Abstract
Objective:
To evaluate the role of pre-pregnancy obesity as a mediator in the association between race-ethnicity and severe maternal morbidity.
Methods:
We conducted an analysis on a population-based retrospective cohort study using 2010–2014 birth records linked with hospital discharge data in New York City. A multivariable logistic regression mediation model on a subgroup of the sample consisting of normal weight and obese women (n=409,021) calculated the mediation effect of obesity in the association between maternal race-ethnicity and severe maternal morbidity, and the residual effect not mediated by obesity. A sensitivity analysis was conducted excluding the severe maternal morbidity cases due to blood transfusion.
Results:
Among 591, 455 live births, we identified 15,158 cases of severe maternal morbidity (256.3 per 10,000 deliveries). The severe maternal morbidity rate among obese women was higher than that of normal weight women (342 vs. 216 per 10,000 deliveries). Black women had a severe maternal morbidity rate nearly three times higher than White women (420 vs. 146 per 10,000 deliveries) and the severe maternal morbidity rate among Latinas was nearly twice that of White women (285 per 10,000 deliveries). Among women with normal or obese BMI only (n=409,021), Black race was strongly associated with severe maternal morbidity (aOR 3.02, 95%CI 2.88–3.17) but the obesity-mediated effect represented only 3.2% of the total association (aOR 1.03, 95%CI 1.02–1.05). Latina ethnicity was also associated with severe maternal morbidity (aOR 2.01, 95%CI 1.90–2.12) and the obesity-mediated effect was similarly small: 3.4% of the total association (aOR 1.02, 95%CI 1.01–1.03). In a sensitivity analysis excluding blood transfusion, severe maternal morbidity cases found a higher mediation effect of obesity in the association with Black race and Latina ethnicity (15.3% and 15.2% of the total association, respectively).
Conclusion:
Our findings indicate that prepregnancy obesity, a modifiable factor, is a limited driver of racial-ethnic disparities in overall severe maternal morbidity.
Précis
While both Black race and Latina ethnicity were strongly associated with severe maternal morbidity, obesity only mediated 3% of the observed associations suggesting other factors deserve more attention.
Introduction
Maternal mortality in the United States is the highest among developed countries at 17.4 per 100,000 live births1. Severe maternal morbidity occurs 100 times more frequently than mortality and is an important indicator for studying obstetric quality and preventing death. Severe maternal morbidity rose from 49.5 to 144.0 per 10,000 delivery hospitalizations between 1993–20142. The high burden of severe maternal outcomes is linked to racial disparities in maternal health. Black women are 3–4 times more likely to die in childbirth as compared to White women and 60% of these deaths are preventable3,4. They are also 2–3 times as likely to experience severe maternal morbidity5,6. New York City (NYC) has a higher severe maternal morbidity rate than the national average (241.0 per 10,000 delivery hospitalizations, 2008–2014), with Black-White racial disparities mirroring national trends7.
Racial disparities in severe maternal outcomes are multifactorial and not well understood. Conventional wisdom emphasizes differences in clinical comorbidities. For example, Black women have twice the rate of prepregnancy obesity (BMI > =30 kg/m2) as that of White women (40% versus 20%)8,9. Some evidence supports an independent association between obesity and severe maternal morbidity10–17.
Given the scarce existing data on how severe maternal morbidity risks operate within the context of disparities, more robust analytic approaches are needed in order to understand their roles on the causal pathway. Our objective was to elucidate the contribution of prepregnancy obesity to the association between race-ethnicity and severe maternal morbidity.
Methods
We performed a retrospective cohort study using birth records linked with hospital discharge abstract data for all delivery hospitalizations in New York City from 01/2010 to 12/2014. The New York State Department of Health and Mental Health (DOHMH) conducted the data linkage. Approximately 125,000 live births are registered annually in NYC by the DOHMH’s Bureau of Vital Statistics (VS) which issues birth certificates with detailed confidential demographic and medical data of high quality and reliability18. The birth certificate data include mother’s age, self-reported race and ethnicity, place of birth, education level, health insurance information, and details regarding the pregnancy such as parity, prenatal care, and delivery. A list of all data elements contained in the NYC birth certificate is available at nyc.gov/html/doh/html/data/vs-summary.shtml19. Almost all (99%) of deliveries in NYC occur in hospitals19. The mandatory New York State Department of Health Statewide Planning and Research Cooperative System (SPARCS) monitors and reports on all inpatient hospital discharges, including details regarding length of stay as well as diagnosis and procedure codes19. Full SPARCS data elements can be found at: www.health.ny.gov/statistics/sparcs/sysdoc/iptable.htm. New York State Department of Health performs audits to check data quality on a quarterly basis. The SPARCS data quantity, quality, and validation protocol can be found at: https://www.health.ny.gov/statistics/sparcs/training/docs/sparcs_data_quantity_quality_protocol_final.pdf. The match rate in the linked dataset was approximately 98%. Multiple births were counted as one delivery.
