Excessive Gestational Weight Gain Associated with Higher Risk of Cesarean Delivery across Body Mass Index Categories in the United States

Abstract

Objectives:

This research investigated associations between rate of gestational weight gain (GWG; “below recommended rate,” “within recommended rate,” and “exceeding recommended rate”) and increased risk of cesarean delivery for women across a range of body mass index (BMI) categories (underweight, normal, overweight, obesity classes I, II, and III).

Study Design:

We used vital records data from the National Center for Health Statistics to include nulliparous, singleton term births with cephalic presentation (n=7,891,653) to women who gave birth in the US from 2014 through 2022. We evaluated the association of GWG on mode of birth (cesarean vs. vaginal delivery) using multivariable linear probability (categorical GWG) and generalized additive (continuous GWG) models.

Results:

Approximately 26% of births were by cesarean delivery and 65% exceeded the recommended rate of GWG. Risk of cesarean delivery increased as BMI increased, and those in pre-pregnancy BMI above normal categories had higher risk of cesarean delivery. Excessive GWG (EGWG) increased the risk of cesarean delivery for all women, but the effect was more pronounced in those with a higher pre-pregnancy BMI. Among women whose BMI was underweight or normal, GWG had a J-shaped relationship with cesarean delivery risk, while women whose pre-pregnancy BMI was categorized as overweight or obese demonstrated a relatively linear increase in risk with any GWG.

Conclusions:

Findings provide valuable insights on EGWG and increased risk of cesarean delivery for women across BMI categories. Additional efforts are needed at both the patient and provider level to reduce EGWG and the risk of cesarean delivery.

INTRODUCTION

Maternal morbidity and mortality rates in the United States (US) exceed those of other high-income countries (1, 2). One of the factors contributing to severe maternal morbidity is the prevalence of cesarean deliveries (3–5). In the US, the percentage of women undergoing cesarean deliveries has risen significantly, from 7% in 1990 to 32% in 2023, surpassing the World Health Organization’s acceptable range of 10-15% (6–8).

Although cesarean delivery can decrease maternal and neonatal morbidity and mortality when medically necessary, unnecessary cesarean delivery increases the risk of short- and long-term adverse outcomes for both the mother and infant (4). Cesarean deliveries are associated with risks of maternal complications, including hemorrhage requiring transfusion, infections, septic shock, uterine rupture, placenta previa, hysterectomy, and complications in future pregnancies (5). Epidemiological studies have shown that a cesarean delivery is associated with a higher incidence of noncommunicable diseases in infants, including asthma, food allergies, and obesity (5). These findings may be partially explained by a lack of exposure to diverse microorganisms encountered during vaginal birth, which contribute to development of the immune system (9). Reducing cesarean deliveries is a priority outlined by the Centers for Disease Control and Prevention Healthy People 2030, the American College of Obstetricians and Gynecologists (ACOG), and the Alliance for Innovation on Maternal Health (10, 11).

One of the noted risk factors for cesarean delivery is excessive gestational weight gain (EGWG). The Institute of Medicine (IOM) and World Health Organization provide specific guidelines for appropriate gestational weight gain (GWG) based on pre-pregnancy body mass index (BMI) in an effort to reduce risks associated with weight gain below or exceeding recommend rates during pregnancy and promote healthy outcomes for mothers and infants (12, 13). With the growing number of women experiencing pre-pregnancy obesity and EGWG (14, 15), exploring the relationship between pre-pregnancy obesity and EGWG may provide insights for reducing the high rates of cesarean deliveries in the US. Although several studies have identified an association between EGWG and cesarean delivery, most of the investigations involved small sample sizes and were conducted among a single racial or ethnic group (16–18). Previous research has not explored whether the relationship between EGWG and cesarean delivery risk varies across BMI categories. Numerous clinical and nonclinical factors associated with cesarean deliveries are difficult to modify; pre-pregnancy BMI and pre-pregnancy health conditions, as well as factors such as maternal age, race, socioeconomic status, and insurance coverage, are rooted in individual circumstances and/or systemic issues that can be challenging to change through discrete interventions (14, 19–21). However, GWG is a potentially modifiable factor. This research investigated associations between EGWG and increased risk of cesarean delivery among low risk (nulliparous, singleton term births with cephalic presentation) for women across a range of BMI categories.

