories. Cox regression analysis (n=746 evaluable data) found that the rate ratio for live births for those with PCOS in BMI 25-35 kg/m2 was 0.89 (95% CI 0.79, 1.00), p=0.04 for the MetS Z-score, when adjusted for BMI, age and current smoking status and the adjusted risk ratio was 0.91(95% CI 0.83, 1.00) using modified Poisson Regression . There is a 10-11% lower live birth rate with a one-unit increase in the MetS z-score independent of BMI in those whose BMI was 25-35 kg/m2 adjusted for BMI and other potential confounders.
Table 3.
Live births for full cohort
| Using Cox Regression | |||||
|---|---|---|---|---|---|
| BMI category | Metabolic Syndrome (n= 254) |
No Metabolic Syndrome (n=492) |
RR (95% CI) |
Adjusted * RR (95% CI) |
p- value |
| #25-35 kg/m2 | 11/77 (14.3%) |
52/175 (29.7%) |
0.83 (0.75, 0.93) |
0.89 (0.79, 1.00a) |
0.04 |
| >35 kg/m2 | 30/176 (17.1%) |
39/199 (19.6%) |
1.02 (0.92, 1.13) |
1.04 (0.93, 1.15) |
0.50 |
| Using Modified Poisson Regression | |||||
| BMI category | Metabolic Syndrome (n= 254) |
No Metabolic Syndrome (n=492) |
RR (95% CI) |
Adjusted * RR (95% CI) |
p- value |
|
#BMI 25-35 |
11/77 (14.3%) |
52/175 (29.7%) |
0.86 (0.79, 0.93) |
0.92 (0.83, 0.99) |
0.04 |
| BMI >35 |
30/176 (17.1%) |
39/199 (19.6%) |
1.02 (0.94, 1.11) |
1.04 (0.96, 1.13) |
0.34 |
Rate Ratio for one-unit change in MetS z-score adjusted for BMI
Rate Ratio for adjusted BMI, age, current smoking = 0.89(0.79, 1.00)
Upper bound of confidence interval rounded from value of 0.997
Risk ratio for one-unit change in MetS z-score adjusted for BMI
Risk ratio adjusted for BMI, age, current smoking = 0.91 (0.83, 1.00)
Risk ratios of similar magnitude were observed in the analysis of those who completed study follow up. The prevalence of obesity remained similar in this restricted sample 82.2% (484/589).
Table 4 compares the effects of MetS on live birth rates and pregnancy complications by type of OI medication used and by MetS status. Live birth rates were higher with letrozole. The detrimental effects of having MetS on live birth rates can be seen for all BMI groups and more clearly in the BMI category 25-35 kg/m2 (p=.03), more clearly shown in clomiphene users. Overall complication rates were not different by which OI agent was used. However, having MetS was associated with more obstetrical complications in letrozole users compared to those who did not have MetS.
Table 4.
