Racial Disparities in Preterm Birth among Pregnant Women with Obesity

Abstract

Objectives:

We assessed the impact of obesity and racial disparities on preterm birth (PTB) in the United States and sought to determine whether obesity widens the racial-ethnic disparity gap in preterm birth with a focus on non-Hispanic Black and White women.

Methods:

Using birth data for the years 2014–2019 made publicly available by the Centers for Disease Control and Prevention and obtained from the National Vital Statistics System, we conducted a cross-sectional cohort study analyzing a total of 14,864,844 births from 2014 to 2019.

Results:

We observed dose-dependent changes in obesity and PTB by defining obesity in subgroups and PTB in a stratified method. PTB occurred more among non-Hispanic Black women than their non-Hispanic White and Hispanic counterparts. We observed a consistent trend of increased PTB among women with high body mass index. Racial disparity existed in PTB among pregnant obese women, with non-Hispanic Black women exhibiting the greatest risk for PTB.

Conclusions:

Our work further contributes to the growing knowledge of existence of health disparity among the Black population.

In the United States, obesity continues to be a cause of grave concern in reproductive age and pregnant women, with almost 75% of reproductive-age Black women being obese.^1–4 The Centers for Disease Control and Prevention estimated an increase in prepregnancy obesity from 26.1% in 2016 to 29.0% in 2019. The percentage of women with prepregnancy obesity was highest in Blacks, whereas the highest increase in prepregnancy obesity during 2016–2019 was highest for Hispanics.⁵ It is well established that maternal overweight and obesity are risk factors for various adverse outcomes during pregnancy, including gestational diabetes mellitus, preeclampsia, cesarean delivery, and preterm (PTB) birth.^1–4,6 Extremely obese and superobese women have an increasingly higher risk of adverse birth outcomes, with Black women generally possessing the greatest risk of these adverse perinatal outcomes.^1,2,4A

Racial/ethnic disparities also are observed in the prevalence of obesity, with previous reports showing that non-Hispanic Black, Hispanics, and non-Hispanic Others have a higher prevalence compared with Whites and Asian Americans.¹ Significant racial and ethnic disparities also persist in the rates of PTB (defined as a birth before 37 weeks of gestation) and very PTB (VPTB, defined as a birth before 32 weeks of gestation) in the United States.⁷ Although significant racial differences have been identified in general perinatal outcomes, the joint effects of obesity and racial disparities on PTB have not been studied specifically. The aim of this study was to determine whether obesity widens the racial/ethnic disparity gap in PTB, especially among non-Hispanic Black and non-Hispanic White women.

Methods

Using birth data for the years 2014–2019 made publicly available by the Centers for Disease Control and Prevention and obtained from the National Vital Statistics System, we conducted this cross-sectional cohort study. The data files include information regarding sociodemographic and health characteristics and maternal risk factors associated with each live birth. The information was abstracted from the birth certificates filed in the vital statistics offices of each state in the United States and contains information on all of the births occurring within the United States.

Study Variables and Sample

We restricted the study to singletons within the gestational age of 20 to 42 weeks. The study variables included were gestational age, mother’s prepregnancy body mass index (BMI), age, race, nativity, maternal education, birth facility, prenatal care, delivery method, birth attendant, and fetal sex. Maternal prepregnancy BMI was categorized in the dataset as follows: underweight: <18.5, normal: 18.5 to 24.9, overweight: 25.0 to 29.9, obesity type I: 30.0 to 34.9, obesity type II: 35.0 to 39.9, and obesity type III: ≥40.0.⁸ We included mothers characterized as normal and those characterized as obese in this study.