Maternal race-ethnicity was the independent variable of interest in our analysis. We created this variable by combining self-reported race with self-reported Hispanic ethnicity from the birth certificate and coded it in 5 categories: non-Hispanic White (we refer to as White), non-Hispanic Black (we refer to as Black), Latina, Asian, and other. Other race was a formal prespecified category on the birth certificate. Birth certificate data regarding maternal race and Hispanic ethnicity has been previously validated20–22. Delivery hospitalizations were ascertained from International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis, procedure, and diagnosis-related group delivery codes from SPARCS23. The binary outcome of composite severe maternal morbidity was based on a published algorithm from the Centers for Disease Control and Prevention and defined as per specific ICD-9-CM billing codes and hospital length of stay24,25. Prepregnancy BMI (kg/m2) is reliably reported on the NYC birth certificate26 and was used to create categorical BMI as the mediator in the analyses, defined as per routine clinical cutoffs: underweight <18.5, normal weight 18.5–24.9, overweight 25–29.9, obese >=30 27. We excluded improbable prepregnancy BMIs of less than 12 or greater than 55 kg/m2 (n=4,262).
Demographic, social, and medical characteristics of participants were described and differences between women who did and did not experience severe maternal morbidity ascertained using chi-squared tests of independence or Fisher’s exact tests for categorical variables, and two sample t-tests or ANOVA for continuous variables (p<0.05). Differences between participants by maternal race-ethnicity and BMI categories were similarly described.
Multivariable logistic regression models were constructed including the mediation effect of maternal BMI categories in the association between maternal race-ethnicity and severe maternal morbidity. As our aim was to isolate the effect of obesity as compared to normal weight, the mediation analysis was conducted on a subgroup of the entire study sample which consisted of women who were either normal weight or obese (n=409,021) 2,18,23,25,26,28–30,24,27–29. Using normal BMI as the reference, a path analysis multivariable logistic regression mediation model decomposed the total effect of maternal race-ethnicity on severe maternal morbidity into an indirect obesity-mediated effect and a residual effect not mediated by obesity 30,31. Potential variables for inclusion in the analysis were identified by constructing a directed acyclic graph of the main effects, i.e. the causal effect of the exposure (race-ethnicity) on the outcome (severe maternal morbidity) and of the mediator (obesity) on the outcome28, and did not include factors on the causal pathways between them 17. Prenatal care utilization and certain pre-existing maternal conditions, such as chronic hypertension or diabetes, were not considered for inclusion in the analysis as they are potential intermediates in the causal pathway between maternal race-ethnicity and severe maternal morbidity.
The path analysis model conceptualized a counterfactual approach which may theoretically be formulated as the response to the following question: “What would be the risk of [ severe maternal morbidity]associated with Black race if Black women had the same probability of being obese as White women?”. The indirect obesity-mediated effect of maternal race–ethnicity on the risk of severe maternal morbidity was calculated from the regression coefficients obtained via adjusted logistic regression models and was expressed as a percentage of the total effect. We conducted similar analyses for each race-ethnicity category with non-Hispanic White race as the reference. Given a high reported false positive rate of severe maternal morbidity with the ICD-9 administrative codes for a blood transfusion32, likely due to the coding of 1–2 unit transfusions, a sensitivity analysis was conducted using a severe maternal morbidity composite outcome excluding blood transfusion. In the overall dataset, missing data were infrequent; variables with the highest rates were prepregnancy BMI (1.6%), and educational level (0.8%). Among the sub-group of normal weight and obese women included in the model, all participants necessarily had values for BMI and missing data for all other variables were omitted from the final analyses. This resulted in 0.2% of participants missing in the final model. All analyses were performed using Stata, V.14.0 SE (Stata Corporation, College Station, Texas, USA) and add-on models developed by Buis30. The investigation was approved by the Institutional Review Boards of the New York City Department of Health and Mental Hygiene, the New York State Department of Health, and the Icahn School of Medicine at Mount Sinai.
Results
Among 591,455 hospital deliveries in 40 NYC hospitals between 2010–2014, we identified 15,158 cases of severe maternal morbidity, a rate of 256.3 per 10,000 deliveries. The severe maternal morbidity rate among obese women was notably higher than among normal weight women (342 vs. 216 per 10,000 deliveries). Black women had the highest rate of severe maternal morbidity (420 per 10,000 deliveries), a rate nearly three times higher than that of White women (146 per 10,000 deliveries), followed by Latina women (285 per 10,000 deliveries) and Asian women (178 per 10,000 deliveries). Women who experienced severe maternal morbidity were also more likely to be less than 20 or older than 35, without a high school diploma, unemployed during their pregnancies, have Medicaid insurance, and have pre-existing medical conditions including chronic hypertension and diabetes . Severe maternal morbidity also occurred more often among women who were multiparous with prior cesarean deliveries. In the current pregnancy, women who experienced severe maternal morbidity were more likely to have a multiple gestation, late initiation of prenatal care, hypertensive disorders of pregnancy, gestational diabetes, abnormal placentation, and have conceived via assisted reproductive technologies (ART) (Table 1).