Effects of metabolic syndrome (MetS) on live birth rates and pregnancy complications by ovulation induction medication in PPCOSII
| Live Births | |||||
|---|---|---|---|---|---|
| Letrozole | P* | Clomiphene Citrate | P* | P** | |
| BMI 16.7-67.5 | 0.06 | <.01 | |||
| No MetS | 76/248(30.7%) | 57/245(23.3%) | 0.06 | ||
| MetS | 27/125(21.6%) | 15/129(11.6%) | 0.03 | ||
| Obese BMI >25 kg/m2 | 81/314(25.8%) | 0.17 | 51/313(16.3%) | 0.03 | <0.01 |
| No MetS | 54/189 (28.6%) | 37/185 (20.0%) | 0.06 | ||
| MetS | 27/125 (21.6%) | 14/128 (10.9%) | 0.02 | ||
| BMI 25-35 kg/m2 | 36/128(28.1%) | 0.14 | 27/123 (22.0%) | 0.03 | 0.26 |
| No MetS | 29/91 (31.8%) | 23/83 (27.7%) | 0.55 | ||
| MetS | 7/37 (18.9%) | 4/40 (10.0%) | 0.34 | ||
| BMI > 35 kg/m2 | 45/186(24.2%) | 0.66 | 24/190(12.6%) | 0.63 | <0.01 |
| No MetS | 25/98 (25.5%) | 14/102 (13.7%) | 0.04 | ||
| MetS | 20/88 (22.7%) | 10/88 (11.4%) | 0.05 | ||
| Pregnancy Complication | |||||
| Letrozole | P* | Clomiphene Citrate | P* | P** | |
| All BMI | 35/72(48.6%) | 0.01 | 50/103(48.5%) | 0.87 | 0.99 |
| No MetS | 31/76(40.8%) | 28/57(49.1%) | 0.33 | ||
| MetS | 19/27(70.4%) | 7/15(46.7%) | 0.19 | ||
| Obese BMI >25 kg/m2 | 44/81(54.3%) | 0.04 | 27/51 (52.9%) | 0.16 | 0.88 |
| No MetS | 25/54(46.3%) | 20/37 (54.1%) | 0.47 | ||
| MetS | 19/27(70.4% ) | 7/14 (50.0%) | 0.20 | ||
| BMI 25-35 kg/m2 | 13/36(36.1%) | 0.20 | 14/27 (51.8%) | 0.33 | 0.21 |
| No MetS | 9/29(31.0%) | 13/23 (56.5%) | 0.09 | ||
| MetS | 4/7(57.1%) | 1/ 4 (25.0%) | 0.55 | ||
| BMI > 35 kg/m2 | 31/45(68.9%) | 0.43 | 13/24(54.2%) | 0.77 | 0.23 |
| No MetS | 16/25(64.0%) | 7/14(50.0%) | 0.40 | ||
| MetS | 15/20(75.0%) | 6/10(60.0%) | 0.43 | ||
Chi-Square or Fisher’s Exact test.
P-value within medication for MetS.(categorical)
P-value between medications.
Pregnancy complications are limited to those with a live birth and refer to having had a complication during pregnancy (any). This includes preterm labor, preeclampsia, hyperemesis, gestational diabetes, IUGR, incompetent cervix, premature rupture of the membranes, and/or other
Comments
Principal findings of the study
Having MetS in the presence of obesity was associated with a reduced live birth rate and is a risk factor for gestational diabetes and macrosomia for women with PCOS who underwent OI in the PPCSOII trial. Obesity, with and without MetS, was very prevalent among women in PPCOSII as was the incidence of a subsequent pregnancy complication for those who conceived. Letrozole had a higher live birth rate, but it was associated with a higher rate of pregnancy complications in the presence of MetS. Clomiphene users had fewer live births in the presence of the MetS. Overall, pregnancy complication rates were not different between the two OI medications
Results of the study in the context of other observations
The overall prevalence of MetS was 34% (254/746) in our study. This is similar to the overall prevalence in the United States by NHANES report.7 Of interest, a study from China reported a lower prevalence (27%) of MetS in infertile women undergoing IVF for PCOS than we found in this US cohort.13 Lim et al. reported an increased prevalence of MetS in overweight and obese women with PCOS OR 1.88 ( 95% 1.16, 3.04) compared to women without PCOS in their meta-analysis.28 They also found a higher prevalence of MetS with older age, Hispanic ethnicity and lower socioeconomic status. 7
Severe obesity is associated with systemic chronic inflammation. We hypothesize that obesity and MetS are both risk factors for lower live birth rates with OI and likely contribute to the higher rates of gestational diabetes and macrosomia. This is easier to delineate in those whose BMI was 25-35 kg/m2. Separate effects of MetS from obesity were very difficult to disentangle when the BMI was >35 kg/m2 in PPCOSII. In a recent study of PCOS women who had IVF treatment in China, they found a lower cumulative live birth rate in PCOS women with MetS than those without MetS. The prevalence of obesity was significantly lower in their cohort.13 There is limited literature on the impact of MetS on fertility treatment success following OI in an obese population to compare to our findings.
