Objective
To evaluate the association between body mass index (BMI) and good perinatal outcomes after in vitro fertilization (IVF) among women with polycystic ovary syndrome (PCOS).
Design
Retrospective cohort study using 2012-2015 Society for Assisted Reproductive Technology Clinic Outcomes Reporting System data.
Setting
Fertility clinics.
Patient(s)
To identify patients most likely to have PCOS, we included women with a diagnosis of ovulation disorder and serum antimüllerian hormone >4.45 ng/mL. Exclusion criteria included age ≥ 41 years, secondary diagnosis of diminished ovarian reserve, preimplantation genetic testing, and missing BMI or primary outcome data.
Intervention(s)
None
Main Outcome Measure(s)
Good perinatal outcome, defined as a singleton live birth at ≥ 37 weeks with birth weight ≥ 2,500 g and ≤ 4,000 g.
Result(s)
The analysis included 9,521 fresh, autologous IVF cycles from 8,351 women. Among women with PCOS, the proportion of cycles with a good perinatal outcome was inversely associated with BMI: underweight 25.1%, normal weight 22.7%, overweight 18.9%, class I 18.4%, class II 14.9%, and class III or super obesity 12.2%. After adjusting for confounders, women in the highest BMI category had 51% reduced odds of a good perinatal outcome compared with normal weight women (adjusted odds ratio 0.49, 95% confidence interval 0.36-0.67).
Conclusion(s)
Among women with PCOS undergoing fresh, autologous IVF, the odds of a good perinatal outcome decline with increasing BMI. Women with PCOS should be counseled that the odds of achieving a good perinatal outcome decrease as their weight increases.
Key Words
The prevalence of obesity in the United States has steadily risen over the past two decades, yielding an increasing prevalence of obesity among women entering pregnancy (
1
, 2
). Maternal obesity contributes to poor obstetrical outcomes, including miscarriage, stillbirth, gestational diabetes, and preeclampsia (3
, 4
, 5
, 6
, 7
). Obesity is also associated with adverse outcomes after assisted reproductive technology (ART), including a reduction in live birth rates with increasing body mass index (BMI) (8
, 9
).- Provost M.P.
- Acharya K.S.
- Acharya C.R.
- Yeh J.S.
- Steward R.G.
- Eaton J.L.
- et al.
Pregnancy outcomes decline with increasing body mass index: analysis of 239,127 fresh autologous in vitro fertilization cycles from the 2008-2010 society for assisted reproductive technology registry.
Fertil Steril. 2016; 105: 663-669
Polycystic ovary syndrome (PCOS) is a common endocrinopathy associated with obesity and infertility. The prevalence of infertility is approximately 50% among women with PCOS, whereas obesity rates range from 30 to 60% (
10
, 11
). Women with PCOS undergoing IVF are at higher risk of miscarriage and obstetrical complications such as gestational diabetes, hypertensive disorders of pregnancy, preterm delivery, and large or small for gestational age infants (12
, 13
, 14
, 15
).Despite the known association between PCOS and obesity, there are limited data on the effect of obesity on in vitro fertilization (IVF) outcomes among women with PCOS, and no prior studies have examined adverse perinatal outcomes such as multiple birth, prematurity, or extremes of birth weight (
16
, 17
, 18
). As obesity rates continue to climb, we expect to see a growing number of women with obesity and PCOS seeking treatment for infertility. Thus, we aimed to examine the association between BMI and the odds of a good perinatal outcome after fresh autologous IVF in women with PCOS. We hypothesized that the odds of a good perinatal outcome would decrease with increasing BMI.Materials and methods
This was a retrospective cohort study and was declared exempt by the Duke Health institutional review board (Pro00100912) as the data were de-identified. All 2012-2015 fresh, autologous IVF cycles with patient age <41 and no use of preimplantation genetic testing were queried from the Society for Assisted Reproductive Technology Clinic Outcomes Reporting System (SART CORS). The SART CORS database contains comprehensive data from >90% of all clinics performing ART cycles in the United States. The data were collected through voluntary submission, verified by SART, and then reported to the Centers for Disease Control and Prevention in compliance with the Fertility Clinic Success Rate and Certification Act of 1992 (Public Law 102-493). The SART maintains HIPAA-compliant business associates' agreements with reporting clinics. In 2004, after a contract change with Centers for Disease Control and Prevention, SART gained access to the SART CORS data system for the purposes of conducting research. The data in the SART CORS are validated annually, with select clinics having on-site visits for chart review based on an algorithm for clinic selection (
19
). During each visit, data reported by the clinic were verified with information recorded in patients’ charts. In 2012, records for 2,045 cycles at 35 clinics were randomly selected for full validation, along with 238 egg or embryo banking cycles (19