Starting in 2014, the National Center for Health Statistics changed the method of measuring gestational age in these data files from those based on the last menstrual period to obstetric estimate of gestational age; therefore, obstetric estimate was used in this study to obtain the gestational age. Fetal gestational ages were categorized as follows: PTB refers to all deliveries occurring at <37 weeks, whereas term birth refers to deliveries from 37 weeks to 42 weeks.⁹ PTB was further subclassified into 3 phenotypes: extreme preterm (<28 weeks), very preterm (28–31 weeks), and moderate-to-late preterm (32–36 weeks, 6 days).⁶

We categorized PTB as medically induced if the labor was induced or the trial of labor was attempted in case of cesarean deliveries, or both; otherwise, the preterm delivery was categorized as spontaneous. The mother’s age was categorized as younger than 20 years; 20 to 29 years; 30 to 39 years; and 40 years old and older. Maternal race was classified as non-Hispanic White, non-Hispanic Black, Hispanic, or Others. Mother’s nativity was categorized into “born in US” and “born outside US”; highest educational attainment was grouped as “less than high school” and “high school graduate, General Educational Development Test completed, or higher.” Birth facility included categories of “hospital” and “others” and birth attendants were subgrouped into “medical professionals” and “others.” Prenatal care was categorized as the one “starting in first trimester,” “starting in second or third trimester,” and “no prenatal care.” Delivery method was subgrouped as vaginal or cesarean section.

Statistical Analysis

The study was performed using publicly available deidentified data and it was approved as exempt by the institutional review board of our institution. All of the statistical analyses were performed using R version 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria) and RStudio version 1.1.423. Pearson χ² tests were conducted to identify differences of proportions of term and preterm birth across various birth characteristics. We calculated the prevalence of preterm birth by sociodemographic and birth characteristics. Next, we calculated the rate of preterm birth phenotypes across maternal BMI subgroups. We used adjusted logistic regression modeling to compute adjusted odds for the association between prepregnancy maternal BMI and preterm birth. Furthermore, to study the differential impact of maternal obesity on spontaneous versus medically induced preterm birth, we generated adjusted logistic regression models. Lastly, we evaluated the adjusted association between obesity phenotypes and PTB. Maternal age, race, nativity and educational qualification, birth facility, prenatal care, delivery method, birth attendant, and fetal sex were the covariates used to adjust in the models.

Results

In this study, we analyzed a total of 14,864,844 births from 2014 to 2019, 908,432 of which were preterm (6.1%). Table 1 displays the sociodemographic characteristics of PTBs and term births. Younger (younger than 20 years old) and older women (40 years old and older) had a higher prevalence of PTB, with a prevalence rate of 8.4% for women younger than 20 years old and 8.7% for women 40 years old and older. Non-Hispanic Black women had the highest prevalence of PTB at 10.1%, which was double the rate for non-Hispanic White women. The prevalence of PTB in women born in the United States was higher compared with the women born outside the United States, with a rate of 6.2% and 5.7%, respectively. Mothers with less than a high school–level education had the highest prevalence of PTB (8.5%) compared with the women with a higher level of education than high school (5.7%). Mothers who did not receive prenatal care had a higher prevalence (13.4%) of PTB when compared with women who received prenatal care in the first and second or third trimesters, respectively (5.3%, 7.9%). Women who delivered via cesarean section had a prevalence rate of 8.9% for PTB, which was double the prevalence rate for women delivering vaginally (4.9%).

Table 1.

Sociodemographic characteristics of women with preterm and term birth

	Term		Preterm		Prevalence
	N = 13,956,412	% = 93.9	N = 908,432	% = 6.1
Maternal age, y
<20	725,438	5.2	66,398	7.3	8.4
20–29	6,899,534	49.4	439,629	48.4	6.0
30–39	5,917,682	42.4	363,119	40.0	5.8
≥40	413,758	3.0	39,286	4.3	8.7
Race
NH-White	7,503,556	53.8	398,868	43.9	5.0
NH-Black	1,868,983	13.4	210,818	23.2	10.1
Hispanic	3,136,657	22.5	213,372	23.5	6.4
NH-Others	1,334,934	9.6	78,367	8.6	5.5
Missing	112,282	0.8	7007	0.8	5.9
Mother’s nativity
Born in US	10,906,120	78.1	722,949	79.6	6.2
Born outside US	3,025,869	21.7	182,943	20.1	5.7
Missing	24,423	0.2	2540	0.3	9.4
Education
Less than high school	1,743,179	12.5	161,856	17.8	8.5
High school graduate, GED completed, or higher	12,060,130	86.4	735,406	81.0	5.7
Unknown	153,103	1.1	11,170	1.2	6.8
Birth facility
Hospital	13,710,812	98.2	902,823	99.4	6.2
Others	245,108	1.8	5557	0.8	2.2
Unknown	492	0.0	52	0.0	9.6
Prenatal care
1st trimester	10,685,194	76.6	595,315	65.5	5.3
2nd or 3rd trimester	2,794,472	20.0	239,581	26.4	7.9
No prenatal care/missing	476,746	3.4	73,536	8.1	13.4
Delivery method
Vaginal	9,785,203	70.1	502,072	55.3	4.9
Cesarean section	4,166,606	29.9	406,024	44.7	8.9
Missing	4603	0.0	336	0.0	6.8
Birth attendant
Medical professionals	13,708,745	98.2	899,154	99.0	6.2
Other	240,610	1.7	8560	0.9	3.4
Unknown	7057	0.1	718	0.1	9.2
Sex of fetus
Female	6,832,538	49.0	419,438	46.2	5.8
Male	7,123,874	51.0	488,994	53.8	6.4

