Research
www.AJOG .org
OBSTETRICS
The association of body mass index with serum angiogenic markers in normal and abnormal pregnancies Chloe A. Zera, MD, MPH; Ellen W. Seely, MD; Louise E. Wilkins-Haug, MD, PhD; Kee-Hak Lim, MD; Samuel I. Parry, MD; Thomas F. McElrath, MD, PhD OBJECTIVE: Because obesity is a risk factor for placental dysfunction, we hypothesized that maternal body mass index (BMI) would be associated with alterations in serum angiogenic markers. STUDY DESIGN: We included 2399 singleton pregnancies with and
without placental dysfunction in a prospective longitudinal cohort study of angiogenic markers. We modeled the relationship between categorical and continuous BMI, soluble fms-like tyrosine kinase-1 (sFlt-1), and placental growth factor (PlGF) over gestation, stratified by pregnancy outcome. RESULTS: In women with normal pregnancies, a higher BMI was associated with lower sFlt-1 values across gestation (P < .0001),
lower PlGF in the second and third trimesters (P < .0001), and lower rate of change in PlGF (P < .0001). Similar relationships were seen between maternal BMI, sFlt-1 (P < .0001), and PlGF (P ¼ .0005) in women with clinically evident placental dysfunction. CONCLUSION: The sFlt-1 value is inversely associated with maternal
BMI. The pattern of change in PlGF is also dependent on maternal BMI, indicating that obese women may have abnormalities in angiogenesis near term. Key words: angiogenic markers, obesity, placenta, preeclampsia, soluble fms-like tyrosine kinase-1
Cite this article as: Zera CA, Seely EW, Wilkins-Haug LE, et al. The association of body mass index with serum angiogenic markers in normal and abnormal pregnancies. Am J Obstet Gynecol 2014;210:xx-xx.
N
early 25% of women have a body mass index (BMI) in the obese range (30 kg/m2) prior to pregnancy.1 Maternal obesity is associated with abnormal placental function clinically apparent as preeclampsia,2,3 clinically unrecognized intrauterine growth restriction,4,5 placental abruption, and potentially stillbirth.6,7 The mechanisms underlying the association between maternal obesity and ischemic placental disease are not well understood; however, there is a large body of literature on
nonpregnant individuals to support that obesity, like preeclampsia, is a state of systemic inflammation characterized by endothelial dysfunction and abnormalities in angiogenesis.8-10 When taken together, these associations suggest the possibility that maternal obesity is associated with endothelial damage leading to abnormal placental function. An evolving literature has linked the development of placental dysfunction with the expression of abnormal quantities of placental angiogenic proteins. In
From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology (Drs Zera, Wilkins-Haug, and McElrath), and Division of Endocrinology, Diabetes, and Hypertension, Department of Medicine (Dr Seely), Brigham and Women’s Hospital, Harvard Medical School (Drs Zera, Seely, Wilkins-Haug, Lim, and McElrath), and Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Beth IsraeleDeaconess Medical Center (Dr Lim), Boston, MA, and Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Pennsylvania Medical Center, Philadelphia, PA (Dr Parry). Received Dec. 18, 2013; revised March 6, 2014; accepted March 10, 2014. This study was supported by Abbott Diagnostics Research grant 9MZ-04-06N03. The authors report no conflict of interest. Presented, in part, as a poster at the 30th annual meeting of the Society for Maternal-Fetal Medicine, Chicago, IL, Feb. 1-6, 2010. Reprints: Chloe Zera, MD, MPH, Division of Maternal-Fetal Medicine, Brigham and Women’s Hospital, 75 Francis St., Boston, MA 02115.