). The full validation included a review of 1,318 cycles for which pregnancy was reported. Among the nondonor cycles, 331 were multiple-fetus pregnancies. Ten out of 11 data fields selected for validation were found to have discrepancy rates of ≤5%. The exception was the diagnosis field, which, depending on the diagnosis, had a discrepancy rate between 2.1% and 9.2% (19
).The analysis included women with a diagnosis of “ovulation disorder polycystic ovaries.” Women with a secondary infertility diagnosis of diminished ovarian reserve were excluded. Women were also excluded if their indication for ART was designated as “other” with “hypothalamic,” “hypogonadotropic hypogonadism,” or “hypo-hypo” in the comments. The SART CORS data field “ovulation disorder polycystic ovaries” is broadly defined as “one or more disorders causing reduced fecundity associated with structural, anatomic, or functional impairment of one or both ovaries.” Therefore, this category includes women with ovulatory disorders other than PCOS. Recognizing this limitation, we used antimüllerian hormone (AMH) as a surrogate marker to help differentiate PCOS from diminished ovarian reserve or primary ovarian insufficiency. Although not part of the formal diagnostic criteria, AMH is known to be higher in patients with PCOS (
20
, 21
, 22
, 23
). We selected an AMH cutoff of 4.45 ng/mL, which has been shown to have 76% sensitivity and 75% specificity for PCOS (20
). Women with an AMH below this cutoff point were excluded from the analyses.Patients were assigned to BMI categories according to the World Health Organization BMI guidelines: underweight (< 18.5 kg/m2), normal (18.5–24.9 kg/m2), overweight (25.0–29.9 kg/m2), class I obesity (30.0–34.9 kg/m2), class II obesity (35.0–39.9 kg/m2), class III obesity (40.0–49.9 kg/m2), and super obesity (≥ 50.0 kg/m2) (
24
). Women were excluded if BMI data was missing. Super obesity was left as a standalone category for descriptive tables, but because of the low frequency and lack of events, this category was grouped with class III obesity for statistical analyses. Race and ethnicity were categorized as “Hispanic or Latina,” “Non-hispanic White,” “Non-hispanic Black,” “Other,” and “Unknown,” as reported by each clinic to the SART CORS. The term “Unknown” was used to identify those patients without race or ethnicity documented in the medical record.The primary outcome was a “good perinatal outcome,” defined as a singleton live birth at ≥ 37 weeks with birth weight ≥ 2,500 g and ≤ 4,000 g. Women were further excluded if data were missing for any component of the composite primary outcome. Secondary outcomes included live birth rate, defined as the proportion of cycles with a live birth entered in SART CORS; multiple birth rate, calculated as the proportion of live births with two or more infants; and prematurity rate, defined as the proportion of live births occurring < 37 weeks gestational age. Low birth weight was defined as birth weight < 2,500g, and macrosomia was defined as birth weight > 4,000g. Implantation rate, as defined by SART CORS, was the greater of the number of fetal hearts on ultrasound or the number of live births plus still births, divided by the total number of embryos transferred. The clinical pregnancy rate was defined as the proportion of cycles with a gestational sac on first-trimester ultrasound. Miscarriage was defined as clinical pregnancy ending before 24 completed weeks of gestation. Ovarian hyperstimulation syndrome was defined as abdominal distension and discomfort with nausea, vomiting, diarrhea, ovarian enlargement, and ultrasonic evidence of ascites.
Analyses were performed in SAS 9.4 (SAS Institute Inc., Cary, NC) at a two-tailed significance level of 0.05. All available cycles within the study period for a woman were included in the analysis. Marginal models using generalized estimating equations (GEE) were implemented to account for the correlation between repeated IVF cycles. A GEE-type model with a binomial distribution and logit link was implemented to test the association between BMI and a good perinatal outcome with and without the following covariates: age, race, parity, infertility diagnoses, and smoking status. Secondary outcomes were analyzed using GEE-type models without adjustment for covariates. An autoregressive(
1
) working correlation structure was used along with the robust variance estimator for all models.Cycles with missing values for the primary outcome were excluded from the analysis. Two cycles with missing parity were also excluded. Given that BMI was missing for > 10% of cycles, we confirmed that outcomes were clinically similar between cycles with missing and nonmissing BMI. Cycles with missing values in BMI were then excluded from the analysis. As a sensitivity analysis, an additional BMI category was created for the cycles with missing BMI. The unadjusted and adjusted primary outcome models were then refitted with this additional BMI category included as a covariate.