Table 1:
Characteristics of women experiencing severe maternal morbidity and those not experiencing severe maternal morbidity
| Total n=591,455 n(%) |
Did not experience SMM n=576,297 (97.4%) n(%) |
Experienced SMM n=15,158 (2.6%) n(%) |
p* | ||
|---|---|---|---|---|---|
| Prepregnancy body mass-index (kg/m2) [mean ±standard deviation] | 25.5 ±10.8 | 25.5 ±10.8 | 26.9 ±12.6 | <0.001 | |
| Underweight (<18.5) | 33,146(5.7) | 32,426 (5.6) | 720 (4.8) | <0.001 | |
| Normal weight (18.5–24.9) | 314,047(54.1) | 307,255 (53.3) | 6,792 (44.8) | ||
| Overweight (25–29.9) | 137,990(23.8) | 134,207 (23.3) | 3,783(25.0) | ||
| Obese (>=30) | 94,974 (16.4) | 91,727 (15.9) | 3,247(21.4) | ||
| Missing | 11,298 (1.9) | 10,682 (1.9) | 616 (4.1) | ||
| Socio-demographic characteristics | |||||
| Maternal age (years) [mean ±standard deviation] | 29.6 ±6.2 | 29.6 ±6.1 | 29.9 ±6.8 | <0.001 | |
| Maternal age (years) [categorical] | <0.001 | ||||
| <20 | 28,182 (4.7) | 27,237 (4.8) | 945 (6.2) | ||
| 20–34 | 424,020 (71.8) | 414,250 (71.8) | 9,770 (64.5) | ||
| >=35 | 132,217 (22.3) | 128,214 (22.3) | 4,003 (26.4) | ||
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Race | <0.001 | ||||
| White | 185,095 (31.3) | 182,386 (31.7) | 2,709 (17.9) | ||
| Black | 122,067 (20.6) | 116,938 (20.3) | 5,129 (33.8) | ||
| Latina | 177,768 (30.1) | 172,693 (30.0) | 5,075 (33.5) | ||
| Asian | 97,496 (16.5) | 95,765 (16.6) | 1,731 (11.4) | ||
| Other | 1,407 (0.2) | 1,365 (0.2) | 42 (0.3) | ||
| Missing | 7,622 (1.3) | 7,150 (1.2) | 472 (3.1) | ||
| Foreign born | 0.208 | ||||
| Yes | 274,568 (46.4) | 267,455 (46.4) | 7,113 (46.9) | ||
| No | 316,887 (53.6) | 308,842 (53.6) | 8,045 (53.1) | ||
| Missing | 0 | 0 | 0 | ||
| Education | <0.001 | ||||
| Less than high school | 126,668 (21.4) | 122,799 (21.3) | 3,869 (25.5) | ||
| High school | 128,974 (21.8) | 125,506 (21.8) | 3,468 (22.9) | ||
| Some college or associates degree | 127,750 (21.6) | 124,285 (21.6) | 3,465 (22.9) | ||
| College degree or greater | 199,101 (33.7) | 195,308 (33.9) | 3,793 (25.0) | ||
| Missing | 8,962 (1.5) | 8,399 (1.5) | 563 (3.7) | ||
| Employed during pregnancy | <0.001 | ||||
| Yes | 286,846 (48.5) | 280,472 (48.7) | 6,374 (42.1) | ||
| No | 296,615 (50.2) | 288,340 (50.0) | 8,275 (54.6) | ||
| Missing | 7,994 (1.3) | 7,485 (1.3) | 509 (3.3) | ||
| Insurance | <0.001 | ||||
| Medicaid | 348,936 (59.0) | 339,058 (58.8) | 9,878 (65.2) | ||
| Private | 221,742 (37.5) | 217,384 (37.7) | 4,358 (28.8) | ||
| Uninsured | 7,839 (1.3) | 7,555 (1.3) | 284 (1.9) | ||
| Other payer | 5902 (1.0) | 5,704 (1.0) | 198 (1.4) | ||
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Medicaid | <0.001 | ||||
| Yes | 348,936 (59.0) | 339,058 (58.8) | 9,878 (65.2) | ||
| No | 235,483 (39.8) | 230,642 (40.0) | 4,840 (31.9) | ||
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Medical history | |||||
| Chronic hypertension | 12,498 (2.1) | 11,552 (2.0) | 946 (6.2) | <0.001 | |
| Missing | 0 | 0 | 0 | ||
| Pre-existing diabetes | 4,968 (0.8) | 4,614 (0.8) | 354 (2.3) | <0.001 | |
| Missing | 0 | 0 | 0 | ||
| Any other notable pre-existing medical condition | 10,092 (1.7) | 9,490 (1.7) | 602 (4.1) | <0.001 | |
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Obstetric history | |||||
| Parity by delivery type | <0.001 | ||||
| Primiparous | 260,160 (44.0) | 253,879 (44.1) | 6,281 (41.4) | ||
| Multiparous without prior cesarean delivery | 236,830 (40.0) | 232,340 (40.3) | 4,490 (29.6) | ||
| Multiparous with prior cesarean delivery | 90,882 (15.4) | 86,648 (15.0) | 4,234 (27.9) | ||
| Missing | 3,583 (0.6) | 3,430 (0.6) | 153 (1.0) | ||
| If parous | |||||
| Prior cesarean delivery | 90,882 (27.7) | 86,648 (27.2) | 4,234 (48.5) | <0.001 | |