GED, General Educational Development test; NH, non-Hispanic.

Figure 1 shows the rates of the various PTB phenotypes categorized by maternal BMI. Moderate-to-late preterm was the most predominant PTB phenotype across all three obesity types and normal-weight women. There was a consistent increase in all three PTB phenotypes (extreme preterm, very preterm, and moderate-to-late preterm), with change in mother’s BMI category from normal to type 1, type 2, and type 3 obesity, respectively. The rate of PTB was directly proportional to the mother’s obesity subtype.

Fig. 1.

Rates of preterm birth phenotypes by maternal BMI. BMI, body mass index.

Table 2 shows the association between prepregnancy BMI and PTB after adjustment for other characteristics. The risk of PTB was greater among obese compared with nonobese mothers (odds ratio [OR] 1.08, 95% confidence interval [CI] 1.08–1.09). Another maternal–fetal characteristic predictive of PTB was maternal age 40 years old and older (OR 1.19, 95% CI 1.18–1.21) when compared with women younger than 20 years. The age groups 20 to 29 and 30 to 39 years showed a less significant association between maternal BMI and PTB (OR 0.82, 95% CI 0.81–0.83; OR 0.85, 95% CI 0.84–0.86), respectively. In comparison to non-Hispanic White, the highest association among race, BMI, and PTB was exhibited in non-Hispanic Black (OR 1.82, 95% CI 1.81–1.83) and Hispanics (hazard ratio 1.2, 95% CI 1.19–1.21), with non-Hispanic-Others having the lowest risk (OR 1.18, 95% CI 1.17–1.19).

Table 2.

Association between prepregnancy BMI and preterm birth after adjustment for other characteristics

	OR (95% CI)	P
Maternal BMI
Normal	Reference
Obese	1.08 (1.08–1.09)	<0.01
Maternal age, y
<20	Reference
20–29	0.82 (0.81–0.83)	<0.01
30–39	0.85 (0.84–0.86)	<0.01
≥40	1.19 (1.18–1.21)	<0.01
Race
NH-White	Reference
NH-Black	1.82 (1.81–1.83)	<0.01
Hispanic	1.2 (1.19–1.21)	<0.01
NH-Others	1.18 (1.17–1.19)	<0.01
Mother’s nativity
Born in US	Reference
Born outside US	1.23 (1.18–1.28)	<0.01
Education
Less than high school	Reference
High school graduate, GED completed, or higher	0.73 (0.72–0.73)	<0.01
Birth facility
Hospital	Reference
Others	0.44 (0.42–0.45)	<0.01
Prenatal care
1st trimester	Reference
2nd or 3rd trimester	1.43 (1.42–1.44)	<0.01
No prenatal care/missing	2.54 (2.52–2.56)	<0.01
Delivery method
Vaginal	Reference
Cesarean section	1.84 (1.84–1.85)	<0.01
Birth attendant
Medical professionals	Reference
Other	0.96 (0.93–0.98)	<0.01
Sex of fetus
Female	Reference
Male	1.11 (1.10–1.11)	<0.01

BMI, body mass index; CI, confidence interval; GED, General Educational Development Test; NH, non-Hispanic; OR, odds ratio.