Results
Among the 10,921 identified cycles, 118 (1.1%) were excluded because of a diagnosis of diminished ovarian reserve or hypothalamic amenorrhea (Supplemental Figure 1, available online). Forty-six (0.4%) cycles were excluded for missing good perinatal outcomes, 2 (0.02%) for missing parity, and 1,234 (11.3%) for missing BMI. A total of 9,521 cycles among 8,351 unique women were included in the analysis. The mean age across all cycles was 31.5 years, and the median AMH was 7.9 ng/mL (Table 1). Nearly half of the cycles (45.0%) had a secondary cause of infertility in addition to PCOS, most commonly male factor (32.6%).
TABLE 1Demographic and baseline characteristics among women with PCOS who underwent fresh, autologous IVF, 2012-2015.
| Characteristic | Underweight (N=231) | Normal (N=4134) | Overweight (N=2254) | Class I Obesity (N=1556) | Class II Obesity (N=937) | Class III Obesity (N=381) | Super Obesity (N=28) | Total (N=9521) |
|---|---|---|---|---|---|---|---|---|
| Age (y), mean (SD) | 30.5 ± 3.6 | 31.3 ± 3.5 | 31.5 ± 3.7 | 31.6 ± 3.8 | 31.6 ± 3.7 | 31.7 ± 3.6 | 31.9 ± 4.1 | 31.4 ± 3.6 |
| AMH (ng/ml), median (25th percentile, 75th percentile) | 7.8 (5.8, 11.3) | 8.0 (5.9, 11.5) | 8.2 (6.1, 12.0) | 8.0 (5.9, 11.1) | 7.6 (5.9, 10.7) | 7.5 (5.9, 10.0) | 7.2 (5.7, 10.4) | 8.0 (5.9, 11.3) |
| Race, N (%) | ||||||||
| Hispanic or Latina | 3 (1.3%) | 143 (3.5%) | 136 (6.0%) | 109 (7.0%) | 81 (8.6%) | 21 (5.5%) | 1 (3.6%) | 494 (5.2%) |
| Non-hispanic Black | 5 (2.2%) | 100 (2.4%) | 150 (6.7%) | 118 (7.6%) | 58 (6.2%) | 27 (7.1%) | 5 (17.9%) | 463 (4.9%) |
| Non-hispanic White | 121 (52.4%) | 2354 (56.9%) | 1141 (50.6%) | 820 (52.7%) | 553 (59.0%) | 224 (58.8%) | 17 (60.7%) | 5230 (54.9%) |
| Other | 32 (13.9%) | 521 (12.6%) | 275 (12.2%) | 157 (10.1%) | 44 (4.7%) | 5 (1.3%) | 0 (0.0%) | 1034 (10.9%) |
| Unknown | 70 (30.3%) | 1016 (24.6%) | 552 (24.5%) | 352 (22.6%) | 201 (21.5%) | 104 (27.3%) | 5 (17.9%) | 2300 (24.2%) |
| Nulligravid, N (%) | 159 (68.8%) | 2526 (61.1%) | 1263 (56.0%) | 850 (54.6%) | 524 (55.9%) | 216 (56.7%) | 17 (60.7%) | 5555 (58.3%) |
| Nulliparous, N (%) | 192 (83.1%) | 3343 (80.9%) | 1783 (79.1%) | 1227 (78.9%) | 756 (80.7%) | 293 (76.9%) | 25 (89.3%) | 7619 (80.0%) |
| Smoker, N (%) | 10 (4.3%) | 129 (3.1%) | 111 (4.9%) | 89 (5.7%) | 61 (6.5%) | 28 (7.3%) | 3 (10.7%) | 431 (4.5%) |
| Additional infertility diagnoses, n (%) | ||||||||
| Male infertility | 73 (31.6%) | 1339 (32.4%) | 739 (32.8%) | 518 (33.3%) | 309 (33.0%) | 119 (31.2%) | 6 (21.4%) | 3103 (32.6%) |
| Endometriosis | 11 (4.8%) | 251 (6.1%) | 120 (5.3%) | 63 (4.0%) | 36 (3.8%) | 12 (3.1%) | 1 (3.6%) | 494 (5.2%) |
| Uterine factor | 7 (3.0%) | 170 (4.1%) | 108 (4.8%) | 83 (5.3%) | 39 (4.2%) | 13 (3.4%) | 1 (3.6%) | 421 (4.4%) |
| Tubal factor | 16 (6.9%) | 248 (6.0%) | 210 (9.3%) | 154 (9.9%) | 77 (8.2%) | 48 (12.6%) | 2 (7.1%) | 755 (7.9%) |
| Other | 15 (6.5%) | 147 (3.6%) | 98 (4.3%) | 72 (4.6%) | 31 (3.3%) | 22 (5.8%) | 2 (7.1%) | 387 (4.1%) |
SD = standard deviation; AMH = anti-Müllerian hormone.
