[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 34.236.145.124. Please contact the publisher to request reinstatement.
[Skip to Content Landing]
Table 1.  
Severe Maternal Morbidity
Severe Maternal Morbidity
Table 2.  
Demographic and Pregnancy Characteristics by Prepregnancy Body Mass Index, Singleton Births, Washington State, 2004-2013a
Demographic and Pregnancy Characteristics by Prepregnancy Body Mass Index, Singleton Births, Washington State, 2004-2013a
Table 3.  
Labor and Delivery Characteristics by Prepregnancy Body Mass Index, Singleton Births, Washington State, 2004-2013a
Labor and Delivery Characteristics by Prepregnancy Body Mass Index, Singleton Births, Washington State, 2004-2013a
Table 4.  
Mortality and Severe Maternal Morbidity (Per 10 000 Singleton Births) by Prepregnancy Body Mass Index, Washington State, 2004-2013a
Mortality and Severe Maternal Morbidity (Per 10 000 Singleton Births) by Prepregnancy Body Mass Index, Washington State, 2004-2013a
Table 5.  
Adjusted Odds Ratios and Adjusted Rate Differences Per 10 000 Births for Severe Maternal Morbidity by Prepregnancy Body Mass Index, Singleton Births, Washington State, 2004-2013a
Adjusted Odds Ratios and Adjusted Rate Differences Per 10 000 Births for Severe Maternal Morbidity by Prepregnancy Body Mass Index, Singleton Births, Washington State, 2004-2013a
Supplement 2.

eTable 1. Sources of Data

eTable 2. Severe Maternal Morbidity Conditions and ICD-9-CM Codes

eTable 3. Unadjusted Association Between Pre-Pregnancy BMI and Severe Maternal Morbidity, Washington State, USA, 2004-2013

eTable 4. Sensitivity Analysis: Adjusted Odds Ratios and Adjusted Rate Differences per 10 000 Births for Severe Maternal Morbidity by Pre-Pregnancy Body Mass Index (BMI), Singleton Births, Adjusted for Chronic Hypertension and Pre-Pregnancy Diabetes Mellitus, Washington State, USA, 2004-2013

eTable 5. Demographic Characteristics of Women With Missing Body Mass Index (BMI)

eTable 6. Severe Maternal Morbidity Among Women With Missing Body Mass Index (BMI)

eTable 7. Sensitivity Analysis: Adjusted Odds Ratios and Adjusted Rate Differences per 10 000 Births for Severe Maternal Morbidity by Pre-Pregnancy Body Mass Index (BMI), With Multiple Imputation for Missing BMI Values, Singleton Births, Washington State, USA, 2004-2013

eTable 8. Sensitivity Analysis: Adjusted Odds Ratios and Adjusted Rate Differences per 10 000 Births for Severe Maternal Morbidity by Pre-Pregnancy Body Mass Index (BMI), With Adjustment for Lower and Higher Than Recommended Gestational Weight Gain, Singleton Births, Washington State, USA, 2004-2013

eTable 9. Sensitivity Analysis: Adjusted Odds Ratios and Adjusted Rate Differences per 10 000 Births for Severe Maternal Morbidity by Pre-Pregnancy Body Mass Index (BMI), Singleton Births, Also Adjusted for Drug Dependence, Washington State, USA, 2004-2013