[email protected]. 0002-9378/$36.00 ª 2014 Mosby, Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2014.03.020
animal models, abnormal uteroplacental blood flow is associated with an increase in the antiangiogenic protein, soluble fms-like tyrosine kinase-1 (sFlt-1), which binds and antagonizes both placental growth factor (PlGF) and vascular endothelial growth factor.11 Although the analyses were limited by sample size, observational data from the Calcium for Preeclampsia Prevention trial suggested that the ratio of sFlt-1 to PlGF increases several weeks before the onset of clinical disease in women destined to develop preeclampsia.12 In vitro, the adipokine visfatin alters the expression of vascular endothelial growth factor, providing a potential link between obesity and altered angiogenesis at the maternal-fetal interface.13 In vivo data connecting maternal obesity with altered placental angiogenesis are limited, however. Although some have attempted to examine the associations between maternal obesity and alterations of serum angiogenic markers in preeclamptic pregnancies, their conclusions are limited by small sample size, homogenous population, and a narrow range of maternal BMI.14,15 Furthermore, factors affecting angiogenic
MONTH 2014 American Journal of Obstetrics & Gynecology
1.e1
Research
www.AJOG.org
Obstetrics
TABLE 1
Participant characteristics, by BMI category First trimester BMI, kg/m2 Characteristic
<25 (n [ 1292)
Age, mean (SD)
31.3 (5.4)
25 to <30 (n [ 584) 30.9 (6.2)
‡30 (n [ 523) 30.4 (5.7)
Total (n [ 2399)
P value
31.0 (5.7)
.005 < .0001
Race/ethnicity Caucasian, n, %
865 (67.0%)
319 (54.6%)
220 (42.1%)
1404 (58.5%)
African American, n, %
172 (13.3%)
146 (25.0%)
214 (40.9%)
532 (22.2%)
98 (7.6%)
61 (10.5%)
69 (13.2%)
228 (9.5%)
110 (8.5%)
31 (5.3%)
6 (1.2%)
147 (6.1%)
47 (3.6%)
27 (4.6%)
14 (2.7%)
88 (3.7%)
22.0 (1.9)
27.1 (1.4)
35.9 (5.7)
26.3 (6.3)
< .0001
424 (32.8%)
156 (26.7%)
105 (20.1%)
685 (28.6%)
< .0001
Current smoker, n, %
31 (2.4%)
17 (2.9%)
28 (5.4%)
76 (3.2%)
.005
History of diabetes, n, %
15 (1.2%)
7 (1.2%)
25 (4.8%)
47 (2.0%)
< .0001
History of hypertension, n, %
24 (1.9%)
24 (4.1%)
52 (9.9%)
100 (4.2%)
< .0001
Gestational diabetes, n, %
35 (2.7%)
20 (3.4%)
51 (9.8%)
106 (4.4%)
< .0001
GA at delivery, mean (SD)
39.0 (1.7)
38.9 (1.8)
38.6 (2.1)
38.9 (1.8)
.004
Hispanic, n, % Asian, n, % Other or missing, n, % 2
BMI at baseline, kg/m , mean (SD) Nulliparous, n, %
Birthweight, g, mean (SD) Placental ischemic disease, n, %a SGA, n, % Preeclampsia, n, % b
Abruption, n, %
3278 (520)
3311 (532)
3301 (635)
3292 (550)
.12
137 (10.6%)
77 (13.2%)
107 (20.5%)
321 (13.4%)
< .0001
89 (6.9%)
34 (5.8%)
37 (7.1%)
161 (6.7%)
.63
51 (4.0%)
45 (7.7%)
86 (16.4%)
182 (7.6%)
< .0001
7 (0.5%)
6 (1.0%)
2 (0.4%)
15 (0.6%)
.34
BMI, body mass index; GA, gestational age; SGA, small for gestational age. a
Patient may have more than 1 of SGA, preeclampsia, and abruption; b Compared with Fisher exact test.
Zera. BMI and angiogenic markers. Am J Obstet Gynecol 2014.
markers in normal pregnancies are not well understood. We hypothesize that maternal obesity is associated with abnormalities in placental angiogenesis. We therefore investigated the effect of maternal BMI on serum angiogenic markers in normal pregnancies and those affected by clinical placental dysfunction.