The median number of oocytes retrieved ranged from 19 in the normal BMI group to 15 in the group with class III obesity and super obesity (Table 2). The number of supernumerary embryos cryopreserved decreased from a median of three in the normal weight and overweight categories to two in women with any stage of obesity. There were no significant differences in the use of intracytoplasmic sperm injection or cleavage stage versus blastocyst transfer between BMI categories.
TABLE 2Fresh autologous IVF cycle characteristics among women with PCOS, 2012-2015.
| Cycle characteristic | N | Underweight | Normal | Overweight | Class I obesity | Class II obesity | Class III obesity or super obesity | P |
|---|---|---|---|---|---|---|---|---|
| Total FSH dose, IU | 9320 | 1802.2 ± 832.4 | 1900.6 ± 899.6 | 2058.0 ± 926.0 | 2302.5 ± 1010.2 | 2634.3 ±1198.9 | 2900.6 ± 1267.6 | |
| Beta (95% CI) | -93.7 (-209.8–22.3) | Reference | 155.7 (105.5–205.8) | 391.5 (333.2–449.9) | 686.2 (602.6–769.9) | 1022.6 (896.2–1149.1) | <.001 | |
| Oocytes retrieved | 9521 | 18.0 (13.0, 25.0) | 19.0 (13.0, 26.0) | 18.0 (12.0, 25.0) | 18.0 (12.0, 25.0) | 16.0 (10.0, 23.0) | 15.0 (9.0, 22.0) | |
| Beta (95% CI) | -0.78 (-2.26–0.69) | Reference | -0.78 (-1.35 to -0.21) | -1.23 (-1.88 to -0.58) | -2.86 (-3.64 to -2.09) | -4.03 (-5.10 to -2.95) | <.001 | |
| Embryos cryopreserved | 9521 | 4.0 (1.0, 7.0) | 3.0 (0.0, 7.0) | 3.0 (0.0, 7.0) | 2.0 (0.0, 6.0) | 2.0 (0.0, 5.0) | 2.0 (0.0, 5.0) | |
| Beta (95% CI) | -0.02 (-0.62–0.57) | Reference | -0.35 (-0.60 to -0.10) | -0.85 (-1.11 to -0.58) | -1.06 (-1.40 to -0.73) | -1.33 (-1.73 to -0.93) | <.001 | |
| ICSI N (%) | 9521 | 158 (68.4) | 2854 (69.0) | 1593 (70.7) | 1053 (67.7) | 640 (68.3) | 285 (69.7) | |
| Odds ratio (95% CI) | 0.96 (0.72–1.29) | Reference | 1.09 (0.97–1.22) | 0.94 (0.83 – 1.07) | 1.00 (0.85 ˗ 1.17) | 1.04 (0.83 ˗ 1.31) | .488 | |
| Assisted hatching, N (%) | 9521 | 39 (16.9) | 675 (16.3) | 418 (18.5) | 303 (19.5) | 168 (17.9) | 83 (20.3) | |
| Odds ratio (95% CI) | 1.03 (0.71–1.49) | Reference | 1.18 (1.03–1.35) | 1.23 (1.06 – 1.44) | 1.13 (0.93 – 1.37) | 1.35 (1.04 – 1.74) | .036 | |
| Blastocyst transfer N(%) | 6834 | 122 (75.3) | 2146 (74.0) | 1209 (74.0) | 848 (73.9) | 486 (71.3) | 206 (66.9) | |
| Odds ratio (95% CI) | 1.10 (0.75–1.60) | Reference | 0.99 (0.86–1.14) | 0.97 (0.83–1.14) | 0.87 (0.72–1.05) | 0.72 (0.55–0.93) | .178 | |
| Embryos transferred, | 6834 | 2.0 (1.0, 2.0) | 2.0 (1.0, 2.0) | 2.0 (1.0, 2.0) | 2.0 (1.0, 2.0) | 2.0 (1.0, 2.0) | 2.0 (1.0, 2.0) | |
| Beta (95% CI) | -0.03 (-0.12–0.06) | Reference | 0.07 (0.04–0.11) | 0.08 (0.04–0.12) | 0.12 (0.08–0.17) | 0.16 (0.09–0.23) | <.001 |
FSH = follicle stimulating hormone; ICSI = intracytoplasmic sperm injection.
∗ P value for overall association between BMI and cycle characteristic.
a Reported as mean (standard deviation).
b Reported as median (25th percentile, 75th percentile).
c Reported among cycles with transfer attempted.
With increasing BMI, women had lower rates of implantation, clinical pregnancy, and live birth (Table 3). Elevated BMI was also significantly associated with miscarriage, macrosomia, and prematurity, but not multiple birth or low birth weight. Among women with BMI in the overweight range or higher, rates of ovarian hyperstimulation syndrome declined with increasing BMI.