METHODS

Data and Population

The study used vital records data from the National Center for Health Statistics and was approved as an exempt study by the University of Arkansas for Medical Sciences Institutional Review Board. The study population included singleton live births to women who gave birth in the United States from January 1, 2014 through December 31, 2022. Our goal was to examine the effects of EGWG on cesarean delivery in low-risk pregnancies. Therefore, the study sample was limited to live births that were: singleton, term (gestational age of 37-42 weeks), cephalic presentation at birth, and first birth in the total birth order to exclude women who had a prior cesarean delivery.

Variable Definitions

The primary outcome of interest was mode of birth (cesarean vs. vaginal delivery). The main risk factor of interest was EGWG. EGWG was defined by the rate of GWG in the second and third trimesters. Although many epidemiological studies focus on total weight gain, evaluating the rate of weight gain as represented in the IOM guidelines (12) provided a more nuanced assessment and allowed us to account for the natural variations in gestational length, i.e., weeks gestation (22). For example, a woman with a delivery at 40 weeks may gain more weight and stay within the recommended ranges than a woman with a delivery at 36 weeks. In addition, for comparison with other published studies, we also completed the analysis using total weight gain as a predictor (see Supplemental Appendix) which yielded similar results.

Rate of GWG was defined as below, within, or exceeding recommended rates depending on if GWG per week was below, within, or exceeding the IOM recommended ranges based on pre-pregnancy BMI (underweight: 1.0–1.3 lbs/week [454–590 g/week]; normal: 0.8–1.0 lbs/week [363–454 g/week]; overweight: 0.5–0.7 lbs/week [227–318 g/week]; and obesity class I–III: 0.4–0.6 lbs/week [181–272 g/week]). Rate of GWG for the second and third trimesters (GWG per week) was calculated by using the methods documented by Wu et al., 2024 (23), where 2.75 lbs is the midpoint of the suggested GWG range (1.1–4.4 lbs) for the first trimester, (12) gestational age is GA, and 12 completed weeks is the length of the first trimester.

$$RateofGWG=\frac{TotalGWG-2.75lbs}{GA-12weeks}$$

The analysis was adjusted for demographic variables. Payers included Medicaid, private insurance, other (e.g., Indian Health Service or TRICARE), and self-pay. Maternal race/ethnicity was coded as Hispanic (regardless of race) or non-Hispanic American Indian or Alaska Native (AIAN), Asian, White, Black, Native Hawaiian or Pacific Islander (NHPI), or multiracial. The county of residence of the birth mother was coded as being “metro” (i.e., urban) or “non-metro” (i.e., rural) using the 2023 Rural-Urban Continuum Codes from the United States Department of Agriculture (24). The birth year (i.e., 2014-2022) and state of occurrence were also included as fixed effects. Additional covariates included other maternal risk factors such as maternal age (continuous); gestational age (37-42 weeks); pre-pregnancy BMI category (underweight, normal, overweight, obesity class I, obesity class II, or obesity class III); diabetes (pre-pregnancy and gestational); hypertension (pre-pregnancy and gestational); and induction of labor.

Statistical Analyses

We utilized a linear probability model (i.e., ordinary least squares regression) to estimate the risk of cesarean delivery. We estimated marginal predicted probabilities and calculated mean absolute risk differences (RD). The linear model included all covariates, the main effect of EGWG, the main effect of BMI, and an interaction between BMI and EGWG. The interaction was included due to the potential modification of EGWG effects by pre-pregnancy BMI (i.e., a multiplicative effect of GWG with BMI). Additionally, we fit a generalized additive model to model GWG as a continuous variable using the “mgcv” R package with thin-plate regression splines with a smooth factor interaction between BMI category and GWG (25). From this model, the smooth terms (spline) can be interpreted using the effective degrees of freedom (edf) to measure the complexity of the smooth term in the model. A higher edf implies a more complex, “wiggly” curve capable of capturing finer details in the relationship between the predictor and the response variable (25). If the edf for a smooth term is substantially different across levels of the interacting factor (i.e., BMI categories), it suggests the relationship between the predictor (GWG) and response (cesarean delivery) is markedly different for those factor levels. Calculation and visualizations of the predicted probabilities and calculation of the absolute RDs and the partial derivative of the slopes were created using the “marginaleffects” R package (26). Additional analysis details are included in a supplemental file showing the regression coefficients and other model information. Statistical significance was assumed at p < 0.01, and 99% confidence intervals were reported. All analyses were conducted in R version 4.4.1 (27).

RESULTS

The analytical sample included 7,891,653 singleton live births. Table 1 provides characteristics of the study population split by rate of GWG categorization. Among our sample, approximately 26% of births were by cesarean delivery and 65% exceeded the recommended rate of GWG (see Supplemental Appendix for more details).