1.
Fryar  CD, Carroll  MD, Ogden  CL. Prevalence of overweight, obesity, and extreme obesity among adults: United States, 1960-1962 through 2011-2012. Centers for Disease Control and Prevention website. https://www.cdc.gov/nchs/data/hestat/obesity_adult_11_12/obesity_adult_11_12.pdf. September 2014. Accessed April 3, 2017.
2.
Flegal  KM, Kruszon-Moran  D, Carroll  MD, Fryar  CD, Ogden  CL.  Trends in obesity among adults in the United States, 2005 to 2014.  JAMA. 2016;315(21):2284-2291.PubMedGoogle ScholarCrossref
3.
Branum  AM, Kirmeyer  SE, Gregory  ECW.  Prepregnancy body mass index by maternal characteristics and state: data from the birth certificate, 2014.  Natl Vital Stat Rep. 2016;65(6):1-11.PubMedGoogle Scholar
4.
Guelinckx  I, Devlieger  R, Beckers  K, Vansant  G.  Maternal obesity: pregnancy complications, gestational weight gain and nutrition.  Obes Rev. 2008;9(2):140-150.PubMedGoogle ScholarCrossref
5.
Callaway  LK, Prins  JB, Chang  AM, McIntyre  HD.  The prevalence and impact of overweight and obesity in an Australian obstetric population.  Med J Aust. 2006;184(2):56-59.PubMedGoogle Scholar
6.
Dzakpasu  S, Fahey  J, Kirby  RS,  et al.  Contribution of prepregnancy body mass index and gestational weight gain to adverse neonatal outcomes: population-attributable fractions for Canada.  BMC Pregnancy Childbirth. 2015;15:21.PubMedGoogle ScholarCrossref
7.
Scott-Pillai  R, Spence  D, Cardwell  CR, Hunter  A, Holmes  VA.  The impact of body mass index on maternal and neonatal outcomes: a retrospective study in a UK obstetric population, 2004-2011.  BJOG. 2013;120(8):932-939.PubMedGoogle ScholarCrossref
8.
Yogev  Y, Visser  GHA.  Obesity, gestational diabetes and pregnancy outcome.  Semin Fetal Neonatal Med. 2009;14(2):77-84.PubMedGoogle ScholarCrossref
9.
Cnattingius  S, Bergström  R, Lipworth  L, Kramer  MS.  Prepregnancy weight and the risk of adverse pregnancy outcomes.  N Engl J Med. 1998;338(3):147-152.PubMedGoogle ScholarCrossref
10.
Aune  D, Saugstad  OD, Henriksen  T, Tonstad  S.  Maternal body mass index and the risk of fetal death, stillbirth, and infant death: a systematic review and meta-analysis.  JAMA. 2014;311(15):1536-1546.PubMedGoogle ScholarCrossref
11.
Schummers  L, Hutcheon  JA, Bodnar  LM, Lieberman  E, Himes  KP.  Risk of adverse pregnancy outcomes by prepregnancy body mass index: a population-based study to inform prepregnancy weight loss counseling.  Obstet Gynecol. 2015;125(1):133-143.PubMedGoogle ScholarCrossref
12.
O’Brien  TE, Ray  JG, Chan  WS.  Maternal body mass index and the risk of preeclampsia: a systematic overview.  Epidemiology. 2003;14(3):368-374.PubMedGoogle ScholarCrossref
13.
Abdollahi  M, Cushman  M, Rosendaal  FR.  Obesity: risk of venous thrombosis and the interaction with coagulation factor levels and oral contraceptive use.  Thromb Haemost. 2003;89(3):493-498.PubMedGoogle Scholar
14.
Bodnar  LM, Catov  JM, Klebanoff  MA, Ness  RB, Roberts  JM.  Prepregnancy body mass index and the occurrence of severe hypertensive disorders of pregnancy.  Epidemiology. 2007;18(2):234-239.PubMedGoogle ScholarCrossref
15.
Durst  JK, Tuuli  MG, Stout  MJ, Macones  GA, Cahill  AG.  Degree of obesity at delivery and risk of preeclampsia with severe features.  Am J Obstet Gynecol. 2016;214(5):651.e1-651.e5.PubMedGoogle ScholarCrossref
16.
Birth data quality technical notes. Washington State Department of Health website. https://www.doh.wa.gov/Portals/1/Documents/5300/TechnicalNotes.pdf. July 2016. Accessed July 26, 2017.
17.
Comprehensive Hospital Abstract Reporting System (CHARS): hospital inpatient discharge database reports 2010-2016. Washington State Department of Health website. https://www.doh.wa.gov/ForPublicHealthandHealthcareProviders/HealthcareProfessionsandFacilities/DataReportingandRetrieval/HospitalInpatientDatabaseCHARS. Accessed July 26, 2017.
18.
Procedure manual for submitting discharge data UB-40 and 837I 5010. Washington State Department of Health website. https://www.doh.wa.gov/Portals/1/Documents/5300/CHARS-UB04-5010-CompanionGuide-R5.pdf. January 28, 2011 (revised September 21, 2013). Accessed July 26, 2017.
19.
Lydon-Rochelle  MT, Holt  VL, Cárdenas  V,  et al.  The reporting of pre-existing maternal medical conditions and complications of pregnancy on birth certificates and in hospital discharge data.  Am J Obstet Gynecol. 2005;193(1):125-134.PubMedGoogle ScholarCrossref
20.
Lydon-Rochelle  MT, Holt  VL, Nelson  JC,  et al.  Accuracy of reporting maternal in-hospital diagnoses and intrapartum procedures in Washington State linked birth records.  Paediatr Perinat Epidemiol. 2005;19(6):460-471.PubMedGoogle ScholarCrossref
21.
Joseph  KS, Liu  S, Rouleau  J,  et al.  Severe maternal morbidity in Canada, 2003 to 2007: surveillance using routine hospitalization data and ICD-10CA codes.  J Obstet Gynaecol Can. 2010;32(9):837-846.PubMedGoogle ScholarCrossref
22.
Centers for Disease Control and Prevention (CDC). Severe maternal morbidity in the United States. CDC website. https://www.cdc.gov/reproductivehealth/maternalinfanthealth/severematernalmorbidity.html. 2013. Accessed July 26, 2017.
23.
Kramer  MS, Ananth  CV, Platt  RW, Joseph  KS.  US black vs white disparities in foetal growth: physiological or pathological?  Int J Epidemiol. 2006;35(5):1187-1195.PubMedGoogle ScholarCrossref
24.
Martin  JA, Hamilton  BE, Osterman  MJK, Driscoll  AK, Mathews  TJ.  Births: final data for 2015.  Natl Vital Stat Rep. 2017;66(1):1.PubMedGoogle Scholar
25.
VanderWeele  TJ, Mumford  SL, Schisterman  EF.  Conditioning on intermediates in perinatal epidemiology.  Epidemiology. 2012;23(1):1-9.PubMedGoogle ScholarCrossref
26.
Schisterman  EF, Cole  SR, Platt  RW.  Overadjustment bias and unnecessary adjustment in epidemiologic studies.  Epidemiology. 2009;20(4):488-495.PubMedGoogle ScholarCrossref
27.
Joseph  KS.  Incidence-based measures of birth, growth restriction, and death can free perinatal epidemiology from erroneous concepts of risk.  J Clin Epidemiol. 2004;57(9):889-897.PubMedGoogle ScholarCrossref
28.
American College of Obstetricians and Gynecologists.  ACOG Committee opinion no. 548: weight gain during pregnancy.  Obstet Gynecol. 2013;121(1):210-212.PubMedGoogle ScholarCrossref
29.
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.PubMedGoogle ScholarCrossref
30.
Sebire  NJ, Jolly  M, Harris  JP,  et al.  Maternal obesity and pregnancy outcome: a study of 287,213 pregnancies in London.  Int J Obes Relat Metab Disord. 2001;25(8):1175-1182.PubMedGoogle ScholarCrossref
31.
Sattar  N, Clark  P, Holmes  A, Lean  ME, Walker  I, Greer  IA.  Antenatal waist circumference and hypertension risk.  Obstet Gynecol. 2001;97(2):268-271.PubMedGoogle Scholar
32.
WHO, UNICEF, UNFPA, The World Bank and the United Nations Population Division. Trends in maternal mortality: 1990 to 2013: estimates by WHO, UNICEF, UNFPA, The World Bank and the United Nations Population Division. World Health Organization website. http://www.who.int/reproductivehealth/publications/monitoring/maternal-mortality-2013/en/. 2014. Accessed August 1, 2017.
33.
Ogden  CL, Carroll  MD, Fryar  CD, Flegal  KM. Prevalence of obesity among adults and youth: United States, 2011-2014: NCHS data brief no. 219. Centers for Disease Control and Prevention website. https://www.cdc.gov/nchs/data/databriefs/db219.pdf. 2015. Accessed August 1, 2017.
34.
National Center for Chronic Disease Prevention and Health Promotion. BRFSS prevalence and trends data. Centers for Disease Control and Prevention website. https://www.cdc.gov/brfss/brfssprevalence/. Accessed March 15, 2017.
35.
Goldstein  RF, Abell  SK, Ranasinha  S,  et al.  Association of gestational weight gain with maternal and infant outcomes: a systematic review and meta-analysis.  JAMA. 2017;317(21):2207-2225.PubMedGoogle ScholarCrossref
36.
Braekkan  SK, Siegerink  B, Lijfering  WM, Hansen  JB, Cannegieter  SC, Rosendaal  FR.  Role of obesity in the etiology of deep vein thrombosis and pulmonary embolism: current epidemiological insights.  Semin Thromb Hemost. 2013;39(5):533-540.PubMedGoogle ScholarCrossref
Original Investigation
November 14, 2017

