Methods
This metaanalysis was conducted according to the Metaanalysis of Observational Studies in Epidemiology guidelines.
Literature search
A literature search was done by the investigators in PubMed, MEDLINE, Embase, and CENTRAL from January 1989 to July 15, 2015. In addition, Google, Google Scholar, and references of selected articles were used to identify other studies. We used the following keywords: preeclampsia, pregnancy-induced hypertension, gestational hypertension, pregnancy complication, egg, oocyte, ovum, donation, and donor.
Eligibility criteria
We included comparative studies that described pregnancies that were achieved through oocyte donation with the subsequent generation of preeclampsia or gestational hypertension as an outcome and compared them with pregnancies that were achieved through other methods of ART or natural conception. Gestational hypertension is defined as a new-onset elevated blood pressure (mild, ≥140/90 mm Hg; severe, ≥160/110 mm Hg) after 20 weeks of gestation without proteinuria or end-organ failure. Before 2013, preeclampsia was diagnosed when gestational hypertension was accompanied by proteinuria (≥0.3 g/24 h). In 2013, the American College of Obstetricians and Gynecologists (ACOG) replaced proteinuria as a necessary criterion for preeclampsia diagnosis with signs and symptoms of end-organ injuries. The definitions of preeclampsia and gestational hypertension that were used for inclusion were based on the regional standards and guidelines in place at the time of each study.
Comorbidities (such as, gestational diabetes mellitus, HELLP (hemolysis, elevated liver enzymes, and low platelet count syndrome), morbid obesity, preterm labor, and multiple gestations) were not exclusion criteria. Abstracts, reviews, case studies, editorials, and noncomparative primary studies were excluded. The studies that had nonspecific “hypertensive disorders” as their outcome were also excluded. No language restrictions were applied.
Quality assessment
The Methodological Index for Non-Randomized Studies (MINORS) was used to assess the quality of nonrandomized studies. This framework consists of 12 items that evaluate a study’s validity, methods, and completeness of reporting elements. In the MINORS criteria, a comparative study is assigned a score of 0–2 for each of the 12 items included, for a maximum score of 24. Higher scores are indicative of greater methodologic quality.
Two investigators assessed each study independently and compared their scores afterwards to reach a consensus. If an agreement could not be reached, a third investigator was consulted.
Data extraction
The data from oocyte donation pregnancies, which lasted at least until week 20 of gestation, along with the control group, were extracted in a 2 × 2 contingency table. The data for nonoocyte donation ART (such as, in vitro fertilization, intracytoplasmic sperm injection, and insemination) were collected under the ART label. The data on spontaneous conception groups who did not use any type of assisted reproduction were collected separately under the natural conception label. Another investigator confirmed the extracted data independently. Disagreements were resolved by consulting a third investigator.
Data synthesis
Studies were classified into 4 groups based on their outcomes and control groups: (1) preeclampsia as the outcome and other methods of ART as the control, (2) preeclampsia as the outcome and natural conception as the control, (3) gestational hypertension as the outcome and other ART methods as the control, and (4) gestational hypertension as the outcome and natural conception as the control. It was possible for a study to be assigned to >1 group depending on whether they included both preeclampsia and gestational hypertension as the outcome or both ART and natural conception as the control.
Metaanalysis was performed with Review Manager software (version 5.3; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark). The Mantel-Haenszel model was used to analyze the dichotomous variables to produce an odds ratio (OR) for each outcome with a 95% confidence interval (CI). For each outcome, the heterogeneity of the study was assessed with the use of Chi 2 test and I 2 statistics. When no degree of heterogeneity was detected (I 2 = 0%), we used a fixed-effects model. When some degree of heterogeneity was present (I 2 > 0%), we used a random-effects model. Funnel plot analysis was used to assess publication bias by plotting ORs against standard errors.
Results
Study characteristics
The conducted search identified 523 studies for initial review, of which 19 were deemed to meet preidentified inclusion and exclusion criteria ( Figure 1 ). There were no randomized control trials found. From the 19 selected studies, 6 were case-control, and 13 were retrospective cohort studies. Thirteen studies had other methods of ART as their only comparison ; 3 studies had natural conception as their only comparison, and 3 studies included both comparison groups. In regards to outcomes, 5 studies included only preeclampsia ; 4 studies included only gestational hypertension, and 10 studies included both outcomes. Ten of the included studies originated from the United States ; 8 originated from Europe, and 1 originated from Israel. A total of 86,515 pregnant women were included and observed during their pregnancies. The characteristics of the patients in the studies are listed in Table 1 . The range of maternal age in the oocyte donation, other methods of ART, and natural conception groups were 33.5-46.2, 33-44, and 30.7-44.1, respectively. One study included only singleton pregnancies ; 2 studies that only included twin or multiple pregnancies, and 16 studies that did not have such restrictions. Of the 16 studies, 6 studies provided separate data for singleton and multiple subgroups.