| Missing | 0 | 0 | 0 | ||
| Prior preterm delivery | 10,664 (1.8) | 10,127 (1.8) | 537 (3.5) | <0.001 | |
| Missing | 0 | 0 | 0 | ||
| Current pregnancy | |||||
| Conception via ART | 7,710 (1.3) | 7,334 (1.3) | 376 (2.6) | <0.001 | |
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Multiple gestation | 11,539 (2.0) | 10,557 (1.8) | 982 (6.5) | <0.001 | |
| Missing | 0 | 0 | 0 | ||
| Prenatal care initiation [trimester] | <0.001 | ||||
| 1 | 413,115 (72.3) | 403,476 (72.4) | 9,639 (68.2) | ||
| 2 | 121,317 (21.2) | 117,985 (21.2) | 3,332 (23.6) | ||
| 3 | 37,037 (6.5) | 35,868 (6.4) | 1,169 (8.3) | ||
| Missing | 19,986 (3.4) | 18,968 (3.3) | 1,018 (6.7) | ||
| Prenatal visits (n) [mean ±standard deviation] | |||||
| 11.1 ±3.9 | |||||
| 11.1 ±3.9 | 10.6±4.5 | <0.001 | |||
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Hypertensive disorders of pregnancy | <0.001 | ||||
| Normotensive | 550,774 (93.1) | 538,608 (93.5) | 12,166 (80.3) | ||
| Non-proteinuric gestational hypertension | 14,197 (2.4) | 13,721 (2.4) | 476 (3.1) | ||
| Preeclampsia | 26,483 (4.5) | 23,967 (4.2) | 2,516 (16.6) | ||
| Missing | 1 (0.0) | 1 (0.0) | 0 | ||
| Gestational diabetes | 40,944 (6.9) | 39,613 (6.9) | 1,331 (8.8) | <0.001 | |
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Placenta previa or placenta accreta | 4,702 (0.8) | 3,772 (0.7) | 930 (6.1) | <0.001 | |
| Missing | 0 | 0 | 0 | ||
| Tobacco use | 13,309 (2.3) | 12,917 (2.3) | 392 (2.7) | <0.001 | |
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Delivery characteristics | |||||
| Gestational age (weeks) [mean ±standard deviation] | 38.7 ±2.0 | 38.7 ±2.0 | 37.6 ±3.3 | <0.001 | |
| Gestational age (weeks) | <0.001 | ||||
| <28 | 3,301 (0.6) | 2,933 (0.5) | 368 (2.4) | ||
| 28–31 | 4,722 (0.8) | 4,132 (0.7) | 590 (3.9) | ||
| 32–36 | 39,642 (6.7) | 37,139 (6.4) | 2,503 (16.5) | ||
| >=37 | 543,789 (91.9) | 532,092 (92.3) | 11,697 (77.2) | ||
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Noncephalic presentation | 25,861 (4.4) | 24,542 (4.3) | 1,319 (9.0) | <0.001 | |
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Labor onset | 0.02 | ||||
| Spontaneous | 497,276 (85.1) | 484,855 (85.1) | 12,421 (84.4) | ||
| Induced | 87,143 (14.9) | 84,846 (14.9) | 2,297 (15.6) | ||
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Mode of delivery | <0.001 | ||||
| Vaginal | 373,923 (63.2) | 369,587 (64.1) | 4,336 (28.6) | ||
| Vaginal instrumental | 18,942 (3.2) | 18,545 (3.2) | 397 (2.6) | ||
| Cesarean delivery with trial of labor | 66,341 (11.2) | 63,176 (11.0) | 3,165 (20.9) | ||
| Cesarean delivery without trial of labor | 124,572 (21.1) | 117,780 (20.4) | 6,792 (44.8) | ||
| Missing | 7,677 (1.3) | 7,029 (1.3) | 468 (3.1) | ||
| Newborn characteristics | |||||
| Neonatal death | 43 (0.01) | 3 (0.0) | 40 (0.3) | <0.001 | |
| Missing | 0 | 0 | 0 | ||
| Birth weight (g) [mean ±standard deviation] | 3,245 ±563 | 3,251 ±554 | 3,050 ±807 | <0.001 | |
| Missing | 7,036 (1.2) | 6,596 (1.1) | 440 (2.9) | ||
| Delivery facility characteristics | |||||
| Annual delivery volume (quartiles) | <0.001 | ||||
| 1 | 150,448 (25.4) | 145,013 (25.2) | 5,435 (35.9) | ||
| 2 | 147,769 (25.0) | 143,601 (24.9) | 4,168 (27.5) | ||
| 3 | 161,097 (27.2) | 157,534 (27.3) | 3,563 (23.5) | ||
| 4 | 132,141 (22.3) | 130,149 (22.6) | 1,992 (13.1) | ||
| Missing | 0 | 0 | 0 | ||
Chi-squared test of independence or 2 sample t-test
Compared to uncomplicated deliveries, severe maternal morbidity deliveries were more likely to occur preterm or very preterm, have noncephalic presentation, and occur via cesarean. Severe maternal morbidity rates were higher in facilities with lower delivery volumes (Table 1).