Association Between Prepregnancy Body Mass Index and Severe Maternal Morbidity

Author Affiliations
  • 1Department of Obstetrics and Gynaecology, University of British Columbia and the Children’s and Women’s Hospital and Health Centre of British Columbia, Vancouver, Canada
  • 2School of Population and Public Health, University of British Columbia, Vancouver, Canada
  • 3Department of Medicine, University of British Columbia and Women’s Hospital and Health Centre of British Columbia, Vancouver, Canada
JAMA. 2017;318(18):1777-1786. doi:10.1001/jama.2017.16191
Key Points

Question  Is prepregnancy body mass index (BMI) associated with severe maternal morbidity?

Findings  In this cohort study that included 743 630 pregnant women in Washington State between 2004 and 2013, low and high prepregnancy BMI, compared with normal BMI, were significantly associated with increased risk of a composite adverse outcome of severe maternal morbidity or mortality that included maternal death and conditions leading to serious sequelae (eg, adjusted absolute risk increase of 28.8 per 10 000 for underweight and 61.1 per 10 000 for class 3 obese women).

Meaning  Low and high prepregnancy BMI were associated with a statistically significant but small increase in the risk of severe maternal morbidity or mortality.

Abstract

Importance  Although high body mass index (BMI) is associated with adverse birth outcomes, the association with severe maternal morbidity is unclear.

Objective  To examine the association between prepregnancy BMI and severe maternal morbidity.

Design, Setting, and Participants  Retrospective population-based cohort study including all singleton hospital births in Washington State, 2004-2013. Demographic data and morbidity diagnoses were obtained from linked birth certificates and hospitalization files.

Exposures  Prepregnancy BMI (weight in kilograms divided by height in meters squared) categories included underweight (<18.5), normal BMI (18.5-24.9), overweight (25.0-29.9), obesity class 1 (30.0-34.9), obesity class 2 (35.0-39.9), and obesity class 3 (≥40).