TABLE 3Associations between BMI and cycle outcomes among women with PCOS undergoing fresh, autologous IVF, 2012-2015.
| Cycle outcome | N | Underweight | Normal | Overweight | Class I obesity | Class II obesity | Class III obesity or super obesity | P |
|---|---|---|---|---|---|---|---|---|
| Implantation | --- | 0.49 | 0.48 | 0.43 | 0.40 | 0.37 | 0.32 | --- |
| Clinical pregnancy n (%) | 9521 | 105 (45.5) | 1828 (44.2) | 958 (42.5) | 657 (42.2) | 362 (38.6) | 138 (33.7) | |
| Odds ratio (95% CI) | 1.04 (0.80–1.35) | Reference | 0.93 (0.84–1.03) | 0.93 (0.82–1.04) | 0.79 (0.68–0.92) | 0.65 (0.52–0.81) | <.001 | |
| Live birth n (%) | 9521 | 93 (40.3) | 1591 (38.5) | 796 (35.3) | 527 (33.9) | 293 (31.3) | 107 (26.2) | |
| Odds ratio (95% CI) | 1.07 (0.82–1.39) | Reference | 0.87 (0.78–0.97) | 0.82 (0.72–0.93) | 0.73 (0.62–0.84) | 0.57 (0.45–0.72) | <.001 | |
| Miscarriage n (%) | 4048 | 11 (10.5) | 208 (11.4) | 148 (15.4) | 112 (17.0) | 67 (18.5) | 30 (21.7) | |
| Odds ratio (95% CI) | 0.80 (0.41–1.57) | Reference | 1.41 (1.12–1.77) | 1.67 (1.30–2.15) | 1.74 (1.27–2.37) | 2.31 (1.50–3.57) | <.001 | |
| Multiple birth n (%) | 3407 | 23 (24.7) | 407 (25.6) | 239 (30.0) | 127 (24.1) | 81 (27.6) | 38 (35.5) | |
| Odds ratio (95% CI) | 0.95 (0.59–1.55) | Reference | 1.24 (1.03–1.50) | 0.92 (0.73–1.16) | 1.11 (0.84–1.47) | 1.61 (1.06–2.43) | .055 | |
| Low birth weight n (%) | 3380 | 21 (22.8) | 386 (24.4) | 190 (24.1) | 125 (23.9) | 79 (27.3) | 30 (28.3) | |
| Odds ratio (95% CI) | 0.91 (0.55–1.51) | Reference | 0.98 (0.81–1.20) | 0.97 (0.77–1.23) | 1.16 (0.88–1.54) | 1.23 (0.79 –1.90) | .815 | |
| Macrosomia n (%) | 3380 | 2 (2.2) | 68 (4.3) | 48 (6.1) | 45 (8.6) | 23 (8.0) | 4 (3.8) | |
| Odds ratio (95% CI) | 0.49 (0.12–2.05) | Reference | 1.42 (0.97–2.08) | 2.00 (1.32–3.03) | 2.07 (1.23–3.48) | 0.88 (0.31–2.46) | .002 | |
| Prematurity n (%) | 3394 | 22 (23.7) | 402 (25.3) | 241 (30.4) | 151 (28.8) | 94 (32.4) | 36 (33.6) | |
| Odds ratio (95% CI) | 0.91 (0.56–1.49) | Reference | 1.28 (1.06–1.55) | 1.19 (0.96 ˗ 1.49) | 1.41 (1.07–1.84) | 1.50 (0.99–2.28) | .024 | |
| OHSS n (%) | 9521 | 37 (16.0) | 683 (16.5) | 320 (14.2) | 197 (12.7) | 104 (11.1) | 41 (10.0) | |
| Odds ratio (95% CI) | 0.96 (0.67–1.38) | Reference | 0.84 (0.72– 0.97) | 0.73 (0.62 ˗ 0.87) | 0.63 (0.51 ˗ 0.79) | 0.56 (0.40 –0.79) | <.001 |
CI = confidence interval; OHSS = ovarian hyperstimulation syndrome.
∗ P value for overall association between BMI and outcome.
a Implantation rate presented as ratio of implantations per number of embryos transferred.
b Reported among all cycles.
c Reported among cycles with clinical pregnancy.
d Reported among cycles with live birth.
The proportion of cycles with a good perinatal outcome was highest in the women with underweight BMI (25.1%) and decreased with increasing BMI to a low of 12.2% in the group with class III obesity or super obesity (Table 4, unadjusted odds ratio [OR] 0.47, 95% confidence interval [CI], 0.35–0.65; P<.001, compared with normal weight group).