Socio-demographic differences between participants by race-ethnicity were remarkable for higher levels of education, lower rates of unemployment and lower rates of Medicaid insurance among White mothers as compared to mothers of all other races and ethnicities (Appendix 1). Black and Latina women were also more likely to be overweight or obese, and all race-ethnicity categories were more likely to have history of cesarean delivery as compared to White women (Appendix 1).
Compared to normal weight women, obese women in our study sample were more likely to be American-born, have a lower level of education, be unemployed, have Medicaid insurance, and have a prior cesarean delivery (Appendix 2). The majority of severe maternal morbidity cases were due to receiving a blood transfusion (202.4 per 10,000 deliveries), followed by emergent hysterectomy (14.1 per 10,000 deliveries), and mechanical ventilation (13.6 per 10,000 deliveries).
In the total population, Black race was associated with composite severe maternal morbidity (OR 2.95, 95%CI 2.82–3.10) and with obesity (OR 3.90, 95%CI 3.83–3.98), as was Latina ethnicity; however to a lesser extent (Appendix 3). Obesity was also associated with composite severe maternal morbidity (OR 1.60, 95%CI 1.53, 1.67) and the association remained significant after adjusting for age, Medicaid insurance, and parity (aOR 1.37, 95%CI 1.31–1.43) (Table 2). In the mediation path analysis adjusted for covariables among the subgroup of women with normal or obese BMI only, Black race remained associated with severe maternal morbidity (aOR 3.02, 95%CI 2.88–3.17); however, the obesity-mediated indirect effect was small (aOR 1.03, 95%CI 1.02–1.05) or 3.2% of the total association (Table 3). Latina ethnicity was also associated with severe maternal morbidity (aOR 2.01, 95%CI 1.90–2.12) and the obesity-mediated effect was 3.4% of the total association. When the overweight BMI category was tested as the mediator, the mediation effects were small and not statistically significant (data not shown).
Table 2:
Association between categorical body mass index and severe maternal morbidity*
| Prepregnancy body mass-index (kg/m2) | Unadjusted OR (95%CI) |
Adjusted† aOR (95%CI) |
|
| Normal weight (18.5–24.9) | 1.00 | 1.00 | |
| Underweight (<18.5) | 1.00 (0.93–1.09) | 0.99 (0.91–1.07) | |
| Overweight (25–29.9) | 1.28 (1.22–1.33) | 1.16 (1.11–1.21) | |
| Obese (>=30) | 1.60 (1.53–1.67) | 1.37 (1.31–1.43) | |
Logistic regression, N=580,157 for unadjusted analysis, N=574,957 for adjusted analysis
Adjusted for: age (continuous), Medicaid insurance (Y/N), parity (primiparous, multiparous without prior cesarean delivery, multiparous with prior cesarean delivery), education (Less than high school, high school, some college or associates degree, college degree or greater)
Table 3.
Obesity-mediated effect of maternal race on severe maternal morbidity*
| Total unadjusted OR (95%CI) | Total adjusted aOR† (95%CI) | Direct aOR† (95%CI) | Obesity-mediated indirect aOR† (95%CI) | |
|---|---|---|---|---|
| N | 409,021 | 405,578 | ||
| Maternal race | ||||
| White | 1.00 | 1.00 | 1.00 | 1.00 |
| Black | 3.02 (2.88–3.17) | 2.90 (2.72–3.09) | 2.80 (2.63–3.00) | 1.03 (1.02–1.05)† |
| Latina | 2.01 (1.90–2.12) | 1.92 (1.81–2.04) | 1.88 (1.77–1.99) | 1.02 (1.01–1.03)†† |
| Asian | 1.25 (1.15–1.35) | 1.20 (1.11–1.29) | 1.21 (1.12–1.30) | 0.99 (0.99–1.00) |
| Other | 2.24 (1.45–3.45) | 2.22 (1.62–3.05) | 2.20 (1.60–3.02) | 1.01 (1.01–1.02) |
Obesity (BMI >= 30kg/m2); reference: normal weight (BMI 18.5–24.9kg/m2)
Adjusted for age (continuous), Medicaid insurance (Yes or No), parity (primiparous, multiparous without prior cesarean delivery, multiparous with prior cesarean delivery), and education level (Less than high school, high school, some college or associates degree, college degree or greater
The sensitivity analysis in which blood transfusion severe maternal morbidity cases were excluded revealed similar disparities by race-ethnicity (Appendix 4). The path analysis found a higher mediation effect of obesity in the association between Black race and the outcome of severe maternal morbidity excluding blood transfusion severe maternal morbidity as compared to the main analysis (15.3% excluding blood transfusion vs 3.2% overall) (table S4). A higher mediation effect was also noted in the sensitivity analysis for Latina ethnicity (15.2% versus 3.4%) (table S4).
Discussion
Black and Latina women had two- to three-fold higher rates of severe maternal morbidity compared to White women even after adjusting for covariables such as Medicaid insurance and maternal education level – social risk factors which are often presumed to be explanatory for racial-ethnic disparities – as well as age and parity. . However, only 3 to 15% of the association between race or Latina -ethnicity and severe maternal morbidity is mediated by prepregnancy obesity, indicating that other factors may be more important when considering interventions to reduce racial disparities in all-cause severe maternal morbidity.
