Association of pre-pregnancy obesity versus excessive gestational weight gain with adverse maternal outcomes in the United States

Abstract

Objective:

Pre-pregnancy BMI >30 kg/m² or obesity (PPO) and excessive gestational weight gain (eGWG) are associated with increased risks for adverse maternal outcomes. There is little is known regarding the individual effects of PPO and eGWG. The objective of this study was to compare the effects of PPO and eGWG with adverse maternal outcomes.

Methods:

This was a retrospective cohort study of singleton, live births in the United States in 2018 using data from the National Vital Statistics System. Pregnancies complicated by pregestational diabetes, chronic hypertension, and with unknown maternal body mass index (BMI) were excluded. IOM recommendations was used to define GWG. These births were divided into four groups: 1) normal pre-pregnancy BMI and normal GWG, 2) normal pre-pregnancy BMI and eGWG, 3) PPO and normal GWG and 4) PPO and eGWG. Outcomes such as Gestational diabetes, hypertensive disorders of pregnancy, cesarean delivery, maternal transfusion, and medical intensive care unit (MICU) admissions were compared among groups using ANOVA, and multivariable.

Results:

Of the 1,432,602 births included in the study, 22.2%, 42.3%, 10.1%, and 25.4% were in Groups 1–4 respectively. Compared with pregnancies in Group 1 (adjusted OR (95%CI)), in Groups 2–4 the risk for gestational diabetes 0.72 (0.70–0.74), 2.80 (2.72–2.88) and 2.28 (2.22–2.34) respectively. All groups had higher risk of hypertensive disease of pregnancy 1.58 (1.53–1.64), 3.88 (3.74–4.03) and 5.07 (4.90–5.23); eclampsia: 1.61 (1.33–1.95), 2.99 (2.42–3.69) and 3.57, (2.97–4.29), and cesarean delivery: 1.21 (1.19–1.23), 1.97 (1.92–2.02), and 2.45 (2.40–2.50) in Groups 2–4 respectively.

Conclusion:

Both PPO and eGWG are independently associated with higher odds of gestational diabetes, gestational hypertension, preeclampsia and Cesarean sections, with the highest risk among pregnancies with both PPO and eGWG. This data supports the importance of pre-pregnancy weight management in preventing adverse pregnancy outcomes.

Introduction:

Incidence of obesity in the US has increased dramatically in last few decades (1,2). There is a concordant increase in rates of obesity among women of reproductive age, with rates approaching 40% (3), with rates of obesity. Obesity is associated with changes in the immune, metabolic, and inflammatory changes, which may explain many of the adverse outcomes associated with obesity (4–6), including gestational diabetes mellitus and hypertensive disease of pregnancy (7).

Excessive gestational weight gain (GWG), as defined by the Institute of Medicine (8), is also associated with adverse maternal outcomes such as gestational diabetes mellitus, gestational hypertension, and higher rates of cesarean delivery (9–11) Furthermore, excessive GWG is associated with an increased risk of severe maternal morbidity, including eclampsia, heart failure, higher need for transfusions, and mechanical ventilation (12)

Whether pre-pregnancy obesity (PPO) has a more detrimental effect on pregnancy outcomes than excessive GWG has not been explored. The aim of this study was to compare the association of adverse maternal outcomes among pregnant women with PPO to those with normal prepregnancy BMI but with excessive GWG. Since obesity is related to increase in the general inflammatory state associated with immune-metabolic alterations in the body (4–6), we hypothesized that given the longer standing obesity, women with PPO would have worse outcomes as compared to the shorter-term obesity among pregnant women who had normal pre-pregnant BMI but had excessive GWG.

Methods:

This was a cross-sectional, retrospective cohort study of singleton, live births among women who had had first prenatal care by 12 weeks gestation in the United States for 2018. Women with pregestational diabetes and chronic hypertension were excluded. Data were extracted from the National Vital Statistics System, which is based on a summary of annual birth certificate data that are submitted by mandatory reporting procedures within each state. This study was exempt from institutional review board approval given that this data is deidentified and publicly available at www.cdc.gov/nchs/nvss. Normal pre-pregnancy BMI was defined as patients with pre-pregnancy BMI between 18 and 25 kg/m². Obesity was defined as those with a pre-pregnancy BMI >30 kg/m². The 2009 Institute of Medicine GWG guidelines were used to define normal and excessive GWG, which were based on the pre-pregnancy BMI (8). As there were preterm deliveries included in this cohort, weight gain was adjusted for the gestational age at the time of delivery. Thus, we used the recommended rates of weight gain per week in the second trimester (8). To differentiate the PPO from excessive GWG, we included only those pregnant women who had their first examination before 12 weeks of pregnancy.