Main Outcomes and Measures  Composite severe maternal morbidity or mortality included life-threatening conditions and conditions leading to serious sequelae (eg, amniotic fluid embolism, hysterectomy), complications requiring intensive care unit admission, and maternal death. Logistic regression was used to obtain adjusted odds ratios (ORs) and adjusted rate differences with 95% confidence intervals, adjusted for confounders (eg, maternal age and parity).

Results  Overall, 743 630 women were included in the study (mean age, 28.1 [SD, 6.0] years; 41.4% nulliparous). Prepregnancy BMI was distributed as follows: underweight, 3.2%; normal weight, 47.5%; overweight, 25.8%; obesity class 1, 13.1%; obesity class 2, 6.2%; and obesity class 3, 4.2%. Rates of severe maternal morbidity or mortality were 171.5, 143.2, 160.4, 167.9, 178.3 and 202.9 per 10 000 women, respectively. Adjusted ORs were 1.2 (95% CI, 1.0-1.3) for underweight women; 1.1 (95% CI, 1.1-1.2) for overweight women; 1.1 (95% CI, 1.1-1.2) for women with class 1 obesity; 1.2 (95% CI, 1.1-1.3) for women with class 2 obesity; and 1.4 (95% CI, 1.3-1.5) for women with class 3 obesity compared with women with normal BMI. Absolute risk increases (adjusted rate differences per 10 000 women, compared with women with normal BMI) were 28.8 (95% CI, 12.2-47.2) for underweight women, 17.6 (95% CI, 10.5-25.1) for overweight women, 24.9 (95% CI, 15.7-34.6) for women with class 1 obesity, 35.8 (95% CI, 23.1-49.5) for women with class 2 obesity, and 61.1 (95% CI, 44.8-78.9) for women with class 3 obesity.

Conclusions and Relevance  Among pregnant women in Washington State, low and high prepregnancy BMI, compared with normal BMI, were associated with a statistically significant but small absolute increase in severe maternal morbidity or mortality.

Introduction

Quiz Ref IDThe age-adjusted prevalence of obesity among US women increased from 15.8% in 1960 to 40.4% in 2014.1,2 In 2014, half of pregnant women were either overweight (25.6%) or obese (24.8%).3 Overweight and obesity are associated with elevated risks of preterm birth, large-for-gestational-age live birth, macrosomia, shoulder dystocia, congenital anomalies, birth injury, stillbirth, cerebral palsy, and neonatal and infant death,4-11 while underweight is associated with preterm birth, small-for-gestational-age birth, and neonatal intensive care unit admission.11 Less is known about the association between body mass index (BMI) and potentially life-threatening maternal morbidity. Some studies have shown an association between obesity and pregnancy complications including hypertension in pregnancy and preeclampsia, gestational diabetes, thromboembolism, and cesarean delivery,4,11-15 although 1 study suggested that the overall risk of severe maternal morbidity does not increase with BMI.11

The objective of this study was to examine the association between prepregnancy BMI and severe maternal morbidity or mortality.

Methods

All analyses in this study were performed using publicly accessible deidentified data. An exemption from ethics approval was granted by the Department of Social and Health Services, State of Washington.

Data Sources

Information on all singleton hospital births at 20 to 45 weeks’ gestation in Washington State between January 1, 2004, and December 31, 2013, was obtained from 2 linked data sources: the live birth and fetal death certificates database and the hospitalization database (Comprehensive Hospital Abstract Reporting System [CHARS]). Information obtained from the birth and fetal death certificates was abstracted by trained abstractors using standardized forms (Supplement 1) and included infants’ sex and maternal characteristics such as prepregnancy BMI, age, race, education, marital status, parity, smoking during pregnancy, assisted conception, and year of childbirth. Information on gestational age at delivery, obstetric history (previous infant death, preterm birth, or small-for-gestational-age birth in parous women), labor characteristics (eg, prolonged labor), mode of delivery, gestational hypertension, gestational diabetes, and congenital anomalies was also obtained from birth certificates. CHARS data included up to 9 diagnostic and 9 procedure codes (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM]) related to maternal hospitalization, information on death during hospitalization, the type of health insurance coverage, and intensive care unit admission. Prepregnancy hypertension and prepregnancy diabetes mellitus were identified from both data sources; the condition was deemed present if indicated in at least 1 data set (see eTable 1 in Supplement 2 for details).

Completeness and accuracy of birth certificate and CHARS data was monitored by Washington State Department of Health through annual assessments and consistency checks.16-18 Records flagged with inconsistent or out-of-range entries were addressed systematically through hospital review and correction. The frequency of diagnostic and procedure codes was monitored in annual reports.18 Previous validation studies of the linked data set19,20 showed that for the majority of labor and delivery information, the positive and negative predictive values were greater than 80% and 98%, respectively, in birth certificates and in CHARS, compared with a gold standard of manually abstracted and reabstracted data from medical charts.20 Although some preexisting medical conditions had excellent reporting in birth certificates (eg, prepregnancy diabetes mellitus), a combination of both data sources was found to improve underreporting for most chronic diseases (eg, chronic hypertension).19

BMI was calculated as weight in kilograms divided by height in meters squared, using the mother’s self-reported height and prepregnancy weight (US standard certificate of birth, 2003 revision). BMI data were checked for consistency, and maternal BMI values outside the expected range were flagged as potentially erroneous and rectified.3,16 Maternal prepregnancy BMI was classified in the following categories: normal BMI (18.5-24.9), underweight (<18.5), overweight (25.0-29.9), and obesity class 1 (30.0-34.9), class 2 (35.0-39.9), and class 3 (≥40). Gestational age at delivery was based on ultrasound dating; date of last menstrual period was used for women with missing ultrasound data. Multiple births were excluded, because members within a twin or triplet set could not be identified in the data source.