Determinants of severe maternal morbidity can be broadly classified as social or environmental, health system, and individual-level medical characteristics33,34. Individual-level social or environmental factors such as education, insurance, income, and neighborhood poverty account for some of the severe maternal morbidity racial disparity in NYC35,36. Health system factors, such as variation in hospital site and performance, may account for as much as 38–48% of the racial-ethnic disparity in severe maternal morbidity rates in NYC37. Although there is a growing recognition that structural racism contributes to these disparities34, previous research has largely focused on individual-level medical characteristics. Characteristics such as age, specific pre-existing medical conditions, and prior cesarean delivery have been shown to be independent risk factors for severe maternal morbidity; however, there is little evidence supporting their role in racial-ethnic disparities38.
Obesity has been postulated as an important modifiable risk factor contributing to racial-ethnic disparities in maternal health. Our current investigation demonstrates that obesity only mediates a small proportion of the association between race-ethnicity and overall severe maternal morbidity. Notably, prior studies examining the association between race-ethnicity and severe maternal morbidity simply employed BMI as a confounder and not as a mediator 15,39,40. Quantifying the mediation effect allowed for a novel theoretical estimation of disease reduction for a modifiable risk factor for severe maternal morbidity.
The majority of severe maternal morbidity in our cohort was ascertained by blood transfusion. When we excluded morbidity from transfusion alone, we found that obesity is a more important mediator of the association between race-ethnicity and other components of severe maternal morbidity such as hypertensive diseases of pregnancy. These findings are consistent with our previous analysis conducted among immigrant women in France in which we found that 18% of the association between Sub-Saharan maternal place of birth and severe preeclampsia was mediated by obesity41.
Our study had several strengths. The dataset used was large, exhaustive, previously validated, and included a large number of severe maternal morbidity events. The sample was racially and ethnically diverse, and population-based. The definition of severe maternal morbidity was based on a highly utilized CDC algorithm, thus our findings can directly be compared with previous data.
Our study was limited by the reliability of administrative data. Furthermore, severe maternal morbidity is likely underestimated as events occurring after hospital discharge are not included. As noted by others, the CDC definition likely overestimates the rate of severe maternal morbidity due to inclusion of any blood transfusion 32. The sensitivity analysis in which we excluded all cases of blood transfusion addressed this issue.
Although often attributed to personal medical or behavioral factors which disproportionately place blame on women of color, current evidence indicates that racial and ethnic disparities in SMM are often due to structural or health system factors rather than modifiable individual-level factors 6. Further evaluation of other prominent contributors to maternal morbidity such as the health system, clinician, and structural racism will be necessary to help shape future prevention strategies.
Supplementary Material
Acknowledgments
Financial support: This study was funded by the National Institute on Minority Health and Health Disparities (R01MD007651). Ayesha Siddiqui received funding in support of her doctoral research in epidemiology from Sorbonne University, Paris, France. The funders of this study had no role in design and conduct of the study, collection, management, analysis, and interpretation of the data, preparation, review, and approval of the manuscript, or the decision to submit the manuscript for publication.
Footnotes
Financial Disclosure
Elie Azria disclosed that money was paid to him from Lily France. The other authors did not report any potential conflicts of interest.
Each author has confirmed compliance with the journal’s requirements for authorship.
Presented in a virtual poster session at the Annual Pregnancy Meeting of the Society for Maternal-Fetal Medicine on January 29, 2021.
References
- 1.Hoyert DL, Miniño AM. Maternal Mortality in the United States: Changes in Coding, Publication, and Data Release, 2018. Natl Vital Stat Rep. 2020;69(2):1–18. [PubMed] [Google Scholar]
- 2.Severe Maternal Morbidity in the United States. Centers for Disease Control and Prevention; 2020. Accessed January 31, 2020. https://www.cdc.gov/reproductivehealth/maternalinfanthealth/severematernalmorbidity.html#anchor_trends [Google Scholar]