Study groups:

The women were divided into four groups: Group 1) those with normal pre-pregnancy BMI and normal GWG (reference group); Group 2) women with a normal pre-pregnancy BMI and excessive GWG; Group 3) women with PPO and normal GWG, and Group 4) women with PPO and excessive GWG.

Clinical characteristics:

Demographic and clinical characteristics were extracted, including maternal age, self-identified race, parity, history of preterm birth, previous cesarean delivery and neonatal sex. Adverse maternal outcomes were defined as gestational diabetes, hypertensive disease of pregnancy, eclampsia, cesarean delivery, blood transfusions, and maternal intensive care unit (ICU) admission. Hypertensive disease of pregnancy in this dataset was defined as births complicated by gestational hypertension or preeclampsia. Given that the recommended timing for universal screening for gestational diabetes is 24–28 weeks (13), only births at or after 28 weeks were included when analyzing gestational diabetes as an outcome.

Statistical analysis:

Continuous demographic and clinical data were presented as mean ± standard deviation and compared across the four groups using one-way ANOVA. Categorical demographic and clinical data were presented as frequencies and percentages and compared using the chi square test. Multivariable logistic regression analysis was performed to assess the differences in neonatal adverse outcomes among the four study groups, with the group comprised of patients with normal pre-pregnancy BMI and normal GWG as the referent group. Demographic and clinical variables that were significantly different (p<0.05) among the groups were included in the logistic regression. We ran goodness of fit model or Hosmer-Lemeshow test for the logistic regression modelling. These analyses were presented as adjusted odd ratios (aOR), 95% confidence interval (CI). All data analyses were performed on STATA 17.0, StataCorp LLC, College Station, Texas, USA.

Results:

Of the 3,801,534 births in 2018, 1,432,602 (37.7%) met inclusion criteria. The most common reason for exclusion was initiation of prenatal care after 12 weeks gestation (932,549, 24.5%). Of the 1,432,602 births that met inclusion criteria, 318, 709 (22.2%) women had normal pre-pregnancy BMI and normal GWG (Group 1), 606,149 (42.3%) had normal pre-pregnancy BMI and excessive GWG (Group 2), 144,300 (10.1%) had PPO and normal GWG (Group 3), and 363,444 (25.4%) had both PPO and excessive GWG (Group 4).

Demographic characteristics are presented in Table 1. Maternal age and race differed among the groups (p<0.001). Patients with a normal pre-pregnancy BMI and excessive GWG were most likely to be nulliparous compared to the other 3 groups (p<0.001). Among multiparous patients, patients with PPO with or without excessive GWG had higher percentage of previous cesarean section deliveries (Groups 3 and 4: 38.0% and 42.6% respectively as compared to Groups 1 and 2: 22.7% and 26.2% respectively, p<0.001) and history of previous preterm delivery (Groups 3 and 4: 5.7% and 5.8% respectively as compared to Groups 1 and 2: 4.5% and 4.9%).

Table 1:

Demographic and clinical characteristics

Demographic and clinical characteristic	Normal PP BMI and GWG N=318,709	Normal PP BMI and excessive GWG N=606,149	PPO and normal GWG N=144,300	PPO and excessive GWG N=363,444	p-value
Age (years)	29.8±5.5	29.3±5.6	29.4±5.5	29.2±5.6	<0.001
Race					<0.001
White	185,450 (60.1) [308,398]	380,999 (65.2) [584,366]	69,690 (50.3) [138,575]	192,217 (55.2) [348,076]
Black	24,484 (7.9) [308,398]	53,983 (9.2) [584,366]	23,225 (16.8) [138,575]	59,276 (17.0) [348,076]
Hispanic	61,169 (19.8) [308,398]	101,416 (17.4) [584,366]	41,661 (30.1) [138,575]	88,758 (25.5) [348,076]
Asian	37,295 (12.1) [308,398]	47,968 (8.2) [584,366]	3,999 (2.9) [138,575]	7,825 (2.3) [348,076]
Nulliparity	130,090 (40.8)	274,576 (45.3)	43,683 (30.3)	139,543 (38.4)	<0.001
Previous preterm birth*	8,397 (4.5) [188,619]	16,257 (4.9) [331,573]	5,713 (5.7) [100,617]	12,973 (5.8) [223,901]	<0.001
Previous cesarean delivery*	69,304 (22.7) [188,619]	154,889 (26.2) [331,573]	49,986 (38.0) [100,617]	145,292 (42.6) [223,901]	<0.001
Neonatal male	158,692 (49.8)	318,616 (52.5)	72,237 (50.1)	189,916 (52.3)	<0.001