The women with PCOS in the PPCOSII trial who had MetS also reported longer times trying to conceive before they entered the PPCOSII trial. This begs the question - for women with PCOS trying to conceive, not participating in a research study, who never seek out medical attention to get pregnant, how does having MetS affect time to conception? Our findings suggest that reducing metabolic dysfunction before pregnancy might make a difference.
The use of the MetS Z-score as a continuous predictor was helpful. It provided more power to discern effects in this relatively small sample size. It is possible that with a larger data set the effects of MetS may have been easier to separate from the potential effects of obesity across the spectrum of BMIs.
The overall risk of GDM in this cohort was 22.3% (39/175). This is higher than often reported (approximately 9%) in the United States.29 Our findings are consistent with other studies that have reported a 2-3 fold higher risk of GDM in women with PCOS.30, 31 Chanzi et al, in a cohort of 508 pregnant women in Greece, reported MetS increased the risk of GDM RR=3.17 (95% CI 1.06-9.50). 32
Excessive weight gain in pregnancy is associated with an increased the risk of preeclampsia and eclampsia. 33 Of interest, the women with MetS in PPCOSII gained less weight during their pregnancy than those without it. The precise cause of preeclampsia is not completely. Some have suggested a significant role of generalized inflammatory response and metabolic dysfunction.34 It is possible that lower weight gain in the MetS group during pregnancy in our study somewhat blunted the association. In a case-control study by Palomba et al., women with PCOS had more preeclampsia compared to women without (8% versus 2%).35 Naver et al. reported preeclampsia in women with PCOS who had hyperandrogenism, but not in PCOS women without hyperandrogenism, even after adjusting for BMI and parity. 36
The increased susceptibility of women with PCOS for higher pregnancy complications might be explained by various mechanisms, possibly including obesity, hyperandrogenism, more insulin resistance and lipid abnormalities. Hyperandrogenism is associated with microscopic alterations in early trophoblast invasion and placentation.37 Hyperinsulinemia is common in PCOS and is considered a contributor. 38 Hyperlipidemia may contribute to maternal and fetal morbidity, increased fetal adiposity, higher birth weight, and greater risk of GDM and preeclampsia.39–41, 42 With limited hepatic de novo lipogenesis in fetus 43, maternal lipids provide an important fuel for fetal fat accretion 44. However excess highly atherogenic cardiometabolic lipids are detrimental. Elevated maternal triglycerides and associated elevated free fatty acids contribute to macrosomia. 45, 46
Weight gain in pregnancy is a risk factor for pregnancy complications. 35, 47 The observation that the women with MetS in PPCOS gained less weight during pregnancy might be explained by patient’s compliance with weight gain recommendations by the Institute of Medicine (IOM). Lower weight gain is recommended for higher BMI groups to minimize pregnancy complications.48 Women who gain more weight than IOM guidelines exhibit a higher risk of maternal and neonatal morbidity, irrespective of their pre-pregnancy weight. 33. Metabolic abnormalities vary among different PCOS phenotypes, so to fairly compare various studies reporting associations between PCOS and pregnancy complications, it is important to consider which diagnostic criteria for PCOS were utilized.1, 49
Strengths and limitations
The major strength of this analysis is its prospective design, and that fasting serum was collected for cardiometabolic dysfunction parameters pre-conception. All serum analyses were batched and performed at a single core lab. Recruitment for participants of PPCOSII was from 11 centers in the USA. This adds to the diversity and increases the generalizability. In most other studies examining the associations between metabolic markers with pregnancy outcome, samples were collected during pregnancy instead of pre-conception. Obtaining measurements while a patient is pregnant might not necessarily reflect metabolic dysfunction at the time a pregnancy is established.