Table 1:

Characteristics of the study population of singleton births.

Characteristic	Below Rec. N = 1,629,350	Within Rec. N = 1,129,670	Exceeding Rec. N = 5,132,633	p-value1
Delivery Method				<0.001
Spontaneous	73%	72%	67%
Forceps	0.9%	1.0%	1.0%
Vacuum	5.3%	5.3%	4.7%
Cesarean	21%	22%	27%
Pre-pregnancy BMI				<0.001
Underweight	7.4%	8.1%	2.1%
Normal	52%	54%	45%
Overweight	16%	19%	29%
Obesity I	10%	9.5%	14%
Obesity II	7.3%	5.3%	5.9%
Obesity III	6.8%	3.7%	3.4%
Mother’s Age				<0.001
< 20 years	15%	12%	12%
20-24 years	28%	26%	29%
25-29 years	27%	29%	30%
30-34 years	21%	24%	22%
35-39 years	7.5%	8.0%	6.9%
>= 40 years	1.5%	1.4%	1.1%
Pre-pregnancy diabetes	0.7%	0.5%	0.6%	<0.001
Pre-pregnancy hypertension	1.5%	1.2%	1.6%	<0.001
Gestational diabetes	7.5%	5.7%	4.6%	<0.001
Gestational hypertension	5.7%	5.8%	9.7%	<0.001
Induction of Labor	33%	33%	38%	<0.001

¹Pearson’s Chi-squared test

Note: Rec. = Recommended Rate

Figure 1 displays the average predicted probability of cesarean delivery by BMI category and rate of GWG categorization averaged over covariates. Inspection of the regression coefficients indicated an interaction between GWG and BMI (all coefficients; p < 0.01).

Figure 1:

Predicted population average probability of cesarean delivery with 99% confidence intervals.

Table 2 provides the mean absolute risk difference for cesarean delivery between BMI categories based on marginal predicted probabilities from the multivariable linear probability model. Overall, the risk of cesarean delivery increased as BMI increased such that those in higher pre-pregnancy BMI categories had higher risk of cesarean delivery (Table 2). Additionally, the magnitude of increased risk between adjacent BMI categories (as indicated by the risk differences in Table 2) increased as BMI increased. For example, women who had obesity class I BMI had an approximately 6.23% increase (p < 0.001) in the risk of cesarean delivery compared to those with an overweight pre-pregnancy BMI; this is much larger than the 1.43% increase in risk for women with a normal pre-pregnancy BMI compared to women with an underweight pre-pregnancy BMI.

Table 2:

Mean absolute risk difference for cesarean delivery between BMI categories.

BMI Comparison	RD	z	p-value	99% C.I.
Normal vs Underweight	1.43%	11.02	<0.001	[1.09%, 1.76%]
Overweight vs Normal	5.62%	127.33	<0.001	[5.51%, 5.74%]
Obesity I vs Overweight	6.23%	104.42	<0.001	[6.08%, 6.39%]
Obesity II vs Obesity I	6.52%	72.04	<0.001	[6.29%, 6.76%]
Obesity III vs Obesity II	9.03%	71.86	<0.001	[8.71%, 9.36%]

Table 3 provides the mean absolute risk difference for cesarean delivery between BMI categories, stratified by pre-pregnancy BMI. A higher rate of GWG was associated with a greater risk of cesarean delivery. However, as shown by the risk differences in Table 3, the magnitude of increased cesarean delivery risk due to increased rate of GWG (e.g., below recommended rate vs. within recommended rate or within recommended rate vs. exceeding recommended rate) depended upon a woman’s pre-pregnancy BMI. Higher pre-pregnancy BMIs were associated with larger differences in cesarean delivery risk due to increased rate of GWG. For example, EGWG for a woman with normal pre-pregnancy BMI was associated with a 4.48% increase in the risk of cesarean delivery and EGWG for a woman with a pre-pregnancy BMI who was in the obesity class I range was associated with a 5.86% increase in risk (Table 3). Additionally, a rate of GWG within recommendation carried a slightly elevated risk of cesarean delivery compared to a rate of GWG below recommendation across all BMI categories (RDs 0.91-3.89%; all differences p<0.001).

Table 3:

Mean absolute risk difference for cesarean delivery between GWG categories, stratified by pre-pregnancy BMI category.