| Study | Country | Design | Study period | Inclusion criteria | Exclusion criteria | Control group | Mean maternal age, yr | Parity |
|---|---|---|---|---|---|---|---|---|
| Cobo et al, 2014 | Spain | RC | 2007-2012 | Live birth or stillbirth ≥24 weeks of gestation; IVF with own oocyte or donated oocyte | Pregnancy loss at <24 weeks of gestation | IVF | OD, 41.2; ART, 35.7 | N/A |
| Fox et al, 2014 | United States | RCC | 2005-2012 | Twin, >20 weeks of gestation | Monochorionic monoamniotic placentation, previous diagnosis of hypertension | IVF | N/A | N/A |
| Gundogan et al, 2010 | United States | RCC | 2004-2006 | Placental deliveries with OD | <24 weeks of gestation | IVF | OD, 43; ART, 37.3 | N/A |
| Henne et al, 2007 | United States | RC | 1997-2002 | OD pregnancies | N/A | NC | OD, 42.3; NC, 36.8 | OD, 0.59; NC, 1.22 |
| Keegan et al, 2007 | United States | RCC | 1999-2003 | OD/IVF patients <35 or ≥40 years old | Triplet pregnancies, frozen embryo transfers, monitored at program satellite offices | IVF | OD, 42.6; ART, 35.1 | N/A |
| Klatsky et al, 2010 | United States | RC | 1998-2005 | OD pregnancies | N/A | IVF | OD, 40.2; ART, 39.8 | OD, 0.27; ART, 0.24 |
| Krieg et al, 2008 | United States | RC | 2001-2005 | OD pregnancies at >38 years old | N/A | IVF | OD, 42.7; ART, 41.3 | OD, 0.32; ART, 0.35 |
| Le Ray et al, 2012 | France | RC | 2008-2010 | Women who gave birth at >43 years old | N/A | IVF, NC | OD, 46.2; ART, 44.0; NC, 44.1 | OD, 0.3; ART, 0.9; NC, 1.4 |
| Levron et al, 2014 | Israel | RC | 2005-2011 | OD pregnancies beyond first trimester; singleton | Congenital anomalies and chromosomal abnormality | IVF | OD, 45; ART, 41 | Nulliparous: OD, 51%; ART, 44% |
| Malchau et al, 2013 | Denmark | RC | 1995-2010 | OD pregnancies resulted in birth | N/A | IVF, ICSI, NC | OD, 37.1;ART, 33.4; NC, 30.7 | OD, 1.3; ART, 1.39; NC, 1.85 |
| Porreco et al, 2005 | United States | RCC | 1998-2004 | OD pregnancies at >45 years old | N/A | NC | N/A | N/A |
| Salha et al, 1999 | UK | RCC | 1992-1997 | Pregnancies with gamete donation delivered at ≥24 weeks of gestation | N/A | Insemination, embryo donation, NC | OD, 38.1; ART, 31.9; NC, 37.2 | Primigravida: OD, 85%; ART, 89%; NC, 82% |
| Sekhon et al, 2014 | United States | RC | 2005-2013 | Twin, at >24 weeks of gestation | Monochorionic monoamniotic placentation | IVF | OD, 43; ART, 41.9 | Nulliparous: OD, 71.1%; ART, 73.1% |
| Söderstörm et al, 1998 | Finland | RC | 1991-1996 | OD pregnancies with birth at ≥24 weeks of gestation or ≥500-g newborn infant | N/A | IVF | OD, 33.5; ART, 33.4 | Primipara: OD, 84%; ART, 69% |
| Stoop et al, 2012 | Belgium | RC | 1999-2008 | OD pregnancies with birth at >20 weeks of gestation | Pregnancies after preimplantation genetic diagnosis, testicular sperm extraction, or use of donor sperm | IVF | OD, 36; ART, 36 | OD, 0.23; ART, 0.23 |
| Tranquilli et al, 2013 | Italy | RC | not specified | 26 cases of ICSI embryo transfer with OD | N/A | ICSI | OD, 42.7; ART, 37.5 | N/A |
| Van Dorp et al, 2014 | Netherland | RCC | 1992-2009 | OD pregnancies with birth at >24 weeks of gestation | Cycles without embryo transfer | IVF | OD, 36.4; ART, 36.7 | Nulliparous: OD, 78%; ART, 64% |
| Wiggins and Main, 2005 | United States | RCC | 1999-2004 | OD pregnancies | N/A | IVF | OD, 41.9; ART, 37.7 | Nulliparous: OD, 70%; ART, 74% |
| Wolff et al, 1997 | United States | RC | 1992-1995 | OD pregnancies | N/A | NC | OD, 41.5; NC, 42.7 | Nulliparous: OD, 70%; NC, 39% |
Risk of bias in included studies
All 19 studies were assessed for methodologic quality with the use of MINORS criteria ( Table 2 ). There was high concordance between the 2 reviewers; as a result, a third reviewer was involved in only a few instances. Criteria that received a low score in the majority of studies that were assessed included “prospective collection of data” (0% of studies included this), “unbiased assessment of the study endpoint” (5% of studies included this), and “prospective calculation of study size” (5% of studies included this). The remaining 9 criteria were reported by most studies with various degrees of adequacy. Overall, the total MINORS scores of the studies were similar, ranging from 14–20, with a median score of 17.