Outcome Measures

The primary outcome of severe maternal morbidity or mortality was defined as a composite outcome that included life-threatening conditions, conditions leading to serious sequelae, complications requiring intensive care unit admission, and maternal death during the hospitalization for childbirth. Life-threatening conditions and conditions leading to serious sequelae were identified using a list of such conditions previously developed by the Canadian Perinatal Surveillance System.21 This included maternal conditions associated with a high case-fatality rate (eg, amniotic fluid embolism), those involving organ failure (eg, acute renal failure) or leading to adverse sequelae (eg, intracranial hemorrhage), and those requiring high resource use (eg, peripartum hysterectomy) (Table 1; eTable 2 in Supplement 2).

ICD-9-CM codes for potentially lifesaving interventions that the Centers for Disease Control and Prevention recognize as indicative of severe maternal morbidity (eg, invasive hemodynamic monitoring) were also included.22 All severe morbidity was grouped into the following categories: (1) antepartum hemorrhage requiring blood transfusion, eg, placenta previa, placental abruption; (2) respiratory morbidity, eg, obstetric pulmonary embolism, respiratory arrest; (3) thromboembolism, eg, deep venous thrombosis; (4) cerebrovascular morbidity, eg, cerebral venous thromboembolism; (5) cardiac morbidity, eg, acute myocardial infarction; (6) eclampsia; (7) severe postpartum hemorrhage requiring transfusion; (8) sepsis, eg, major puerperal infection; (9) acute renal failure; (10) obstetric shock; (11) disseminated intravascular coagulation; (12) uterine rupture; (13) complications of anesthesia and obstetric intervention, eg, shock due to anesthesia; (14) severe morbidity requiring potentially lifesaving intervention, eg, transfusion, mechanical ventilation, invasive hemodynamic monitoring; and 15) other, eg, acute liver failure (Table 1; eTable 2 in Supplement 2). These categories defined specific severe morbid conditions and were not mutually exclusive.

Statistical Analyses

Logistic regression was used to estimate adjusted odds ratios (ORs) and 95% CIs expressing the association between BMI and severe maternal morbidity or mortality. Associations between BMI and severe maternal morbidity or mortality were adjusted for demographic and prepregnancy characteristics, including maternal education (high school graduation or higher vs less than high school graduation), marital status (single, widowed, or separated vs married or common law), race/ethnicity (Hispanic, African American, Native American, and other vs non-Hispanic white), parity (nulliparous, parity ≥4 vs parity 1-3), assisted conception (no vs yes), smoking during pregnancy (no vs yes), type of health insurance (Medicaid, private vs other), year of birth, and fetal sex (female vs male).

Race/ethnicity was included as a potential confounder in regression analyses because race/ethnicity is associated with both BMI and adverse pregnancy outcomes3,23 and is not in the causal pathway between BMI and these outcomes. Race/ethnicity was recorded in the birth certificate as mother’s self-identified race/ethnicity with the categories non-Hispanic white, black or African American, Native American or Alaska Native, and other categories (eg, Asian Indian, Chinese, Filipino, Japanese, also including “other” as an open-ended category). Hispanic origin was also self-reported and recorded as a separate category (Supplement 1).24

Except for BMI, the proportion of missing values for all covariates in the primary analyses did not exceed 3% (eTable 1 in Supplement 2). Complete-case regression analyses were performed. Estimated associations between BMI and severe maternal morbidity or mortality were also expressed in terms of absolute increases in rates using adjusted rate differences and 95% CIs.

Sensitivity Analyses

In the first sensitivity analysis, regression models were used to adjust for underlying chronic hypertension and prepregnancy diabetes mellitus in addition to other factors. These 2 conditions were not included in the primary analyses because they represent potential mediators in the causal pathway between BMI and severe maternal morbidity. However, the data source did not contain sufficient detail to ascertain the temporal directionality between these chronic conditions and BMI.

Mode of delivery was not included as a potential confounder because cesarean delivery is in the causal pathway between high BMI and severe morbidity and because in some clinical situations cesarean delivery is a sequela of the outcome, eg, severe antepartum hemorrhage is a severe maternal morbidity and an indication for cesarean delivery.25-27

In the second sensitivity analysis, the problem of missing values for prepregnancy BMI (approximately 9% of eligible women) was addressed with multiple imputation (proc MI) using Markov Chain Monte Carlo methods, whereby prepregnancy demographic and clinical factors were used to impute missing BMI values.