M ATERIALS
AND
M ETHODS
Study population This study is a secondary analysis of data from a prospective longitudinal cohort study designed to evaluate the utility of sFlt-1 and PlGF as markers for the diagnosis of preeclampsia.16 Women were recruited from 3 urban academic medical centers in Boston, MA, and Philadelphia, PA, from 2006 to 2008. Participants were
eligible for enrollment if the estimated gestational age was 15 weeks or less, they were at least 18 years old, and they were able to provide informed consent and had no more than 3 fetuses. For this analysis, we excluded women missing firsttrimester BMI (n ¼ 41) as well as those who did not have serum samples at 3 or more study visits (n ¼ 53). Of the remaining 2544 women, we excluded 143 with multiple gestations and 2 with stillbirths, leaving 2399 participants in the final analysis. We considered women with preeclampsia, placental abruption, or delivery of a small-for-gestational-age infant to have clinical placental dysfunction.17 Preeclampsia was defined as systolic blood pressure elevation of at least 140 mm Hg or diastolic blood
1.e2 American Journal of Obstetrics & Gynecology MONTH 2014
pressure of at least 90 mm Hg after 20 weeks’ gestation, in association with proteinuria, either spot urine protein/ creatinine ratio of greater than 0.20 or at least 300 mg or 24 hours16,18; each case was reviewed by a consensus committee of site primary investigators (K.-H.L., S.I.P., or T.F.M.). The diagnosis of placental abruption was based on clinician diagnosis at the time of delivery and was abstracted from the medical record by trained reviewers. We defined small for gestational age as a birthweight less than the 10th percentile (z-score less than e1.28) by published sex-specific growth curves.
Exposure assessment Maternal BMI was calculated using weight and height measured in the first
www.AJOG.org
Obstetrics
Research
TABLE 2
Angiogenic markers over time, by BMI category and presence of placental disease First-trimester BMI, kg/m2 Variable
<25
25 to <30
‡30
Total
Normal pregnancies, n (%)
1155 (55.6)
507 (24.4)
416 (20.0)
2078
P valuea
sFlt, pg/mL, median (IQR) <15 wks
5.4 (3.7e7.7)
5.0 (3.5e7.2)
4.1 (2.6e5.9)
5.0 (3.4e7.3)
< .0001
16-20 wks
6.5 (4.4e9.8)
6.0 (4.0e8.7)
5.2 (3.5e7.3)
6.0 (4.0e7.0)
< .0001
24-28 wks
6.2 (4.1e9.8)
5.6 (3.8e8.1)
4.6 (3.2e7.4)
5.8 (3.8e8.7)
< .0001
34-38 wks
10.4 (7.0e15.6)
9.3 (6.7e13.8)
8.4 (5.9e11.9)
9.6 (6.6e14.3)
< .0001
PlGF, ng/mL, median (IQR) <15 weeks
20.3 (14.5e31.4)
21.9 (14.8e32.3)
20.7 (14.9e31.4)
20.7 (14.6e31.7)
.28
16-20 wks
146.6 (103.4e197.5)
127.1 (93.4e184.1)
122.9 (88.8e169.7)
136.3 (97.2e191.4)
< .0001
24-28 wks
495.8 (330.5e713.0)
431.2 (279.8e630.0)
378.8 (258.8e552.6)
452.9 (304.4e663.4)
< .0001
34-38 wks
410.6 (177.1e785.4)
355.7 (175.0e658.9)
350.9 (187.8e678.4)
377.7 (177.6e740.5)
.16
107 (33.3)
321
Pregnancies with ischemic placental disease, n (%)
137 (42.7)
77 (24.0)
sFlt in pg/mL, median (IQR) <15 wks
4.8 (3.2e7.4)
4.7 (3.6e7.6)
3.7 (2.4e5.9)
4.5 (3.0e6.9)
.0009
16-20 wks
6.7 (4.8e10.2)
6.7 (4.6e9.3)
4.7 (3.3e8.5)
6.0 (4.3e9.3)
.01
24-28 wks
7.8 (4. 9e12.8)
5.9 (4.2e10.5)
5.3 (3.5e9.6)
6.3 (4.2e11.1)