Our analysis has some important limitations. People who volunteer for a clinical trial are different from those who do not. Women were excluded from PPCOSII if they had diabetes mellitus or other uncontrolled chronic medical conditions. This provides a lower prevalence and possibly a less severe metabolic dysfunction in this cohort. The majority (66.5%) were non-Hispanic White, and this may affect generalization. Women diagnosed using NIH criteria (both androgen excess and anovulation) are more likely to have severe insulin resistance and more prevalent MetS.50 The PPCOSII trial used the Rotterdam criteria for PCOS and this definition may not include women at the same heightened risk of metabolic dysfunction as likely found with NIH criteria. This is a relatively small sample size cohort, and we might not have adequate power to detect some differences. Multiple comparisons in the studied sample could have resulted in spurious associations. Survival analysis methods considers total duration of persons ‘at risk of pregnancy’ and provides options for censoring those who were lost to follow up yet attempted some cycles. Those lost to follow up had higher BMIs, were older, were more likely non-Hispanic Black, Hispanic and/or of lower socioeconomic status. These are risk factors for higher rates of pregnancy complications. 51 It is possible that the true detrimental effect size of MetS for success or complications is larger than we found. Although we conducted our analysis using stratification by BMI, and adjusted for potential confounders; there can always be unknown confounders.
Conclusions and clinical implications
Obesity and MetS were prevalent in women with PCOS in PCOSII, as was the incidence of pregnancy complications after OI. MetS was associated with a lower live birth rate after OI and more gestational diabetes and macrosomia. Detrimental effects of having this cardiometabolic phenotype on live birth rates could be discerned independent of obesity (and likely additive) most clearly for those with a BMI >25-35 kg/m2. Although having MetS was associated with more pregnancy complications, overall letrozole users had more live births. The obstetrical complication rate was no different compared to clomiphene. For clomiphene users, having MetS was associated with fewer live births. MetS is a risk factor for poorer OI outcomes in PCOS. Our findings stress the importance of recognizing MetS in PCOS women desiring infertility treatments, and suggest that diagnosing this cardiometabolic syndrome and reversing it before infertility treatments is desirable.
AJOG at a Glance:
Why was this study conducted: To determine if metabolic syndrome, independent of obesity, affects ovulation induction live birth and pregnancy complication rates in women with the polycystic ovary syndrome in the PPCOSII trial, and to determine if there are differences by which ovulation inducing agent was used.
What are the key findings: Presence of metabolic syndrome in obese women with polycystic ovary syndrome reduces live birth success with ovulation induction.
What does this study add to what is already known: Metabolic syndrome, prevalent in obese women with polycystic ovary syndrome, is an independent and likely an additive risk factor that reduces the live birth rate following ovulation induction.
Effects of metabolic syndrome independent of obesity were most easily shown for persons whose BMI was 25-35 kg/m2.
Letrozole had a higher live birth rate than clomiphene with or without the metabolic syndrome.
Pregnancies achieved with letrozole were more likely to experience obstetrical complications in the presence of metabolic syndrome. However, overall, they were not different by which ovulation induction medication was used.
Acknowledgments
Disclosure Summary:
SA has no conflict of interest.
Dr. Hansen reports grants from Roche Diagnostics and Ferring International Pharma science Center US outside the submitted work.
Dr. Peck reports support from Ferring Pharmaceuticals outside the submitted work.