In another sensitivity analysis, lower- and higher-than-recommended weight gain during pregnancy were added as covariates to the regression analyses because weight gain is a potentially modifiable risk factor. Weight gain was calculated by subtracting prepregnancy weight (self-reported) from the weight at delivery (measured by a clinician and recorded in birth or fetal death certificates). BMI category–specific weight gain during pregnancy as recommended by the American College of Obstetricians and Gynecologists28 was used to identify women outside the optimal range (12.7 to 18.1 kg for underweight women, 11.3 to 15.9 kg for normal-weight women, 6.8 to 11.3 kg for overweight women, and 5.0 to 9.1 kg for obese women).28

Another sensitivity analysis examined opioid drug dependence as a potential confounder of the association between BMI and severe maternal morbidity or mortality.

All analyses were carried out using SAS version 9.4 (SAS Institute Inc).

Results
Study Population

There were 952 212 live births and stillbirths in Washington State between January 1, 2004, and December 31, 2013. Births that occurred out of state, multiple births, births before 20 weeks’ gestation, and births to women younger than 15 years or older than 60 years were excluded (35 598 mothers [3.7%]), as were births that occurred out of hospital (24 716 mothers [2.6%]) and births that could not be matched with hospital records (64 609 mothers [6.8%]). Women with missing information on BMI (83 659 mothers [8.8%]) were excluded from the primary analyses. The study population included 743 630 women.

Overall, 49.3% of women were overweight or obese (25.8% were overweight, 13.1% obese class 1, 6.2% obese class 2, and 4.2% obese class 3), while 47.5% were of normal BMI and 3.2% were underweight. Underweight women and those with normal BMI were younger and included a higher proportion of women of Hispanic and other race/ethnicity and a higher proportion of nulliparous women. Obese and underweight women had a higher rate of smoking during pregnancy, and obese women had higher rates of preexisting diabetes and chronic hypertension (Table 2). Obese women also had higher rates of cesarean delivery, labor induction, previous cesarean delivery, prior infant death, preterm birth or small-for-gestational-age birth, hypertension in pregnancy, and gestational diabetes (Table 3). Women with class III obesity had higher rates of breech presentation and lower rates of chorioamnionitis and precipitous labor than women with normal BMI.

The rate of severe maternal morbidity or mortality was lowest in women with normal BMI (143.2 per 10 000 women) and highest among women with class 3 obesity (202.9 per 10 000) (Table 4). Crude odds ratios and rate differences are reported in eTable 3 in Supplement 2.

Multivariable Analyses

Adjusted odds ratios showed that severe maternal morbidity or mortality rates were significantly higher among underweight, overweight, and obese women compared with women who had normal BMI (adjusted OR, 1.2 [95% CI, 1.0-1.3] for underweight women; 1.1 [95% CI, 1.1-1.2] for overweight women; 1.1 [95% CI, 1.1-1.2] for women with class 1 obesity; 1.2 [95% CI, 1.1-1.3] for women with class 2 obesity; and 1.4 [95% CI, 1.3-1.5] for women with class 3 obesity (Table 5).

Absolute increases in rates of severe maternal morbidity or mortality were small, with adjusted rate differences of 28.8 (95% CI, 12.2-47.2) per 10 000 for underweight women; 17.6 (95% CI, 10.5-25.1) per 10 000 for overweight women; 24.9 (95% CI, 15.7-34.6) per 10 000 for women with class 1 obesity; 35.8 (95% CI, 23.1-49.5) per 10 000 for women with class 2 obesity; and 61.1 (95% CI, 44.8-78.9) per 10 000 for women with class 3 obesity (Table 5).

Compared with women with normal BMI, underweight women had significantly higher rates of antepartum hemorrhage and acute renal failure and were more likely to receive potentially lifesaving interventions, while overweight women had higher rates of acute renal failure (Table 5). Women with class 1 obesity had statistically significantly higher rates of thromboembolism, cerebrovascular morbidity, sepsis, acute renal failure, and complications of obstetric interventions, while women with class 2 obesity had significantly higher rates of respiratory morbidity, eclampsia, sepsis, acute renal failure, and complications of obstetric interventions (Table 5). Women with class 3 obesity had higher rates of respiratory morbidity (including pulmonary embolism [Table 1]), thromboembolism, cerebrovascular morbidity, cardiac morbidity, eclampsia, sepsis, acute renal failure, complications of obstetric interventions, and ICU admission (Table 5). On the other hand, women with class 3 obesity had significantly lower rates of severe antepartum and postpartum hemorrhage requiring transfusion (Table 5).

Sensitivity Analyses

Although the association between BMI and severe maternal morbidity or mortality was attenuated after additional adjustment for chronic hypertension and prepregnancy diabetes mellitus, it remained statistically significant. The largest changes were in the associations between class 3 obesity and cardiac morbidity (adjusted OR decreased from 3.5 [95% CI, 2.5-5.0] to 2.0 [95% CI, 1.4-2.9]; adjusted rate difference, 3.8 [95% CI, 1.4-7.3] per 10 000) and acute renal failure (adjusted OR decreased from 3.9 [95% CI, 2.0-7.7] to 2.5 [95% CI, 1.2-5.2]; adjusted rate difference, 1.7 [95% CI, 0.3-4.5] per 10 000) (eTable 4 in Supplement 2).