.005
34-38 wks
15.6 (8.5e31.9)
14.1 (9.1e29.0)
14.6 (9.1e25.6)
.42
13.1 (9.3e21.0)
PlGF, ng/mL, median (IQR) <15 wks
20.8 (14.5e31.4)
22.0 (16.0e40.3)
20.4 (15.7e31.1)
20.9 (15.2e32.4)
.30
16-20 wks
130.1 (95.7e178.1)
139.1 (97.9e204.0)
107.3 (70.3e151.8)
128.9 (87.1e177.5)
.01
24-28 wks
382.3 (247.8e588.9)
341.4 (224. 8e547.9)
318.2 (173.9e472.3)
349.0 (215.3e527.3)
.02
34-38 wks
174.4 (78.7e350.1)
169.0 (108.8e434.3)
181.5 (78.4e371.8)
173.7 (81.6e378.8)
.91
BMI, body mass index; IQR, interquartile range; PlGF, placental growth factor; sFlt-1, soluble fms-like tyrosine kinase-1. a
P value by analysis of the variance across BMI category (cross-sectional at each point in gestation).
Zera. BMI and angiogenic markers. Am J Obstet Gynecol 2014.
trimester. We defined normal weight as a BMI less than 25 kg/m2, overweight as a BMI 25 to less than 30 kg/m2, and obesity as a BMI of 30 kg/m2 or greater.
Outcome assessment Serum samples were collected at 4 study visits (<15 weeks, 16-20 weeks, 24-28 weeks, and 34-38 weeks) and stored at e80 C. sFlt-1 and PlGF were assayed at each time point using the Architect immunoassay (Abbott Diagnostics, Abbott Park, IL).16 All included participants had at least 3 specimens. We analyzed the data in both a cross-sectional and
longitudinal fashion using all available specimens at each gestational age interval.
Covariates Demographic (including maternal age, race, and ethnicity) and historical information (including pregnancy history, smoking status, use of assisted reproduction, a history of diabetes, and a history of hypertension) was collected by a participant questionnaire at the study enrollment. Medical records were abstracted by trained reviewers to obtain relevant pregnancy outcomes.
Statistical analysis We compared baseline sociodemographics and prevalence of selected pregnancy characteristics by BMI category using analysis of variance (ANOVA) for continuous variables, and Mantel-Hanzel c2 tests for categorical variables. Statistical testing of angiogenic markers was performed after logarithmic transformation. For cross-sectional analyses, we compared sFlt-1 and PlGF by BMI category at each time point using ANOVA in women with and without ischemic placental disease. The pattern of change in sFlt-1 and PlGF was each modeled as a
MONTH 2014 American Journal of Obstetrics & Gynecology
1.e3
Research
www.AJOG.org
Obstetrics R ESULTS
FIGURE 1
Baseline characteristics We included 2399 participants in our analysis. The majority (n ¼ 1292, 53.9%) were normal weight, whereas 584 (24.3%) were overweight, and 523 (21.8%) were obese. Overweight and obese women were more likely to be African American or Hispanic than normal-weight women (Table 1). Overweight and obese women were more likely than normal-weight women to have a diagnosis of chronic hypertension, preexisting diabetes, or gestational diabetes (all P < .0001). Of the participants, 321 had clinically evident placental dysfunction including 182 women with preeclampsia. Fewer normal-weight women (n ¼ 137, 10.6%) developed clinical placental disease than overweight (n ¼ 77, 13.2%) or obese (n ¼ 107, 20.5%) women (P < .0001) (Table 1).