PP- prepregnancy, PPO- Prepregnancy Obesity, BMI- Body Mass Index; GWG-Gestational Weight Gain

Data presented as mean±sd or N (%) [N if missing data]

chi square test performed except for age where one way ANOVA was used

^*Among multiparas

The adverse maternal outcomes among the four study groups are detailed in Table 2. Compared to Group 1, Group 3 and Group 4 had higher rates of gestation diabetes [Group 3: aOR 2.28 (95% CI 2.22–2.34) and Group 4: 2.80 (2.72–2.88)]. Rates of gestational diabetes were lower in Group 2:0.72 (0.70–0.74) compared to Group 1. Compared to Group 1, Group 2 had a higher incidence of hypertensive disease of pregnancy 1.58 (1.53–1.64) and eclampsia 1.61 (1.33–1.95). The odds were even higher among women in Group 3 [3.88 (3.74–4.03) and 2.99 (2.42–3.69)] and Group 4 [5.07 (4.90 to 5.23) and 3.57 (2.97–4.29)] for hypertensive disease of pregnancy and preeclampsia respectively. Delivery by Cesarean section was more common in Group 2:1.21 (1.19–1.23), and even higher in Group 3:1.97 (1.92–2.02)) and Group 4:2.45 (2.40–2.50). A similar trend was observed for the maternal ICU admissions during the present pregnancy, though only reached statistical significance for women in Group 4: (1.35 (1.11–1.63)). Number of blood transfusions were significantly higher in Group 2:1.14 1.02–1.27)) as compared to Group 1.

Table 2:

Adjusted OR of maternal outcomes associated with the 3 study groups as compared to Group1 (normal prepregnancy BMI and GWG)

Outcomes	Normal PP BMI and GWG (Referent) N=318,709	Normal PP BMI and excessive GWG N=606,149	aOR, (95% CI)*	PPO and normal GWG N=144,300	aOR, (95% CI)**	PPO and excessive GWG N=363,444	aOR, (95% CI)***
Gestational diabetes****	14,702 (4.7%) [316,193]	19,088 (3.2%) [600,134]	0.69 (0.67–0.71)	17,544 (12.3%) [142,168]	2.92 (2.83–3.00)	35,072 (9.8%) [358,059]	2.27 (2.21–2.33)
Gestational hypertension	10,210 (3.2%)	33,875 (5.6%)	1.60 (1.55 –1.66)	15,087 (10.5%)	3.84 (3.70–3.98)	51,605 (14.2%)	5.00 (4.89 –5.11)
Eclampsia	320 (0.1%)	1,081 (0.2%)	1.61 [1.33–1.95]	439 (0.3%)	3.03 (2.47–3.72)	1,465 (0.4%)	3.80 (3.18– 4.54)
Cesarean delivery	72,982 (22.9%)	159,852 (26.4%)	1.20 (1.18 –1.22)	55,395 (38.4%)	1.98 (1.93–2.02)	156,547 (43.1%)	2.55 (2.52–2.57)
Maternal transfusion	992 (0.31%)	2,038 (0.34%)	1.10 (0.99–1.22)	419 (0.29%)	0.98 (0.85–1.13)	1,310 (0.36%)	1.09 (0.98–1.23)
MICU admission	334 (0.10%)	651 (0.11%)	0.99 (0.82–1.19)	196 (0.14%)	1.27 (1.01–1.60)	569 (0.16%)	1.40 1.16–1.68

BMI body mass index; GWG gestational weight gain; PPO prepregnancy Obesity

Data presented as N (%) [if N is missing]