Dr. Wild reports grants from Quest Diagnostics, AbbVie, Amgen, and the NICHD outside the submitted work
Support:
The Eunice Kennedy Shriver National Institute of Child Health and Human Development (U10 HD077680) and the National Institute of General Medical Sciences (U54GM104938). The Pregnancy in Polycystic Ovary Syndrome II (PPCOSII) clinical trial was supported by grants from the NICHD (U10 HD27049, U10 HD38992, U10HD055925, U10 HD39005, U10 HD38998, U10 HD055936, U10 HD055942, U10 HD055944, and U54-HD29834)
References
- 1.Fauser BC, Tarlatzis BC, Rebar RW, et al. Consensus on women’s health aspects of polycystic ovary syndrome (PCOS): the Amsterdam ESHRE/ASRM-Sponsored 3rd PCOS Consensus Workshop Group. Fertil Steril 2012;97:28–38.e25. [DOI] [PubMed] [Google Scholar]
- 2.Azziz R, Carmina E, Dewailly D, et al. The Androgen Excess and PCOS Society criteria for the polycystic ovary syndrome: the complete task force report. Fertil Steril 2009;91:456–88. [DOI] [PubMed] [Google Scholar]
- 3.Stepto NK, Cassar S, Joham AE, et al. Women with polycystic ovary syndrome have intrinsic insulin resistance on euglycaemic-hyperinsulaemic clamp. Human reproduction (Oxford, England) 2013;28:777–84. [DOI] [PubMed] [Google Scholar]
- 4.Wild RA, Rizzo M, Clifton S, Carmina E. Lipid levels in polycystic ovary syndrome: systematic review and meta-analysis. Fertil Steril 2011;95:1073–9 e1-11. [DOI] [PubMed] [Google Scholar]
- 5.Coviello AD, Legro RS, Dunaif A. Adolescent girls with polycystic ovary syndrome have an increased risk of the metabolic syndrome associated with increasing androgen levels independent of obesity and insulin resistance. J Clin Endocrinol Metab 2006;91:492–7. [DOI] [PubMed] [Google Scholar]
- 6.Alberti KG, Eckel RH, Grundy SM, et al. Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation 2009;120:1640–5. [DOI] [PubMed] [Google Scholar]
- 7.Moore JX, Chaudhary N, Akinyemiju T. Metabolic Syndrome Prevalence by Race/Ethnicity and Sex in the United States, National Health and Nutrition Examination Survey, 1988-2012. Preventing chronic disease 2017;14:E24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Brewer CJ, Balen AH. The adverse effects of obesity on conception and implantation. Reproduction (Cambridge, England) 2010;140:347–64. [DOI] [PubMed] [Google Scholar]
- 9.Luke B Adverse effects of female obesity and interaction with race on reproductive potential. Fertility & Sterility 2017;107:868–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Jonard S, Dewailly D. The follicular excess in polycystic ovaries, due to intra-ovarian hyperandrogenism, may be the main culprit for the follicular arrest. Human reproduction update 2004;10:107–17. [DOI] [PubMed] [Google Scholar]
- 11.Franks S, Stark J, Hardy K. Follicle dynamics and anovulation in polycystic ovary syndrome. Human reproduction update 2008;14:367–78. [DOI] [PubMed] [Google Scholar]
- 12.Machtinger R, Combelles CM, Missmer SA, Correia KF, Fox JH, Racowsky C. The association between severe obesity and characteristics of failed fertilized oocytes. Hum Reprod 2012;27:3198–207. [DOI] [PubMed] [Google Scholar]
- 13.He Y, Yao L, Zhu Q, et al. Influence of metabolic syndrome on female fertility and in vitro fertilization outcomes in PCOS women. American Journal of Obstetrics and Gynecology 2019. [DOI] [PubMed] [Google Scholar]
- 14.Legro RS, Brzyski RG, Diamond MP, et al. Letrozole versus clomiphene for infertility in the polycystic ovary syndrome. The New England journal of medicine 2014;371:119–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Veltman-Verhulst SM, Fauser BCJM, Eijkemans. High singleton live birth rate confirmed after ovulation induction in women with anovulatory polycystic ovary syndrome: validation of a prediction model for clinical practice. Fertility and sterility 2012;98:761–68.e1. [DOI] [PubMed] [Google Scholar]
- 16.Tang T, Glanville J, Hayden CJ, White D, Barth JH, Balen AH. Combined lifestyle modification and metformin in obese patients with polycystic ovary syndrome. A randomized, placebo-controlled, double-blind multicentre study. Human reproduction (Oxford, England) 2006;21:80–9. [DOI] [PubMed] [Google Scholar]