sFlt-1 over time
The median sFlt-1 values at each time point in gestation by BMI category and presence of clinical placental dysfunction are shown. BMI, body mass index; sFlt-1, soluble fms-like tyrosine kinase-1. Zera. BMI and angiogenic markers. Am J Obstet Gynecol 2014.
function of baseline BMI as a continuous variable using mixed linear models to adjust for significant confounders including maternal age, race, smoking, parity, gestational or pregestational diabetes, and chronic hypertension in women with and without placental dysfunction. All
analyses were performed with SAS version 9.2 (SAS Institute Inc, Cary, NC). The study was approved by the institutional review boards at Beth Israel Deaconess Medical Center, Brigham and Women’s Hospital, and the Hospital of the University of Pennsylvania.
Normal pregnancies Among women with normal pregnancy outcomes, the maternal BMI category was inversely associated with sFlt-1 at each time point in pregnancy (P < .0001). When compared with values in normal-weight women, the median sFlt-1 concentration was approximately 10% lower in overweight women and 20% lower in obese women (Table 2). The concentration of sFlt-1 increased over gestation for all participants, and although baseline values differed by BMI category, the pattern of change was similar,
TABLE 3
Linear longitudinal regression: angiogenic markers over time in normal pregnancies Angiogenic marker sFlt-1
PlGF
Adjusteda
Unadjusted Predictor BMI, kg/m
2
Parameter estimate (95% CI)
P value
Parameter estimate (95% CI)
P value
e0.02 (e0.03 to e0.02)
< .0001
e0.02 (e0.03 to e0.02)
< .0001
Time
0.13 (0.08e0.18)
< .0001
BMI*time
0.0006 (e0.001 to 0.002)
.52
BMI, kg/m2
0.003 (e0.003 to 0.008)
.32
Time
1.29 (1.23e1.35)
BMI*time
e0.006 (e0.008 to e0.004)
< .0001 < .0001
BMI, body mass index; CI, confidence interval; PlGF, placental growth factor; sFlt-1, soluble fms-like tyrosine kinase-1. a
Model adjusted for age, race, parity, smoking, diabetes, and chronic hypertension.
Zera. BMI and angiogenic markers. Am J Obstet Gynecol 2014.
1.e4 American Journal of Obstetrics & Gynecology MONTH 2014
0.13 (0.08e0.18) 0.0006 (e0.001 to 0.002) e0.004 (e0.01 to 0.001) 1.29 (1.23e1.35) e0.006 (e0.008 to e0.004)
< .0001 .54 .12 < .0001 < .0001
www.AJOG.org
Obstetrics
Research
lower in obese women than in overweight or normal-weight women (P ¼.01 at 16-20 weeks and P ¼ .02 at 24-28 weeks; Table 2). Although the concentration of PlGF dropped at term in all women with ischemic placental disease, BMI was associated with the rate of change in PlGF (unadjusted P ¼ .02; Table 4). Similar to the women with normal pregnancy outcomes, the adjustment for multiple covariates did not modify the association between BMI and the rate of change in PlGF (P ¼ .005; Table 4).
FIGURE 2
PlGF over time
C OMMENT
The median PlGF values at each time point in gestation by BMI category and presence of clinical placental dysfunction are shown. BMI, body mass index; PlGF, placental growth factor. Zera. BMI and angiogenic markers. Am J Obstet Gynecol 2014.
regardless of BMI category (Figure 1). In a linear longitudinal regression accounting for change over time, each unit BMI increase was associated with a 2% decrease in the geometric mean sFlt-1 value (P < .0001), and adjustment for covariates including age, race, parity, and comorbidities did not modify this relationship (Table 3). The median concentration of PlGF was similar across BMI categories at baseline but was lower in obese than in overweight or normal-weight women in the second and third trimesters (Table 2). PlGF increased markedly from the first to the second trimester in all participants; however, although PlGF dropped near term in normal-weight women, this change was attenuated in overweight and obese women (Figure 2). In linear longitudinal regression, maternal BMI was not associated with the PlGF value but was associated with the pattern of change over time (P < .0001, Table 3). Adjustment for covariates did not change the association
between BMI and the rate of change in PlGF (Table 3).