TABLE 4Association between BMI and good perinatal outcome among women with PCOS, 2012-2015.
| BMI category | Good perinatal outcome | Unadjusted | Adjusted | ||
|---|---|---|---|---|---|
| N (%) | OR (95% CI) | P | OR (95% CI) | P | |
| Underweight | 58 (25.1) | 1.14 (0.84–1.54) | .401 | 1.14 (0.84–1.55) | .396 |
| Normal | 940 (22.7) | Reference | — | Reference | — |
| Overweight | 425 (18.9) | 0.79 (0.70–0.90) | <.001 | 0.83 (0.73–0.94) | .005 |
| Class I obesity | 286 (18.4) | 0.76 (0.66–0.89) | <.001 | 0.81 (0.70–0.94) | .006 |
| Class II obesity | 140 (14.9) | 0.60 (0.49–0.72) | <.001 | 0.62 (0.51–0.75) | <.001 |
| Class III obesity/Super obesity | 50 (12.2) | 0.47 (0.35–0.65) | <.001 | 0.49 (0.36–0.67) | <.001 |
BMI = body mass index; OR = odds ratio; CI = confidence intervals.
∗ Adjusted for age, race, parity, secondary infertility diagnoses, and smoking status.
a Good perinatal outcome defined as a singleton live birth at ≥ 37 weeks with birth weight ≥ 2,500 g and ≤ 4,000 g.
The multivariable GEE model demonstrated that BMI was significantly and inversely associated with the odds of a good perinatal outcome after adjusting for age, race, parity, infertility diagnoses other than PCOS, and smoking status (Table 4). Compared with women with normal BMI, the odds of a good perinatal outcome were reduced by 17% (OR, 0.83; 95%, CI 0.73–0.94), 19% (OR, 0.81; 95% CI, 0.70–0.94), 38% (OR, 0.62; 95% CI, 0.51–0.75), and 51% (OR, 0.49; 95% CI, 0.36–0.67) with overweight, class I obesity, class II obesity, and class III obesity or super obesity, respectively.
To confirm that the findings were not skewed by excluding cycles with missing BMI, a sensitivity analysis including these 1,234 cycles was performed with a “missing BMI” category. When the unadjusted and adjusted primary outcome models were refit with this additional BMI category as a covariate, the results were consistent with those of the primary analysis (Supplemental Table 1, available online).
Discussion
The results of this large national study demonstrate that BMI is negatively associated with the odds of a singleton, term delivery of a normal weight infant after fresh, autologous IVF among women with PCOS. Our data indicate that obesity is strongly associated with worse ART outcomes, even among a study population with a high baseline prevalence of obesity.
Our findings are consistent with those of a retrospective cohort study that used 2008-2010 SART CORS data to demonstrate a negative association between BMI and live birth rates among women with a diagnosis of ovulation disorder (
9
). Because the ovulation disorder diagnosis was not exclusive to PCOS, however, many women with other causes of ovulation dysfunction were inadvertently included in their analysis. Given the limitations of that study, we chose to use an AMH cutoff of > 4.45ng/mL to improve the likelihood of excluding women without PCOS (- Provost M.P.
- Acharya K.S.
- Acharya C.R.
- Yeh J.S.
- Steward R.G.
- Eaton J.L.
- et al.
Pregnancy outcomes decline with increasing body mass index: analysis of 239,127 fresh autologous in vitro fertilization cycles from the 2008-2010 society for assisted reproductive technology registry.
Fertil Steril. 2016; 105: 663-669
20
). We also excluded women with a secondary infertility diagnosis of diminished ovarian reserve or an ART indication consistent with hypothalamic amenorrhea. Therefore, the results of the present study have greater internal validity and are more applicable to the PCOS population.Other existing studies have been too small to categorize women into more than two or three BMI categories, and some were underpowered to detect a meaningful difference in live birth rates (
16
, 17
, 18
). Those that did compare live birth rates did not examine rates of multiple birth or associated neonatal morbidities such as prematurity and low birth weight.The present study is strengthened by its large sample size and the use of clinically relevant WHO BMI categories. It is also strengthened by the use of a composite variable, a good perinatal outcome, advocated as the most clinically relevant measure of IVF success (
25
, 26
, 27
). The use of live birth rate as the primary outcome underestimates the morbidity associated with multiple births, specifically prematurity and low birth weight (28
). Notably, the median number of embryos transferred in the present study population was two across all BMI categories despite national guidelines advocating single embryo transfer to maximize the likelihood of a healthy live birth (26
, 27
, 29
, 30
). Consequently, the multiple birth rate was as high as 36%, with up to one-third of women delivering prematurely. Even in the lowest BMI categories, only one-quarter of women experienced a good perinatal outcome. Our results accentuate the importance of performing single embryo transfer in good-prognosis patients, regardless of BMI.The present study also has limitations, most notably the challenge of identifying a population of women with PCOS from SART CORS. We recognize that AMH level is not a traditional diagnostic criterion for PCOS. However, no distinct category existed for PCOS in SART CORS at the time of data collection. Given that previous studies did not address this limitation at all, we feel that the use of an AMH cutoff significantly reduces the risk of misclassification bias compared with those of prior SART CORS studies in which a PCOS diagnosis was incorrectly assigned to all women with ovulatory dysfunction. Despite the use of an AMH cutoff and the exclusion of women with a clear diagnosis of hypothalamic amenorrhea, we acknowledge that there was no way to definitively exclude all women without PCOS from our cohort. We also acknowledge that some women with PCOS and an AMH level on the lower end of the spectrum may inadvertently have been excluded.