^*Adjusted for race, nulliparity, history of preterm birth, and history of prior cesarean delivery (Goodness of fit (Hosmer-Lemeshow): gestational diabetes (removed race, p=0.174), gestational hypertension (removed race, nulliparity, and prior cesarean delivery, p=0.229), cesarean delivery (removed race, nulliparity, and history of preterm birth, p=0.390), maternal transfusion (removed race, p=0.768), eclampsia, p=0.103; MICU admission, p=0.204)

^**Adjusted for nulliparity, history of preterm birth, and history of prior cesarean delivery (Goodness of fit (Hosmer-Lemeshow): gestational diabetes (removed nulliparity and prior cesarean delivery, p=0.327), gestational hypertension (removed nulliparity and prior cesarean delivery, p=0.114), cesarean delivery (removed nulliparity and history of preterm birth, p=0.521), eclampsia (p=0.444), maternal transfusion (p=0.823), MICU admission (removed nulliparity and history of preterm birth, p=0.10)

^***Adjusted for nulliparity, history of preterm birth, and history of prior cesarean delivery (Goodness of fit (Hosmer-Lemeshow): gestational diabetes (p=0.160), gestational hypertension (univariable analysis as no goodness of fit p>0.05), cesarean delivery (univariable analysis as no goodness of fit p>0.05), maternal transfusion (removed p=0.154), eclampsia (removed race and prior cesarean delivery, p=0.804), MICU admission (removed race, p=0.362).

^****Among births that delivered at >=32 weeks, given that routine screening for gestational diabetes is 24–28 weeks

Discussion:

We found that women with PPO had more adverse maternal outcomes. Patients with PPO and normal GWG had higher odds of adverse maternal outcomes than those who did not have PPO but had excessive GWG. Those who had both pre-pregnancy obesity and excessive GWG had the highest rates of adverse maternal outcomes. These findings may reflect the longer duration of obesity-related immune-metabolic changes (4–6), among those patients with PPO as compared to those with normal prepregnancy BMI and increased GWG.

Excessive GWG has previously been associated with adverse outcomes. Using data from the Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-To-Be (NuMoM2b), Dude et al. found that patients with excessive GWG were more likely to have hypertensive disorders and cesarean delivery than patients who had normal GWG (11). In addition, in a systematic review, Goldstein et al. also found that those with excessive GWG had an increased risk for cesarean delivery (12). Our findings were similar in that patients with excessive GWG had higher rates of hypertensive disease of pregnancy and cesarean delivery. Interestingly, the risk of gestational diabetes was lower in those who had normal prepregnancy BMI and excessive GWG. Increased insulin levels without hyperglycemia have been shown to be protective (14) and shorter duration of increased weight gain such as during pregnancy may reflect this state. This is further supported by the findings that excessive GWG in the first trimester was more strongly associated with development of gestational diabetes mellitus than with any other trimester (15). Thus, it is possible that the longer period of excessive GWG among those who gained weight from first trimester had hyperinsulinemia along with hyperglycemia and therefore developed gestational diabetes. It is also possible that women in Group 1 who gained significantly more weight in the first trimester (as the study included all women who had one visit before 12 weeks that could be in the later part of the first trimester missing the actual weight gain in early first trimester) and had a higher risk for gestational diabetes mellitus than those who were categorized as excessive weight gain who had lower weight gain during the first trimester and more during later trimester. Among the women with PPO and excessive GWG increased admissions to the medical ICU during present pregnancy is not only associated with increased morbidities but could contribute to increased cost of care specially among PPO women (16).

There are several limitations to this study. As with any administrative database, we were limited to the available variables, and thus were not able to assess other outcomes of interest, such as the indication for cesarean delivery. However, these limitations were applicable to all four study groups and thus should not be a source of bias. In addition, we excluded records with missing variables, reducing our population for analysis. Additionally, we were not able to define the timeline for the development of excessive GWG, thus we were not able to assess whether excessive GWG in certain trimesters was more likely to be associated with adverse maternal outcomes. The main strength of this study is the large sample size with the availability of pre-pregnancy weight and gestational weight gain and important adverse maternal outcomes. This sample size allowed us to assess and adjust for potential confounders within the four main groups. Our data emphasizes importance of addressing obesity prior to getting pregnant.

In conclusion, we found that PPO was associated with higher rates of adverse maternal outcomes than normal prepregnancy BMI with excessive GWG. This data highlights the importance of preconception counseling on pre-pregnancy obesity to optimize the maximum benefit from weight loss prior to pregnancy (17,18).