| Study | Methodological Index for Non-Randomized Studies score |
|---|---|
| Cobo et al, 2014 | 18 |
| Fox et al, 2014 | 17 |
| Gundogan et al, 2010 | 14 |
| Henne et al, 2007 | 18 |
| Keegan et al, 2007 | 15 |
| Klatsky et al, 2010 | 20 |
| Krieg et al, 2008 | 15 |
| Le Ray et al, 2012 | 18 |
| Levron et al, 2014 | 17 |
| Malchau et al, 2013 | 16 |
| Porreco et al, 2005 | 17 |
| Salha et al, 1999 | 17 |
| Sekhon et al, 2014 | 20 |
| Söderstörm et al, 1998 | 18 |
| Stoop et al, 2012 | 19 |
| Tranquilli et al, 2013 | 15 |
| Van Dorp et al, 2014 | 17 |
| Wiggins and Main, 2005 | 18 |
| Wolff et al, 1997 | 16 |
To identify evidence of publication bias, we generated funnel plots of the studies that used other methods of ART as the comparison group ( Figure 2 ). All included studies fell within the 95% confidence interval lines. Both graphs look symmetric, which indicates no publication bias. Funnel plots for studies with natural conception as comparison group were not generated because of the low number of studies.
Outcome analysis: preeclampsia
There were 15 studies that reported preeclampsia as their outcomes in comparison with in vitro fertilization or intracytoplasmic sperm injection. After pooling the data for metaanalysis, we found that oocyte donation significantly increases the risk of preeclampsia compared with the other methods of ART comparison group (OR, 2.54; 95% CI, 1.98–3.24; P < .0001; Figure 3 , A). Analysis of the 5 studies that included natural conception as their comparison group also found an increased risk of development of preeclampsia in the oocyte donation patients (OR, 4.34; 95% CI, 3.10–6.06; P < .0001; Figure 3 , B). A subgroup analysis was performed to examine the effects of singleton vs multiple gestations in the oocyte transfer pregnancies with the use of other methods of ART as the comparison. The risk of the development of preeclampsia after oocyte donation was higher in both singleton (OR, 2.24; 95% CI, 1.42–3.53; P = .0005; Figure 4 , A) and multiple (OR, 2.56; 95% CI, 1.84–3.58; P < .0001; Figure 4 , B) gestation groups. A sensitivity analysis was done for studies with ART as a comparison, which scored >18; our results were robust (OR, 2.75; 95% CI, 1.93–3.90; P < .0001).
Outcome analysis: gestational hypertension
In 13 studies, ART was the comparison group, and gestational hypertension was the outcome. The metaanalysis indicated that oocyte donation pregnancies are at higher risk of gestational hypertension compared with other methods of ART pregnancies (OR, 3.00; 95% CI, 2.44–3.70; P < .0001; Figure 5 , A). Only 2 studies with gestational hypertension as an outcome had a comparison group that consisted of women with natural conception pregnancies. However, the risk of gestational hypertension was also shown to be higher in the oocyte-donation pregnancies compared with the natural conception group (OR, 7.94; 95% CI, 1.73–36.36; P = .008; Figure 5 , B). A subgroup analysis of singleton and multiple pregnancies with other methods of ART as the comparison group was conducted. The risk of the development of gestational hypertension was higher in both singleton (OR, 2.86; 95% CI, 2.10–3.90; P < .0001; Figure 6 , A) and multiple (OR, 3.08; 95% CI, 1.95–4.87; P < .0001; Figure 6 , B) gestation groups. A sensitivity analysis was done for studies with ART as comparison, which scored >18; our results were robust (OR, 1.93; 95% CI, 2.35-4.33; P < .0001).
Results
Study characteristics
The conducted search identified 523 studies for initial review, of which 19 were deemed to meet preidentified inclusion and exclusion criteria ( Figure 1 ). There were no randomized control trials found. From the 19 selected studies, 6 were case-control, and 13 were retrospective cohort studies. Thirteen studies had other methods of ART as their only comparison ; 3 studies had natural conception as their only comparison, and 3 studies included both comparison groups. In regards to outcomes, 5 studies included only preeclampsia ; 4 studies included only gestational hypertension, and 10 studies included both outcomes. Ten of the included studies originated from the United States ; 8 originated from Europe, and 1 originated from Israel. A total of 86,515 pregnant women were included and observed during their pregnancies. The characteristics of the patients in the studies are listed in Table 1 . The range of maternal age in the oocyte donation, other methods of ART, and natural conception groups were 33.5-46.2, 33-44, and 30.7-44.1, respectively. One study included only singleton pregnancies ; 2 studies that only included twin or multiple pregnancies, and 16 studies that did not have such restrictions. Of the 16 studies, 6 studies provided separate data for singleton and multiple subgroups.