Additionally, despite the large cohort, there were a limited number of women with super obesity (BMI > 50). However, this reflects reality as many IVF centers have BMI limits in place to prevent dangerous anesthesia complications or the numerous obstetric risks associated with such morbid obesity (
3
, 31
). By combining the class III obesity and super obesity groups, we maintained the ability to show clinically meaningful differences in ART outcomes in women with morbid obesity. Notably, whereas approximately 10% of women were missing BMI data, the sensitivity analysis is reassuring that this missing data does not affect the results. However, this remains a potential source of bias if an unequal proportion of these women with unknown BMI came from a single BMI category.During our study period (2012-2015), the rate of missing AMH data was approximately 50%, likely based on the newness of this variable at that time. However, the proportion of missing AMH data was similar between women, with (49%) and without (50%) a diagnosis of “ovulatory disorder polycystic ovaries.” Moreover, when applying our inclusion/exclusion criteria to the cohort, aside from AMH > 4.45, the rate of a good perinatal outcome was similar between the group with AMH data (19.5%) compared with the group with missing AMH data (20.1%). Taken together, these findings support a lack of bias introduced by this missing AMH data.
Although SART CORS includes abundant IVF cycle data, it does not include obstetric complications that may affect the likelihood of our composite outcome via pathways unrelated to ART; for example, iatrogenic preterm delivery for preeclampsia or macrosomia in the setting of gestational diabetes. Regardless of the causal pathway, these complications are known risks for women with obesity and PCOS and should be accounted for when counseling patients about their chances of a healthy infant after IVF.
Finally, our dataset includes only fresh cycles and may not be generalizable to subsequent frozen embryo transfers (
32
). However, our results show that the women with obesity have fewer cryopreserved embryos for future use, and thus, their cumulative success per oocyte retrieval is likely similarly decreased.Conclusions
This large, national, retrospective study demonstrated that the odds of a good perinatal outcome decline with increasing BMI among women with PCOS. Women with PCOS and elevated BMI should be counseled that a decreased BMI may optimize their odds of achieving term delivery of a normal weight singleton—the ultimate goal of IVF.
Acknowledgments
The authors thank SART for the dataset, as well as all SART members, for providing clinical information to the SART CORS database for use by patients and researchers. Without the efforts of SART members, this research would not have been possible.
The Duke BERD Method Core’s support of this project was made possible (in part) by Grant Number UL1TR002553 from the National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health (NIH) and NIH Roadmap for Medical Research. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of NCATS or NIH.