GWG Comparison	BMI	RD	z	p-value	99% C.I.
Exceeding Rec. vs Within Rec.	Underweight	3.74%	19.92	<0.001	[3.26%, 4.23%]
	Normal	4.48%	74.18	<0.001	[4.33%, 4.64%]
	Overweight	4.59%	47.51	<0.001	[4.34%, 4.84%]
	Obesity I	5.86%	42.74	<0.001	[5.50%, 6.21%]
	Obesity II	5.97%	31.75	<0.001	[5.49%, 6.46%]
	Obesity III	5.21%	22.75	<0.001	[4.62%, 5.79%]
Exceeding Rec. vs Below Rec.	Underweight	5.64%	32.19	<0.001	[5.19%, 6.09%]
	Normal	5.39%	100.82	<0.001	[5.25%, 5.53%]
	Overweight	6.69%	75.19	<0.001	[6.46%, 6.92%]
	Obesity I	8.84%	77.04	<0.001	[8.55%, 9.14%]
	Obesity II	9.75%	67.77	<0.001	[9.38%, 10.12%]
	Obesity III	9.09%	56.57	<0.001	[8.68%, 9.51%]
Within Rec. vs Below Rec.	Underweight	1.90%	10.29	<0.001	[1.42%, 2.37%]
	Normal	0.91%	12.92	<0.001	[0.73%, 1.09%]
	Overweight	2.10%	17.26	<0.001	[1.79%, 2.42%]
	Obesity I	2.98%	18.11	<0.001	[2.56%, 3.41%]
	Obesity II	3.78%	17.92	<0.001	[3.23%, 4.32%]
	Obesity III	3.89%	16.11	<0.001	[3.27%, 4.51%]

Note: Rec. = Recommended Rate

Continuous Analysis of GWG and Visualization

To gain more precision in estimating the relationship between the rate of GWG and cesarean delivery risk, we employed generalized additive models with thin-plate regression splines. Our analysis revealed a complex interplay between pre-pregnancy BMI and GWG. Figure 2 shows the predicted probabilities of cesarean delivery by continuous GWG (Figure 2a) as well as the slope of the generalized additive model with the continuous GWG (Figure 2b). The risk of cesarean delivery initially decreased as the GWG rate increased for women with normal or underweight BMI, reaching a nadir of risk before subsequently rising (Figure 2A). This non-linear relationship between GWG rate and cesarean risk was statistically significant for both normal BMI (edf = 6.09, p < 0.001) and underweight BMI (edf = 3.57, p < 0.001). This suggests an optimal range of GWG for minimizing cesarean risk in those with underweight and normal pre-pregnancy BMI. The association between GWG rate and cesarean risk was more linear (positively) for overweight and obese women (Figure 2A).

Figure 2:

Visualization of the generalized additive model with GWG displaying A) the predicted population average probability of cesarean delivery across the range of GWG rate B) the partial derivatives of the slopes from the model of GWG rate. The error bands around each line represent 99% confidence intervals.

The spline analysis identified a GWG rate of approximately 0.6 pounds/week, 0.54 pounds/week for those with normal pre-pregnancy BMI, and 0.62 pounds/week for those with an underweight pre-pregnancy BMI, as a potential turning point for cesarean risk in women with normal weight and underweight BMI wherein risk is minimized (Figure 2B). However, among women with a pre-pregnancy BMI in the overweight or obese range, any increase in GWG was associated with a higher cesarean delivery risk (Figure 2B).

COMMENT

Principal Findings

This study evaluated the effect of rate of weight gain during pregnancy on risk of cesarean delivery for women with different pre-pregnancy BMI. Women had a higher risk of cesarean delivery when there was a rate of GWG exceeding recommendations across all pre-pregnancy BMI.

Results in the Context of What is Known

Findings are concerning because two-thirds of women had EGWG. Although the overall finding that EGWG is associated with cesarean delivery is consistent with prior literature (16–18, 28, 29), our analysis provides unique insights into this relationship by examining associations between cesarean delivery and both rate of GWG categorizations as well as rate of GWG as a continuous variable. Our analysis also uniquely examines multiple pre-pregnancy BMI categories and draws on the advantages provided by a large national sample.