Compared with the women included in the primary analyses, women with missing BMI were older, included higher proportions of women who were African American, Hispanic, and other race/ethnicity; higher proportions of women with low education and women with Medicaid insurance; and lower proportions of married women and nulliparous women (eTable 5 in Supplement 2). Women with missing BMI had higher rates of severe maternal morbidity or mortality (eTable 6 in Supplement 2). Results of logistic regression after multiple imputation for missing BMI were similar to results from the primary analyses, except for the association between class 3 obesity and antepartum hemorrhage, which was no longer statistically significant (adjusted OR, 0.7 [95% CI, 0.4-1.1]; adjusted rate difference, −2.2 [95% CI, −3.8 to 0.6] per 10 000) (eTable 7 in Supplement 2).

Approximately 47% of women had a higher-than-recommended weight gain during pregnancy, whereas 19% had a lower-than-recommended weight gain. Adjusting for low and high gestational weight gain did not substantially alter the associations between BMI and severe maternal morbidity or mortality (eTable 8 in Supplement 2). Additional adjustment for opioid drug dependence yielded results similar to those from the primary analyses (eTable 9 in Supplement 2).

Discussion

Quiz Ref IDIn this study of pregnant women in Washington State, low and high prepregnancy BMI, compared with normal BMI, were associated with a statistically significant but small absolute increase in severe maternal morbidity or mortality. Rates of severe maternal morbidity or mortality showed a dose-response pattern among women with above-normal BMI, with adjusted odds ratios increasing with BMI from normal to class 3 obesity compared with women with normal prepregnancy BMI.Quiz Ref ID Underweight women had an increased risk of antepartum and postpartum hemorrhage with blood transfusion and renal failure and were more likely to require a potentially lifesaving intervention. Women with BMI in the overweight category and obese women had significantly higher rates of several different subtypes of severe maternal morbidity. Although relative measures showed higher rates of severe maternal morbidity or mortality among underweight, overweight, and obese women compared with women who had a normal BMI, the absolute risk of severe maternal morbidity or mortality was small.

Two studies have examined the association between BMI and severe maternal morbidity.11,29 Lindquist et al29 carried out a case-control study (with severe maternal morbidity defined as including eclampsia, amniotic fluid embolism, acute fatty liver, peripartum hysterectomy, postpartum hemorrhage with therapy, and uterine rupture) and reported no association between BMI and severe maternal morbidity. A study by Schummers et al11 reported no association between BMI and a composite outcome of severe maternal morbidity or mortality and a significant inverse association between BMI and postpartum hemorrhage requiring interventionQuiz Ref ID. It is possible that anemia, which is more common among underweight women,7 exacerbates the effects of hemorrhage and leads to higher rates of blood transfusion. The association between high BMI and eclampsia is consistent with a previously reported 2- to 4-fold increased incidence of preeclampsia among obese women.4,12,14,15,30,31

This study has several strengths, including its large sample size and the breadth of information on maternal characteristics collected consistently over the study period. This facilitated the analysis of a wide variety of specific clinical conditions.

This study has limitations. First, the observational design precludes causal inferences. Second, despite the large study size, there was insufficient statistical power to assess associations between BMI and rare severe morbidity subtypes and maternal death. This latter concern necessitated the creation of composite morbidity or mortality outcomes. Third, the number of maternal deaths in this study was low compared with reports on US maternal mortality (including deaths during pregnancy and 42 days postpartum).32 The smaller number of deaths was because this study focused on maternal deaths during the delivery hospitalization only.

Fourth, reliance on ICD-9-CM codes meant that not all severely morbid conditions could be captured (eg, use of ICD-9-CM codes precluded the identification of eclampsia superimposed on chronic hypertension). As a result, the reported rates and odds ratios for eclampsia and for the composite severe morbidity or mortality may be underestimated among overweight and obese women. Fifth, errors and omissions in diagnostic coding that are inevitable in large databases may have led to underreporting and nondifferential misclassification and may have biased results toward the null. Sixth, data on income were not available, and results were therefore adjusted for characteristics of socioeconomic status such as marital status, education, race, and type of medical insurance.

Quiz Ref IDSeventh, information on maternal education, race, and BMI was self-reported and potentially inaccurate. Although self-reported BMI has not been validated in US birth certificates, patterns of obesity among women from various ethnic backgrounds described in the National Health and Nutrition Examination Survey are consistent with patterns from birth certificate data,3,33 and patterns of prepregnancy obesity by state obtained from birth certificates are generally consistent with patterns obtained from the Behavioral Risk Factor Surveillance System.33,34 Proportions of women in each BMI category in this study were consistent with BMI data previously reported among mothers in the United States.3,33,34 Rates of gestational weight gain that were less or more than recommended values were also similar to those recently reported.35 Eighth, the study population only included women aged 15 to 60 years with singleton births and with linked birth certificates and hospital delivery data, and these restrictions may have affected the generalizability of the findings. Ninth, information on visceral adiposity, which is not entirely captured by BMI and may be the primary characteristic associated with adverse health outcomes in obese women,36 was not available in the data source.