Supplementary data
- Supplementary Material
References
- Prevalence of obesity among adults and youth: United States, 2015-2016.NCHS Data Brief, 2017
- Is obesity still increasing among pregnant women? Prepregnancy obesity trends in 20 states, 2003-2009.Prev Med. 2013; 56: 372-378
- Obesity, obstetric complications and cesarean delivery rate--a population-based screening study.Am J Obstet Gynecol. 2004; 190: 1091-1097
- Extreme obesity and risk of stillbirth among black and white gravidas.Obstet Gynecol. 2007; 110: 552-557
- Obesity and the risk of stillbirth: a population-based cohort study.Am J Obstet Gynecol. 2014; 210 (457.e1–9)
- Obesity is associated with increased risk of first trimester and recurrent miscarriage: matched case-control study.Hum Reprod. 2004; 19: 1644-1646
- Maternal obesity and risk of cesarean delivery: a meta-analysis.Obes Rev. 2007; 8: 385-394
- Effect of overweight and obesity on assisted reproductive technology--a systematic review.Hum Reprod Update. 2007; 13: 433-444
- Pregnancy outcomes decline with increasing body mass index: analysis of 239,127 fresh autologous in vitro fertilization cycles from the 2008-2010 society for assisted reproductive technology registry.Fertil Steril. 2016; 105: 663-669
- Andrology: polycystic ovary syndrome: the spectrum of the disorder in 1741 patients.Hum Reprod. 1995; 10: 2107-2111
- Referral bias in defining the phenotype and prevalence of obesity in polycystic ovary syndrome.J Clin Endocrinol Metab. 2013; 98: E1088-E1096
- Pregnancy outcomes in women with polycystic ovary syndrome undergoing in vitro fertilization.Fertil Steril. 2016; 105: 791-797.e2
- Adverse pregnancy outcomes with assisted reproductive technology in non-obese women with polycystic ovary syndrome: a case-control study.Clin Exp Reprod Med. 2011; 38: 103-108
- [Pregnancy outcomes of patients with polycystic ovary syndrome undergoing in vitro fertilization and embryo transfer].Zhonghua Fu Chan Ke Za Zhi. 2011; 46: 923-927
- A meta-analysis of pregnancy-related outcomes and complications in women with polycystic ovary syndrome undergoing IVF.Reprod Biomed Online. 2019; 39: 281-293
- Effect of body mass index on in vitro fertilization outcomes in women with polycystic ovary syndrome.Am J Obstet Gynecol. 2014; 211 (163.e1–6)
- Overweight and obesity adversely affect outcomes of assisted reproductive technologies in polycystic ovary syndrome patients.Int J Clin Exp Med. 2013; 6: 991-995
- Morbid obesity is associated with lower clinical pregnancy rates after in vitro fertilization in women with polycystic ovary syndrome.Fertil Steril. 2009; 92: 256-261
- Prevention CfDCa. 2005 Assisted reproductive technology success rates: National summary and fertility clinic reports.in: American Society for Reproductive Medicine. Society for Assisted Reproductive Technology, 2007
- Anti-mullerian hormone as a diagnostic and prognostic tool for PCOS patients.J Assist Reprod Genet. 2014; 31: 1311-1316
- Granulosa cell production of anti-Müllerian hormone is increased in polycystic ovaries.J Clin Endocrinol Metab. 2007; 92: 240-245
- Is the plasma anti-Müllerian hormone (AMH) level associated with body weight and metabolic, and hormonal disturbances in women with and without polycystic ovary syndrome?.Eur J Obstet Gynecol Reprod Biol. 2011; 158: 254-259
- Anti-Müllerian hormone in polycystic ovary syndrome and normo-ovulatory women: correlation with clinical, hormonal and ultrasonographic parameters.Middle East Fertil Soc J. 2010; 15: 253-258
- Obesity: preventing and managing the global epidemic: Report of a WHO Consultation. World Health Organization, 2000
- Trends and correlates of good perinatal outcomes in assisted reproductive technology.Obstet Gynecol. 2012; 120: 843-851
- Number of embryos transferred after in vitro fertilization and good perinatal outcome.Obstet Gynecol. 2014; 123: 239-247
- Elective single embryo transfer trends and predictors of a good perinatal outcome--United States, 1999 to 2010.Fertil Steril. 2013; 99: 1937-1943
- Assisted reproductive technology surveillance - United States, 2016.MMWR Surveill Summ. 2019; 68: 1-23
- Patient and cycle characteristics predicting high pregnancy rates with single-embryo transfer: an analysis of the Society for Assisted Reproductive Technology outcomes between 2004 and 2013.Fertil Steril. 2017; 108: 750-756
Elective single-embryo transfer. In: Practice Committee of the Society for Assisted Reproductive Technology. Vol. 97. Fertil Steril 2012:835–842.
- Survey assessing obesity policies for assisted reproductive technology in the United States.Fertil Steril. 2016; 105: 703-706.e2
- Fresh versus frozen embryos for infertility in the polycystic ovary syndrome.N Engl J Med. 2016; 375: 523-533
Article info
Publication history
Published online: December 01, 2022
Accepted:
November 29,
2022
Received in revised form:
November 23,
2022
Received:
June 25,
2022
Footnotes
J.S.H. has nothing to disclose. J.M.W. has nothing to disclose. T.T. has nothing to disclose. K.S.A. has nothing to disclose. J.L.E. consults for Hologic, Inc. and receives textbook royalties from Springer.
Identification
Copyright
© 2022 The Authors. Published by Elsevier Inc. on behalf of American Society for Reproductive Medicine.
User license
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) | How you can reuse 
Elsevier's open access license policy
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0)
Permitted
For non-commercial purposes:
- Read, print & download
- Redistribute or republish the final article
- Text & data mine
- Translate the article (private use only, not for distribution)
- Reuse portions or extracts from the article in other works
Not Permitted
- Sell or re-use for commercial purposes
- Distribute translations or adaptations of the article
Elsevier's open access license policy