- 3.Mehta PK, Kieltyka L, Bachhuber MA, et al. Racial Inequities in Preventable Pregnancy-Related Deaths in Louisiana, 2011–2016: Obstetrics & Gynecology. 2020;135(2):276–283. doi: 10.1097/AOG.0000000000003591 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Berg CJ, Harper MA, Atkinson SM, et al. Preventability of Pregnancy-Related Deaths: Results of a State-Wide Review. Obstetrics & Gynecology. 2005;106(6):1228–1234. doi: 10.1097/01.AOG.0000187894.71913.e8 [DOI] [PubMed] [Google Scholar]
- 5.Creanga AA, Berg CJ, Ko JY, et al. Maternal Mortality and Morbidity in the United States: Where Are We Now? Journal of Women’s Health. 2014;23(1):3–9. doi: 10.1089/jwh.2013.4617 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Howell EA, Egorova NN, Janevic T, et al. Race and Ethnicity, Medical Insurance, and Within-Hospital Severe Maternal Morbidity Disparities: Obstetrics & Gynecology. 2020;135(2):285–293. doi: 10.1097/AOG.0000000000003667 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Severe Maternal Morbidity Rates in New York City 2013–2014. New York City Department of Health and Mental Hygiene; 2018. Accessed December 26, 2019. https://www1.nyc.gov/assets/doh/downloads/pdf/data/severe-maternal-morbidity-data.pdf
- 8.Robbins CL, Zapata LB, Farr SL, et al. Core state preconception health indicators - pregnancy risk assessment monitoring system and behavioral risk factor surveillance system, 2009. MMWR Surveill Summ. 2014;63(3):1–62. [PubMed] [Google Scholar]
- 9.Singh GK, DiBari JN. Marked Disparities in Pre-Pregnancy Obesity and Overweight Prevalence among US Women by Race/Ethnicity, Nativity/Immigrant Status, and Sociodemographic Characteristics, 2012–2014. Journal of Obesity. 2019;2019:1–13. doi: 10.1155/2019/2419263 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Witteveen T, Zwart JJ, Gast KB, Bloemenkamp KW, van Roosmalen J. Overweight and severe acute maternal morbidity in a low-risk pregnant population in the Netherlands. PloS one. 2013;8(9):e74494. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pallasmaa N, Ekblad U, Gissler M, Alanen A. The impact of maternal obesity, age, preeclampsia and insulin dependent diabetes on severe maternal morbidity by mode of delivery—a register-based cohort study. Archives of gynecology and obstetrics. 2015;291(2):311–318. [DOI] [PubMed] [Google Scholar]
- 12.Lisonkova S, Muraca GM, Potts J, et al. Association Between Prepregnancy Body Mass Index and Severe Maternal Morbidity. JAMA. 2017;318(18):1777. doi: 10.1001/jama.2017.16191 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lindquist A, Noor N, Sullivan E, Knight M. The impact of socioeconomic position on severe maternal morbidity outcomes among women in Australia: a national case-control study. BJOG: Int J Obstet Gy. 2015;122(12):1601–1609. doi: 10.1111/1471-0528.13058 [DOI] [PubMed] [Google Scholar]
- 14.Lindquist A, Knight M, Kurinczuk JJ. Variation in severe maternal morbidity according to socioeconomic position: a UK national case–control study. BMJ Open. 2013;3(6):e002742. doi: 10.1136/bmjopen-2013-002742 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Howland RE, Angley M, Won SH, et al. Determinants of Severe Maternal Morbidity and Its Racial/Ethnic Disparities in New York City, 2008–2012. Matern Child Health J. 2019;23(3):346–355. doi: 10.1007/s10995-018-2682-z [DOI] [PubMed] [Google Scholar]
- 16.Leonard SA, Carmichael SL, Main EK, Lyell DJ, Abrams B. Risk of severe maternal morbidity in relation to prepregnancy body mass index: Roles of maternal co‐morbidities and caesarean birth. Paediatr Perinat Epidemiol Published online May 20, 2019:ppe.12555. doi: 10.1111/ppe.12555 [DOI] [PMC free article] [PubMed]
- 17.Siddiqui A, Azria E, Howell EA, et al. Associations between maternal obesity and severe maternal morbidity: Findings from the French EPIMOMS population‐based study. Paediatr Perinat Epidemiol. 2019;33(1):7–16. doi: 10.1111/ppe.12522 [DOI] [PubMed] [Google Scholar]
- 18.Howland RE, Madsen AM, Toprani A, Gambatese M, Mulready-Ward C, Begier E. How Well Do Birth Records Serve Maternal and Child Health Programs? Birth Registration System Evaluation, New York City, 2008–2011. Matern Child Health J. 2015;19(7):1559–1566. doi: 10.1007/s10995-015-1664-7 [DOI] [PubMed] [Google Scholar]
- 19.Severe Maternal Morbidity in New York City, 2008–2012. New York City Department of Health and Mental Hygiene; 2016. Accessed December 19, 2019. https://www1.nyc.gov/assets/doh/downloads/pdf/data/maternal-morbidity-report-08-12.pdf [Google Scholar]
- 20.Baumeister L, Marchi K, Pearl M, Williams R, Braveman P. The validity of information on “race” and “Hispanic ethnicity” in California birth certificate data. Health Serv Res. 2000;35(4):869–883. [PMC free article] [PubMed] [Google Scholar]
- 21.Howland RE, Tsao T-Y. Evaluating Race and Ethnicity Reported in Hospital Discharge Data and Its Impact on the Assessment of Health Disparities: Medical Care. 2020;58(3):280–284. doi: 10.1097/MLR.0000000000001259 [DOI] [PubMed] [Google Scholar]