Pregnancies complicated by clinical placental dysfunction Maternal BMI was also associated with sFlt-1 in women with placental dysfunction. Similar to women with normal pregnancies, at each time point in pregnancy, the median sFlt-1 concentration was lowest in obese women compared with overweight or normal-weight women (Table 2). The concentration of sFlt-1 increased across gestation in all women, regardless of BMI. In linear longitudinal regression analyses, BMI was associated with the sFlt-1 value (P < .0001; Table 4) but not with the pattern of change over time; the adjustment for covariates slightly attentuated but did not negate the association between BMI and sFlt-1 (P ¼.002; Table 4). In women with placental dysfunction, BMI was not associated with baseline PlGF concentration; however, second and early third trimester values were
In this study, we found differences in angiogenic markers by maternal BMI category. Contrary to our hypothesis that obesity would be associated with an antiangiogenic profile categorized by relatively high levels of sFlt-1 and low levels of PlGF, we found that sFlt-1 concentration was inversely associated with maternal BMI across gestation. PlGF was also lower in obese women than in normal-weight women after the first trimester, with differential patterns of change near term. The associations between maternal BMI, sFlt-1, and PlGF were similar in pregnancies affected by placental dysfunction to those observed in women with clinically normal pregnancies. Although maternal BMI is a wellrecognized risk factor for preeclampsia, few authors have looked at the effect of BMI on serum angiogenic markers longitudinally over the course of gestation, particularly as a primary exposure. In their study of angiogenic markers and preeclampsia, Levine et al12 noted an inverse relationship between maternal BMI and sFlt-1 in the third trimester but concluded that maternal obesity did not affect the angiogenic profile. Others have controlled for the impact of maternal BMI in preeclampsia19 but did not specifically evaluate associations between BMI and serum angiogenic markers in a large sample of normal women. One study of normotensive women demonstrated an association between maternal BMI and angiogenic profile but noted a positive correlation rather than the previously seen negative association
MONTH 2014 American Journal of Obstetrics & Gynecology
1.e5
Research
www.AJOG.org
Obstetrics
TABLE 4
Linear longitudinal regression: angiogenic markers over time in pregnancies complicated by placental dysfunction Angiogenic marker sFlt-1
Predictor BMI, kg/m
Parameter estimate (95% CI) 2
BMI*time BMI, kg/m Time BMI*time
0.28 (0.14e0.42) e0.02 (e0.03 to e0.008)
Time PlGF
Adjusteda
Unadjusted
2
P value
Parameter estimate (95% CI)
< .0001 .002
0.23 (0.09e0.38) e0.02 (e0.04 to e0.009)
P value .002 .002
0.002 (e0.003 to 0.006)
.51
0.002 (e0.003 to 0.007)
.39
0.004 (e0.007 to 0.014)
.47
0.002 (e0.01 to 0.01)
.78
1.08 (0.93e1.23) e0.006 (e0.01 to e0.001)
< .0001 .02
1.14 (0.99e1.29) e0.007 (e0.01 to e0.002)
< .0001 .005
BMI, body mass index; CI, confidence interval; PlGF, placental growth factor; sFlt-1, soluble fms-like tyrosine kinase-1. a
Model adjusted for age, race, parity, smoking, diabetes, and chronic hypertension.