For women with a pre-pregnancy BMI in the underweight and normal category, the risk of cesarean delivery decreased with GWG that was within the IOM recommendation (Figure 2). Our results suggested an optimal rate of GWG for minimizing cesarean risk could be 0.6 lbs/week in those with underweight and normal pre-pregnancy BMI. This nuanced finding using the rate of weight gain highlights the importance of considering continuous GWG rates rather than relying on categorical classifications. When categorized, even GWG within the recommended range was associated with an increase in the risk of cesarean delivery compared to GWG below the recommended rate, regardless of BMI classification, whereas a continuous view of GWG indicated that cesarean delivery risk diminished and remained low across a wide range of GWG rates in those that had an underweight and normal BMI pre-pregnancy. Additionally, women with a pre-pregnancy BMI in the overweight or obese categories had an increase in cesarean delivery risk with any GWG. This finding suggests the need to further evaluate the clinical implications of the current GWG recommendations among women who are overweight or obese.

The findings should be interpreted with some caution, as GWG during pregnancy is a complex process. Although lower GWG may reduce the risk of cesarean delivery, GWG below the recommended rate during pregnancy may have other maternal and neonatal risks (18, 30–32). Therefore, any “protective” effect against cesarean delivery associated with GWG below the recommended rate must considered with the potential for other maternal and neonatal risks.

Clinical Implications

Our findings have several implications for clinical practice. Although several risk factors for cesarean delivery are not modifiable during pregnancy, rate of GWG in the second and third trimesters is a modifiable risk factor. GWG interventions could be delivered as part of routine prenatal care or as specific interventions delivered as part of prenatal care. While there is an opportunity to intervene, studies have documented that few obstetrical providers counsel their patients using IOM guidelines for GWG (33–39).

Research Implications

Our findings also have important implications for future research. Studies have documented that obstetrical providers often do not provide GWG counseling consistent with IOM guidelines (33–39); however, there is a lack of research to understand this variation in counseling. There is no published literature on provider-level intervention to increase consistent GWG recommendations and reduce EGWG.

For example, GWG counseling or efforts to reduce EGWG are not included as part of the Alliance for Innovation on Maternal Health’s bundle to reduce primary cesarean birth (40). EGWG interventions including those focused on diet, physical activity, or combined diet and physical activity have demonstrated effectiveness (41, 42); however, EGWG interventions have primarily been conducted among affluent, urban, and suburban White women who are least affected by EGWG (42–45). The lack of research on the effectiveness of interventions to reduce EGWG among rural, minoritized, and low-income women leaves a significant gap in knowledge and perpetuates maternal health inequities (42–45). Future research should focus on examining interventions to reduce EGWG among diverse, low-income, and rural women. Implementation research is also needed to evaluate strategies to increase GWG counseling by providers.

Strengths and Limitations

The study analyzed vital records data from the National Center for Health Statistics birth records to calculate pre-pregnancy BMI and GWG (46). Our analysis examined associations between GWG and risk for cesarean delivery. This type of observational study is informative regarding current practice guidelines; however, associations do not establish causality. There could be confounders such as socioeconomic status or provider practice norms that are not available in the birth record data. Research has documented that women may underestimate pre-pregnancy weight (47); therefore, the total weight gain and rate of weight gain in our study may be overstated. Additionally, this study was restricted to low-risk deliveries (i.e., first in birth order, singleton, term, not presenting as breech), and we controlled for potential covariates (e.g., diabetes and hypertension). The study was also strengthened because we analyzed GWG using both categorical and continuous approaches.

Conclusions

Reducing cesarean delivery is a shared goal of the Centers for Disease Control and Prevention Healthy People 2030, ACOG, and the Alliance for Innovation on Maternal Health (10, 11), because unnecessary cesarean deliveries have negative health consequences for both mother and infant for both current and future deliveries (5, 9). GWG is a modifiable risk factor that if reduced or limited could reduce the risk of cesarean delivery. Although there was hope that the revised 2009 IOM guidelines, which lowered GWG recommendations for higher BMI categories, would reduce EGWG for women in higher BMI categories, studies have documented such reductions did not occur, and EGWG has increased among women with higher BMIs after guidelines were released (48). This study provides valuable insights on EGWG and an increased risk of cesarean delivery for women across BMI categories. Additional efforts are needed at both the patient and provider level to reduce EGWG and risk of cesarean delivery.

At a Glance:

A. Why was this study conducted?

The United States has higher prevalence of cesarean delivery than recommended, and it is important to understand the factors contributing to the high rates.

B. What are the key findings?

Excessive gestational weight gain is associated with higher risk of cesarean delivery for all body mass index categories.

C. What does this study add to what is already known?

While prior studies have demonstrated associations between excessive gestational weight gain and body mass index, this study is unique because it examined multiple pre-pregnancy body mass index categories and included a large national sample.