- 22.New York City Department of Health and Mental Hygiene (NYC DOHMH) Bureau of Vital Statistics. Mother/Parent Worksheet. Accessed January 24, 2021. https://www1.nyc.gov/assets/doh/downloads/pdf/vs/VR203-web.pdf [Google Scholar]
- 23.Kuklina EV, Whiteman MK, Hillis SD, et al. An Enhanced Method for Identifying Obstetric Deliveries: Implications for Estimating Maternal Morbidity. Matern Child Health J. 2008;12(4):469–477. doi: 10.1007/s10995-007-0256-6 [DOI] [PubMed] [Google Scholar]
- 24.Callaghan WM, MacKay AP, Berg CJ. Identification of severe maternal morbidity during delivery hospitalizations, United States, 1991–2003. American Journal of Obstetrics and Gynecology. 2008;199(2):133.e1-133.e8. doi: 10.1016/j.ajog.2007.12.020 [DOI] [PubMed] [Google Scholar]
- 25.Callaghan WM, Creanga AA, Kuklina EV. Severe Maternal Morbidity Among Delivery and Postpartum Hospitalizations in the United States: Obstetrics & Gynecology. Published online October 2012:1. doi: 10.1097/AOG.0b013e31826d60c5 [DOI] [PubMed] [Google Scholar]
- 26.Deputy NP, Sharma AJ, Bombard JM, et al. Quality of Maternal Height and Weight Data from the Revised Birth Certificate and Pregnancy Risk Assessment Monitoring System: Epidemiology. 2019;30(1):154–159. doi: 10.1097/EDE.0000000000000936 [DOI] [PubMed] [Google Scholar]
- 27.World Health Organization. Body Mass Index - BMI. Accessed February 8, 2021. https://www.euro.who.int/en/health-topics/disease-prevention/nutrition/a-healthy-lifestyle/body-mass-index-bmi [Google Scholar]
- 28.Cole SR, Hernán MA. Fallibility in estimating direct effects. International Journal of Epidemiology. 2002;31(1):163–165. doi: 10.1093/ije/31.1.163 [DOI] [PubMed] [Google Scholar]
- 29.Ananth CV, Schisterman EF. Confounding, causality, and confusion: the role of intermediate variables in interpreting observational studies in obstetrics. American Journal of Obstetrics and Gynecology. 2017;217(2):167–175. doi: 10.1016/j.ajog.2017.04.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Buis ML. Direct and indirect effects in a logit model. Stata J. 2010;10(1):11–29. [PMC free article] [PubMed] [Google Scholar]
- 31.Erikson R, Goldthorpe JH, Jackson M, Yaish M, Cox DR. On class differentials in educational attainment. Proc Natl Acad Sci USA. 2005;102(27):9730–9733. doi: 10.1073/pnas.0502433102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Main EK, Abreo A, McNulty J, et al. Measuring severe maternal morbidity: validation of potential measures. American Journal of Obstetrics and Gynecology. 2016;214(5):643.e1-643.e10. doi: 10.1016/j.ajog.2015.11.004 [DOI] [PubMed] [Google Scholar]
- 33.Howell EA, Brown H, Brumley J, et al. Reduction of Peripartum Racial and Ethnic Disparities: A Conceptual Framework and Maternal Safety Consensus Bundle. Journal of Midwifery & Women’s Health. 2018;63(3):366–376. doi: 10.1111/jmwh.12756 [DOI] [PubMed] [Google Scholar]
- 34.Howell EA. Reducing Disparities in Severe Maternal Morbidity and Mortality: Clinical Obstetrics and Gynecology. Published online January 2018:1. doi: 10.1097/GRF.0000000000000349 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Creanga AA, Bateman BT, Kuklina EV, Callaghan WM. Racial and ethnic disparities in severe maternal morbidity: a multistate analysis, 2008–2010. American Journal of Obstetrics and Gynecology. 2014;210(5):435.e1-435.e8. doi: 10.1016/j.ajog.2013.11.039 [DOI] [PubMed] [Google Scholar]
- 36.Louis JM, Menard MK, Gee RE. Racial and Ethnic Disparities in Maternal Morbidity and Mortality: Obstetrics & Gynecology. 2015;125(3):690–694. doi: 10.1097/AOG.0000000000000704 [DOI] [PubMed] [Google Scholar]
- 37.Howell EA, Egorova NN, Balbierz A, Zeitlin J, Hebert PL. Site of delivery contribution to black-white severe maternal morbidity disparity. American Journal of Obstetrics and Gynecology. 2016;215(2):143–152. doi: 10.1016/j.ajog.2016.05.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Gray KE, Wallace ER, Nelson KR, Reed SD, Schiff MA. Population-Based Study of Risk Factors for Severe Maternal Morbidity. Paediatric and Perinatal Epidemiology. 2012;26(6):506–514. doi: 10.1111/ppe.12011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Grobman WA, Bailit JL, Rice MM, et al. Racial and ethnic disparities in maternal morbidity and obstetric care. Obstet Gynecol. 2015;125(6):1460–1467. doi: 10.1097/AOG.0000000000000735 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Platner M, Ackerman C, Howland RE, et al. 64: The influence of obesity and racial/ethnic disparities on severe maternal morbidity. American Journal of Obstetrics and Gynecology. 2020;222(1):S53–S54. doi: 10.1016/j.ajog.2019.11.080 [DOI] [Google Scholar]
- 41.Siddiqui A, Deneux-Tharaux C, Luton D, et al. Maternal obesity and severe preeclampsia among immigrant women: a mediation analysis. Sci Rep 2020;10(1):5215. doi: 10.1038/s41598-020-62032-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.