Mother’s nativity showed that there was a risk association between BMI and PTB in women born outside the United States (OR 1.23, 95% CI 1.18–1.28) when compared with women born in the United States. The time of initiating prenatal care was another important predictor of risk association. Women who received prenatal care early during the first trimester had lower risk association between BMI and PTB (OR 1.43, 95% CI 1.42–1.44) when compared with the women who received prenatal care during their second and third trimesters (OR 2.54, 95% CI 2.52–2.56). When comparing the methods of delivery, mothers who delivered via cesarean section (OR 1.84, 95% CI 1.84–1.85) had a higher risk association between BMI and PTB than mothers who delivered vaginally. In regard to level of education, we found that there was a higher risk association between BMI and PTB in women with a high school or higher level of education (OR 0.73, 95% CI 0.72–0.73) compared with women with less than a high school–level education.

The prevalence of spontaneous PTB in our study was 4.76% (N = 707,797), whereas the prevalence of medically induced preterm birth was 1.35% (N = 200,635). Supplemental Table 1 ([COMP: INSERT SDC URL HERE]) shows the adjusted odds of spontaneous PTB and medically induced PTB, respectively, for obese women as compared with women with a normal BMI and after adjusting for other birth characteristics. We observed that obese women had an 8% higher odds of experiencing spontaneous PTB (OR 1.08, 95% CI 1.08–1.09) and a 37% increased odds of experiencing medically induced PTB (OR 1.37, 95% CI 1.36–1.38), as compared with women of normal BMI. As compared to women younger than 20 years of age, those in the 20- to 29-year-old group had approximately 1.07 times the odds (95% CI 1.06–1.07) of experiencing spontaneous preterm birth. Non-Hispanic Blacks had the highest odds of experiencing spontaneous (OR 1.83, 95% CI 1.82–1.84) and medically induced preterm birth (OR 1.60, 95% CI 1.58–1.62) as compared with other races/ethnicities. Mothers who were born outside the US, those who graduated high school or completed the General Educational Development Test or higher were less likely than their counterparts who were born in the United States and those with less than a high school education, respectively (P < 0.01) to experience spontaneous or medically induced preterm birth. Women receiving prenatal care after their first trimester of pregnancy and those who delivered via cesarean section had an statistically significantly increased odds of spontaneous or medically induced PTB as compared with those who received prenatal care in the first trimester and those who delivered vaginally, respectively.

In Figure 2, we present the adjusted ORs for the association between classes of maternal obesity and PTB. The obesity subtypes are directly proportional to PTB risk. The odds for PTB increased with increasing obesity subtype. Women in obesity class 3 had the greatest odds of PTB (OR 1.14, 95% CI 1.13–1.15) followed by type 2 obesity (OR 1.09, 95% CI 1.09–1.11) and type 1 obesity (OR 1.06, 95% CI 1.05–1.06), when compared with women of normal BMI.

Fig. 2.

Odds of preterm birth among obesity subtypes in comparison to women with normal BMI. BMI, body mass index.

Discussion

In this study, we found that PTB occurred more among non-Hispanic Black women than their non-Hispanic White and Hispanic counterparts. We observed that the occurrence of PTB among varying maternal age groups followed a “U” shape pattern, with the highest prevalence in young mothers (younger than 20 years old) and older mothers (40 years old and older). Notably, PTB occurrence increased with increasing BMI. Our data showed that racial disparity existed in PTB among pregnant women with obesity, with obese non-Hispanic Black mothers exhibiting the greatest adjusted odds for PTB.

When we assessed the independent association between obesity phenotypes and PTB and other maternal characteristics, we found that maternal factors such as age and initiation of prenatal care played a vital role in determining PTB among obese women. Several studies have established the relation between obesity and adverse maternal birth outcomes as well as an association between PTB and obesity.^3,10,11 Snowden et al found racial/ethnic disparities within several adverse birth outcomes and increased rates of adverse outcomes among women with obesity.³ We observed a consistent trend of increased PTB among women with high BMI. This upward trend of PTB among obese women also was observed with increasing maternal age. A similar increase in medically induced PTB among obese mothers has been previously reported in a study from Salihu et al¹; however, in this present study we did not differentiate between medically induced or spontaneous PTB. In the 2009 Salihu et al study, it was reported that obese teenage mothers had an increased risk for medically induced PTB and a lower risk for spontaneous PTB.¹ In our study, we found that older obese mothers had increased odds for a PTB, whereas obese mothers younger than 20 years had lower odds of PTB. Likewise, in this present study, we differentiate between medically induced or spontaneous PTB. This differentiation of PTB shows that obese mothers have a higher risk association for both spontaneous and medically induced PTB.

