Table 3.
Association between donor and recipient body mass index (BMI) and length of gestation and birthweight among donor oocyte recipient singleton live births.
| No. of live births | Length of gestation |
Birth weight |
|||||
|---|---|---|---|---|---|---|---|
| Mean, wk/% <37 wk | Adjusted HR (95% CI)a | Adjusted OR of preterm (95% CI)b | Mean, g/% <2,500 g | Adjusted β (95% CI)c | Adjusted RR of low birth weight (95% CI)d | ||
| Donor BMI, per 2 kg/m2 | 0.99 (0.92–1.06) | 1.05 (0.90–1.22) | 22.8 (−26.3 to 72.0) | 0.98 (0.85–1.14) | |||
| Donor BMI | |||||||
| ≤21 | 203 | 38.3/16.3 | 1.05 (0.88–1.26) | 0.84 (0.55–1.28) | 3197.7/11.3 | −24.0 (−153.6 to 105.5) | 0.96 (0.65–1.41) |
| 21.1–24.9 | 384 | 38.4/15.6 | 1.00 (ref.) | 1.00 (ref.) | 3279.0/8.9 | 0 (ref.) | 1.00 (ref) |
| ≥25 | 78 | 38.4/11.5 | 0.99 (0.76–1.30) | 0.79 (0.42–1.48) | 3394.1/9.0 | 60.7 (−131.9 to 253.4) | 0.94 (0.55–1.60) |
| p-trend4 | 0.61 | 0.91 | 0.44 | 0.99 | |||
| Recipient BMI, per 2 kg/m2 | 1.04 (0.96–1.12) | 1.04 (0.96–1.12) | −12.0 (−34.8 to 10.8) | 1.04 (0.99–1.11) | |||
| Recipient BMI | |||||||
| <18.5 | 11 | 39.5/9.1 | 0.64 (0.43–0.95) | 0.90 (0.28–2.92) | 3395.6/0.0 | 137.2 (−234.6 to 509.0) | 1.14 (0.41–3.16) |
| 18.5–24.9 | 409 | 38.4/15.7 | 1.00 (ref.) | 1.00 (ref.) | 3257.0/9.5 | 0 (ref.) | 1.00 (ref) |
| 25–29.9 | 158 | 38.3/12.0 | 1.13 (0.90–1.41) | 0.79 (0.52–1.19) | 3309.3/10.8 | 11.4 (−136.5–159.2) | 0.98 (0.64–1.48) |
| 30–34.9 | 63 | 38.0/20.3 | 1.01 (0.75–1.35) | 1.02 (0.57–1.83) | 3173.5/11.1 | −89.0 (−302.2 to 124.1) | 1.11 (0.63–1.94) |
| ≥35 | 25 | 38.0/20.8 | 1.43 (0.94–2.22) | 1.93 (0.78–4.75) | 3373.2/4.0 | −138.2 (−409.3 to 133.0) | 1.76 (1.02–3.04) |
| Ptrend | .13 | .37 | .23 | .17 | |||
Note: All models for donor BMI were adjusted for donor’s age and race, recipient’s BMI and age, uterine-factor infertility, polycystic ovary syndrome (PCOS), and retrieval year. All models for recipient BMI were adjusted for donor’s age and BMI, recipient’s age and race, uterine-factor infertility, PCOS, and retrieval year. CI = confidence interval; HR = hazard ratio; OR = odds ratio.
Analyses for gestational length were conducted using cluster-weighted Cox proportional hazard and a robust sandwich covariance estimate to account for the multiple live births per woman in the presence of nonignorable cluster size. Each observation was weighted inversely to the number of live births they contributed to the analysis.
Analyses for preterm birth and low birth weight were conducted using cluster-weighted generalized estimating equations with binomial distribution and logit link function to account for within-person correlations in the presence of nonignorable cluster size. Each observation was weighted inversely to the number of live births they contributed to the analysis.
Analyses for birth weight were conducted using cluster-weighted generalized estimating equations with normal distribution and identity link function to account for within-person correlations in the presence of nonignorable cluster size. Each observation was weighted inversely to the number of live births they contributed to the analysis.
Ptrend was calculated using the median values of each category of BMI as continuous variables.