- 17.Sahin Y, Yirmibes U, Kelestimur F, Aygen E. The effects of metformin on insulin resistance, clomiphene-induced ovulation and pregnancy rates in women with polycystic ovary syndrome. European journal of obstetrics, gynecology, and reproductive biology 2004;113:214–20. [DOI] [PubMed] [Google Scholar]
- 18.Ladson G, Dodson WC, Sweet SD, et al. Effects of metformin in adolescents with polycystic ovary syndrome undertaking lifestyle therapy: a pilot randomized double-blind study. Fertil Steril 2011;95:2595–8.e1-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Legro RS, Barnhart HX, Schlaff WD, et al. Clomiphene, metformin, or both for infertility in the polycystic ovary syndrome. N Engl J Med 2007;356:551–66. [DOI] [PubMed] [Google Scholar]
- 20.Palomba S, de Wilde MA, Falbo A, Koster MP, La Sala GB, Fauser BC. Pregnancy complications in women with polycystic ovary syndrome. Human reproduction update 2015;21:575–92. [DOI] [PubMed] [Google Scholar]
- 21.Teede HJ, Misso ML, Costello MF, et al. Recommendations from the international evidence-based guideline for the assessment and management of polycystic ovary syndrome. Fertil Steril 2018;110:364–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Legro RS, Kunselman AR, Brzyski RG, et al. The Pregnancy in Polycystic Ovary Syndrome II (PPCOS II) trial: rationale and design of a double-blind randomized trial of clomiphene citrate and letrozole for the treatment of infertility in women with polycystic ovary syndrome. Contemporary clinical trials 2012;33:470–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Cooper TG, Noonan E, von Eckardstein S, et al. World Health Organization reference values for human semen characteristics. Human reproduction update 2010;16:231–45. [DOI] [PubMed] [Google Scholar]
- 24.Legro RS, Brzyski RG, Diamond MP, et al. The Pregnancy in Polycystic Ovary Syndrome II study: baseline characteristics and effects of obesity from a multicenter randomized clinical trial. Fertil Steril 2014;101:258–69 e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 2004;81:19–25. [DOI] [PubMed] [Google Scholar]
- 26.Eisenmann JC. On the use of a continuous metabolic syndrome score in pediatric research. Cardiovascular diabetology 2008;7:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Zou G A modified poisson regression approach to prospective studies with binary data. American journal of epidemiology 2004;159:702–6. [DOI] [PubMed] [Google Scholar]
- 28.Lim SS, Kakoly NS, Tan JWJ, et al. Metabolic syndrome in polycystic ovary syndrome: a systematic review, meta-analysis and meta-regression. Obes Rev 2019;20:339–52. [DOI] [PubMed] [Google Scholar]
- 29.Zhu Y, Zhang C. Prevalence of Gestational Diabetes and Risk of Progression to Type 2 Diabetes: a Global Perspective. Current diabetes reports 2016;16:7–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Qin JZ, Pang LH, Li MJ, Fan XJ, Huang RD, Chen HY. Obstetric complications in women with polycystic ovary syndrome: a systematic review and meta-analysis. Reproductive biology and endocrinology : RB&E 2013;11:56–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Vanky E, Stridsklev S, Heimstad R, et al. Metformin versus placebo from first trimester to delivery in polycystic ovary syndrome: a randomized, controlled multicenter study. The Journal of clinical endocrinology and metabolism 2010;95:E448–E55. [DOI] [PubMed] [Google Scholar]
- 32.Chatzi L, Plana E, Daraki V, et al. Metabolic syndrome in early pregnancy and risk of preterm birth. American journal of epidemiology 2009;170:829–36. [DOI] [PubMed] [Google Scholar]
- 33.Truong YN, Yee LM, Caughey AB, Cheng YW. Weight gain in pregnancy: does the Institute of Medicine have it right? Am J Obstet Gynecol 2015;212:362.e1-8. [DOI] [PubMed] [Google Scholar]
- 34.Lau SY, Guild S-J, Barrett CJ, et al. Tumor Necrosis Factor-Alpha, Interleukin-6, and Interleukin-10 Levels are Altered in Preeclampsia: A Systematic Review and Meta-Analysis. American Journal of Reproductive Immunology 2013;70:412–27. [DOI] [PubMed] [Google Scholar]