Zera. BMI and angiogenic markers. Am J Obstet Gynecol 2014.
between BMI and sFlt-1.20 By performing a stratified analysis, we were able to demonstrate a relationship between maternal BMI and levels of angiogenic markers independently of future preeclampsia risk. Higher maternal BMI was associated with lower sFlt-1 and PlGF levels across gestation in women with and without placental dysfunction. Because the pattern of change in sFlt-1 over gestation did not differ by BMI in both women with and without preeclampsia, it is possible that lower levels of sFlt-1 may in fact represent differences in volume of distribution rather than differences in placental expression of sFlt-1 over time. However, relatively low sFlt-1 concentrations in the first trimester have been associated with preeclampsia, growth restriction, and stillbirth19 and may in fact reflect a predisposition to placental pathology in obese women, regardless of pregnancy outcome. Interestingly, although levels of PlGF were also lower in obese and overweight women than in normal-weight women, the pattern of change differed by BMI category at term. Unlike sFlt-1, differences in the rate of change in PlGF cannot be explained by the increased volume of distribution in obese women, suggesting the possibility that expression of placental angiogenic proteins is stimulated at term in obese but not normalweight women. Potential mechanisms underlying the association between maternal obesity
and rapid increases in PlGF near term include maternal microvascular disease affecting uteroplacental blood flow, hypoadiponectinemia, and systemic inflammation with concurrent endothelial damage21 or increased placental oxidative and nitrative stress.22 An alternative hypothesis is that excessive fetal growth resulting from maternal insulin resistance induces demand for expansion of the placental vascular bed not seen in normal-weight women, an idea supported by previously observed positive associations between PlGF measured at delivery with infant birthweight.23 Further study is needed to understand the relationships between placental angiogenesis, maternal insulin resistance, and fetal growth. Our findings must be interpreted in the context of observational study design. Although our results provide epidemiological evidence of altered values of angiogenic markers in obese women both with and without placental dysfunction, we are unable to assess the causal pathways underlying the observed differences over the course of gestation. The strengths of our analysis include directly measured BMI early in pregnancy and validated clinical outcome information; however, we lacked measures of visceral adiposity such as prepregnancy waist circumference, fasting glucose tolerance, and inflammatory markers, which may have provided more insight into potential mechanisms underlying our findings. Furthermore, as a
1.e6 American Journal of Obstetrics & Gynecology MONTH 2014
secondary analysis of a study that excluded women with stillbirths, we were unable to evaluate the association of maternal BMI with an important outcome of interest. Finally, although only 107 women with placental dysfunction were obese, our analyses were not limited by sample size. Despite these limitations, our study extends the literature beyond previously demonstrated associations in smaller cross-sectional data sets and highlights the need to consider the impact of maternal BMI in studies utilizing angiogenic markers as a primary exposure or outcome. In our study population, maternal BMI was linked to differences in measured serum angiogenic markers across gestation. These findings were seen in parallel in women with and without clinically evident placental dysfunction, suggesting that differences in angiogenesis in obese women may not fully explain their increased risk of preeclampsia, placental abruption, and abnormal fetal growth. Future research is needed to fully elucidate the pathways through which maternal obesity may lead to poor obstetric outcomes. REFERENCES 1. Chu SY, Kim SY, Bish CL. Prepregnancy obesity prevalence in the United States, 2004-2005. Matern Child Health J 2009;13: 614-20. 2. Cedergren MI. Maternal morbid obesity and the risk of adverse pregnancy outcome. Obstet Gynecol 2004;103:219-24.