Consistent with previous reports, we found that a racial disparity did exist in the occurrence of PTB.^1–5 Few studies have assessed the association between obesity and PTB and racial disparities that exists within this association. Marshall et al in a 2013 retrospective cohort study of birth records linked the effect of obesity on perinatal outcomes such as preeclampsia.⁴ Although PTB is a complication from preeclampsia, Marshall et al did not assess the racial disparity of the association of PTB among obese women.^1–4,7,8 The stratification of PTB (spontaneous and medically induced) in our present study provides relevant information regarding the racial disparity that exists within the association of PTB and obesity. Non-Hispanic Black obese women were shown to have a greater risk association for both spontaneous and medically induced PTB. This suggests that although non-Hispanic Black women have been known to have higher morbidity that may lead to medically induced PTB, black obese women are still at higher risk for spontaneous PTB. Obesity in non-Hispanic Black women is associated with a high risk of spontaneous and medically induced PTB. Early prenatal care was found to contribute to a reduced occurrence of PTB and reduced odds of PTB among obese women, supporting findings from Lhila and Long in 2011.¹²

In this study, we found that maternal age increased the odds of PTB, with older obese mothers showing greater odds of PTB. Obese mothers who did not initiate prenatal care within the first trimester had the highest likelihood for PTB. Other maternal characteristics that increased the risk of PTB among obese women were method of delivery and mother’s level of education. Obesity has long been associated with adverse outcomes in pregnancy. Non-Hispanic Black women have been reported to exhibit higher rates of obesity than their non-Hispanic White and Hispanic counterparts. Comorbidities that increase the risk of PTB also have been shown to disproportionately affect Black women.^10,11 This may explain why we observed the highest odds of PTB among obese Black women. Social determinants of health (SDoH) characteristics have been shown to drive disparities in birth outcomes.^12–14 Mothers’ education and healthcare coverage were found to be the major SDoH factors that determine racial disparities in birth outcomes.¹² Healthcare coverage determines access to health services and reflects economic status. Non-Hispanic Black women with less access to high-quality healthcare services may be less likely to seek out appropriate prenatal care. Lack of access to care earlier in pregnancy could contribute to higher adverse outcomes among obese non-Hispanic Black women.

There are few limitations of this study. Variables related to lifestyle choices, socioeconomic status, and SDoH have not been included in this study because these were not available in the database. Comorbidities such as hypertension or diabetes mellitus may have confounded our findings in that these variables, which also independently predict PTB, are likely to occur more in obese women than in those with normal BMI.^12–22 Information on prepregnancy and gestational diabetes mellitus and hypertension were not available for all of the states in the United States and were therefore excluded from this study. The major strengths in this study are the large sample size as well as the subclassification of obesity in subgroups. The study also defined PTB in a stratified method. This categorizing of obesity and PTB allowed us to observe dose-dependent changes.

In this study, we found that obesity affected non-Hispanic Black women more compared with their non-Hispanic White and Hispanic counterparts when considering the outcome of PTB. Our work further contributes to the growing knowledge of the existence of health disparities among Black populations. To address this negative health disparity trend, deliberate prepregnancy interventions are required. These encompass a multiprong approach to address socioeconomic issues and other factors that may lead to obesity and related comorbid conditions in non-Hispanic Blacks that predispose non-Hispanic Black women to worse outcomes during pregnancy.

Key Points.

• A higher prevalence of preterm birth exists among young mothers (younger than 20 years) and older mothers (40 years old and older).

• Compared with non-Hispanic White women, non-Hispanic Black women had the highest associations between race, body mass index, and preterm birth.

• Maternal factors, such as age and initiation of their prenatal care, could play a crucial role in deteriming preterm birth among obese women.