- 35.Palomba S, Falbo A, Chiossi G, et al. Low-grade chronic inflammation in pregnant women with polycystic ovary syndrome: a prospective controlled clinical study. The Journal of clinical endocrinology and metabolism 2014;99:2942–51. [DOI] [PubMed] [Google Scholar]
- 36.Naver KV, Grinsted J, Larsen SO, et al. Increased risk of preterm delivery and pre-eclampsia in women with polycystic ovary syndrome and hyperandrogenaemia. BJOG : an international journal of obstetrics and gynaecology 2014;121:575–81. [DOI] [PubMed] [Google Scholar]
- 37.Fornes R, Maliqueo M, Hu M, et al. The effect of androgen excess on maternal metabolism, placental function and fetal growth in obese dams. Scientific reports 2017;7:8066–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Diamanti-Kandarakis E, Dunaif A. Insulin resistance and the polycystic ovary syndrome revisited: an update on mechanisms and implications. Endocrine reviews 2012;33:981–1030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Sanchez-Vera I, Bonet B, Viana M, et al. Changes in plasma lipids and increased low-density lipoprotein susceptibility to oxidation in pregnancies complicated by gestational diabetes: consequences of obesity. Metabolism 2007;56:1527–33. [DOI] [PubMed] [Google Scholar]
- 40.Spracklen CN, Smith CJ, Saftlas AF, Robinson JG, Ryckman KK. Maternal hyperlipidemia and the risk of preeclampsia: a meta-analysis. American journal of epidemiology 2014;180:346–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Wiznitzer A, Mayer A, Novack V, et al. Association of lipid levels during gestation with preeclampsia and gestational diabetes mellitus: a population-based study. Am J Obstet Gynecol 2009;201:482.e1-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Nasioudis D, Doulaveris G, Kanninen TT. Dyslipidemia in pregnancy and maternal-fetal outcome. Minerva Ginecol 2019;71:155–62. [DOI] [PubMed] [Google Scholar]
- 43.McCormick KL, Susa JB, Widness JA, Singer DB, Adamsons K, Schwartz R. Chronic hyperinsulinemia in the fetal rhesus monkey: effects on hepatic enzymes active in lipogenesis and carbohydrate metabolism. Diabetes 1979;28:1064–8. [DOI] [PubMed] [Google Scholar]
- 44.Barrett HL, Dekker Nitert M, McIntyre HD, Callaway LK. Normalizing metabolism in diabetic pregnancy: is it time to target lipids? Diabetes care 2014;37:1484–93. [DOI] [PubMed] [Google Scholar]
- 45.Barbour LA, Farabi SS, Friedman JE, et al. Postprandial Triglycerides Predict Newborn Fat More Strongly than Glucose in Women with Obesity in Early Pregnancy. Obesity (Silver Spring, Md) 2018;26:1347–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Barbour LA, Hernandez TL. Maternal Lipids and Fetal Overgrowth: Making Fat from Fat. Clin Ther 2018;40:1638–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Kominiarek MA, Peaceman AM. Gestational weight gain. American journal of obstetrics and gynecology 2017;217:642–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Institute of M, National Research Council Committee to Reexamine IOMPWG. The National Academies Collection: Reports funded by National Institutes of Health. In: Rasmussen KM, Yaktine AL, eds. Weight Gain During Pregnancy: Reexamining the Guidelines. Washington (DC): National Academies Press (US) Copyright © 2009, National Academy of Sciences., 2009. [PubMed] [Google Scholar]
- 49.Moran LJ, Norman RJ, Teede HJ. Metabolic risk in PCOS: phenotype and adiposity impact. Trends Endocrinol Metab 2015;26:136–43. [DOI] [PubMed] [Google Scholar]
- 50.Welt CK, Gudmundsson JA, Arason G, et al. Characterizing discrete subsets of polycystic ovary syndrome as defined by the Rotterdam criteria: the impact of weight on phenotype and metabolic features. The Journal of clinical endocrinology and metabolism 2006;91:4842–8. [DOI] [PubMed] [Google Scholar]
- 51.Harelick L, Viola D, Tahara D. Preconception health of low socioeconomic status women: assessing knowledge and behaviors. Women’s health issues : official publication of the Jacobs Institute of Women’s Health 2011;21:272–6. [DOI] [PubMed] [Google Scholar]