www.AJOG.org 3. Weiss JL, Malone FD, Emig D, et al. Obesity, obstetric complications and cesarean delivery rate—a population-based screening study. Am J Obstet Gynecol 2004;190:1091-7. 4. Gardosi J, Williams M, Francis A. Stillbirth in pregnancies of obese mothers is associated with increased risk of fetal growth restriction. Am J Obstet Gynecol 2009;206:S223. 5. Gardosi J, Williams M, Francis A. Clinical causes of stillbirth associated with maternal obesity. Am J Obstet Gynecol 2009;201:S223. 6. Chu SY, Kim SY, Lau J, et al. Maternal obesity and risk of stillbirth: a metaanalysis. Am J Obstet Gynecol 2007;197:223-8. 7. Cnattingius S, Bergstrom R, Lipworth L, Kramer MS. Prepregnancy weight and the risk of adverse pregnancy outcomes. N Engl J Med 1998;338:147-52. 8. Challis JR, Lockwood CJ, Myatt L, Norman JE, Strauss JF 3rd, Petraglia F. Inflammation and pregnancy. Reprod Sci 2009;16: 206-15. 9. Ziccardi P, Nappo F, Giugliano G, et al. Reduction of inflammatory cytokine concentrations and improvement of endothelial functions in obese women after weight loss over one year. Circulation 2002;105:804-9. 10. Yudkin JS, Stehouwer CD, Emeis JJ, Coppack SW. C-reactive protein in healthy subjects: associations with obesity, insulin resistance, and endothelial dysfunction: a potential role for cytokines originating from adipose
Obstetrics tissue? Arterioscler Thromb Vasc Biol 1999;19: 972-8. 11. Maynard SE, Min JY, Merchan J, et al. Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest 2003;111:649-58. 12. Levine RJ, Maynard SE, Qian C, et al. Circulating angiogenic factors and the risk of preeclampsia. N Engl J Med 2004;350:672-83. 13. Adya R, Tan BK, Punn A, Chen J, Randeva HS. Visfatin induces human endothelial VEGF and MMP-2/9 production via MAPK and PI3K/akt signalling pathways: novel insights into visfatin-induced angiogenesis. Cardiovasc Res 2008;78:356-65. 14. Suwaki N, Masuyama H, Nakatsukasa H, et al. Hypoadiponectinemia and circulating angiogenic factors in overweight patients complicated with pre-eclampsia. Am J Obstet Gynecol 2006;195:1687-92. 15. Masuyama H, Segawa T, Sumida Y, et al. Different profiles of circulating angiogenic factors and adipocytokines between early- and lateonset pre-eclampsia. BJOG 2010;117:314-20. 16. McElrath TF, Lim KH, Pare E, et al. Longitudinal evaluation of predictive value for preeclampsia of circulating angiogenic factors through pregnancy. Am J Obstet Gynecol 2012;207:407.e1-7. 17. McElrath TF, Hecht JL, Dammann O, et al. Pregnancy disorders that lead to delivery before
Research
the 28th week of gestation: an epidemiologic approach to classification. Am J Epidemiol 2008;168:980-9. 18. ACOG Committee on Practice Bulletins— Obstetrics. Diagnosis and management of preeclampsia and eclampsia. ACOG practice bulletin number 33. Obstet Gynecol 2002;99: 159-67. 19. Smith GC, Crossley JA, Aitken DA, et al. Circulating angiogenic factors in early pregnancy and the risk of preeclampsia, intrauterine growth restriction, spontaneous preterm birth, and stillbirth. Obstet Gynecol 2007;109: 1316-24. 20. Faupel-Badger JM, Staff AC, Thadhani R, et al. Maternal angiogenic profile in pregnancies that remain normotensive. Eur J Obstet Gynecol Reprod Biol 2011;158:189-93. 21. Callaway LK, O’Callaghan M, McIntyre HD. Obesity and the hypertensive disorders of pregnancy. Hypertens Pregnancy 2009;28: 473-93. 22. Roberts VH, Smith J, McLea SA, Heizer AB, Richardson JL, Myatt L. Effect of increasing maternal body mass index on oxidative and nitrative stress in the human placenta. Placenta 2009;30:169-75. 23. Sundrani D, Khot V, Pisal H, et al. Gestation dependent changes in angiogenic factors and their associations with fetal growth measures in normotensive pregnancy. PLoS One 2013;8: e54153.
MONTH 2014 American Journal of Obstetrics & Gynecology
1.e7