Conclusions

Among pregnant women in Washington State, low and high prepregnancy BMI, compared with normal BMI, were associated with a statistically significant but small absolute increase in severe maternal morbidity or mortality.

Back to top
Article Information

Corresponding Author: Sarka Lisonkova, MD, PhD, Department of Obstetrics and Gynaecology, Women’s Hospital and Health Centre of British Columbia, Room C403, 4480 Oak St, Vancouver, BC V6H 3V4, Canada (slisonkova@cfri.ca).

Accepted for Publication: October 4, 2017.

Author Contributions: Dr Lisonkova had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Lisonkova, Muraca, Potts, Chan.

Acquisition, analysis, or interpretation of data: Lisonkova, Muraca, Liauw, Skoll, Lim.

Drafting of the manuscript: Lisonkova, Lim.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Lisonkova, Muraca.

Obtained funding: Lisonkova.

Administrative, technical, or material support: Lisonkova, Muraca, Potts, Lim.

Supervision: Lisonkova, Skoll, Lim.

Conflict of Interest Disclosure: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.

Funding/Support: This study was supported by funding from the Canadian Institutes of Health Research (grants APR-126338 and MAH-15445). Dr Lisonkova is supported by a Scholar Award from the Michael Smith Foundation for Health Research.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: We thank K. S. Joseph, MD, PhD (Department of Obstetrics and Gynaecology, University of British Columbia) for his review and suggestions that helped to improve the manuscript. Dr Joseph did not receive any compensation for his contributions.

References
1.
Fryar  CD, Carroll  MD, Ogden  CL. Prevalence of overweight, obesity, and extreme obesity among adults: United States, 1960-1962 through 2011-2012. Centers for Disease Control and Prevention website. https://www.cdc.gov/nchs/data/hestat/obesity_adult_11_12/obesity_adult_11_12.pdf. September 2014. Accessed April 3, 2017.
2.
Flegal  KM, Kruszon-Moran  D, Carroll  MD, Fryar  CD, Ogden  CL.  Trends in obesity among adults in the United States, 2005 to 2014.  JAMA. 2016;315(21):2284-2291.PubMedGoogle ScholarCrossref
3.
Branum  AM, Kirmeyer  SE, Gregory  ECW.  Prepregnancy body mass index by maternal characteristics and state: data from the birth certificate, 2014.  Natl Vital Stat Rep. 2016;65(6):1-11.PubMedGoogle Scholar
4.
Guelinckx  I, Devlieger  R, Beckers  K, Vansant  G.  Maternal obesity: pregnancy complications, gestational weight gain and nutrition.  Obes Rev. 2008;9(2):140-150.PubMedGoogle ScholarCrossref
5.
Callaway  LK, Prins  JB, Chang  AM, McIntyre  HD.  The prevalence and impact of overweight and obesity in an Australian obstetric population.  Med J Aust. 2006;184(2):56-59.PubMedGoogle Scholar
6.
Dzakpasu  S, Fahey  J, Kirby  RS,  et al.  Contribution of prepregnancy body mass index and gestational weight gain to adverse neonatal outcomes: population-attributable fractions for Canada.  BMC Pregnancy Childbirth. 2015;15:21.PubMedGoogle ScholarCrossref
7.
Scott-Pillai  R, Spence  D, Cardwell  CR, Hunter  A, Holmes  VA.  The impact of body mass index on maternal and neonatal outcomes: a retrospective study in a UK obstetric population, 2004-2011.  BJOG. 2013;120(8):932-939.PubMedGoogle ScholarCrossref
8.
Yogev  Y, Visser  GHA.  Obesity, gestational diabetes and pregnancy outcome.  Semin Fetal Neonatal Med. 2009;14(2):77-84.PubMedGoogle ScholarCrossref
9.
Cnattingius  S, Bergström  R, Lipworth  L, Kramer  MS.  Prepregnancy weight and the risk of adverse pregnancy outcomes.  N Engl J Med. 1998;338(3):147-152.PubMedGoogle ScholarCrossref
10.
Aune  D, Saugstad  OD, Henriksen  T, Tonstad  S.  Maternal body mass index and the risk of fetal death, stillbirth, and infant death: a systematic review and meta-analysis.  JAMA. 2014;311(15):1536-1546.PubMedGoogle ScholarCrossref
11.
Schummers  L, Hutcheon  JA, Bodnar  LM, Lieberman  E, Himes  KP.  Risk of adverse pregnancy outcomes by prepregnancy body mass index: a population-based study to inform prepregnancy weight loss counseling.  Obstet Gynecol. 2015;125(1):133-143.PubMedGoogle ScholarCrossref
12.
O’Brien  TE, Ray  JG, Chan  WS.  Maternal body mass index and the risk of preeclampsia: a systematic overview.  Epidemiology. 2003;14(3):368-374.PubMedGoogle ScholarCrossref
13.
Abdollahi  M, Cushman  M, Rosendaal  FR.  Obesity: risk of venous thrombosis and the interaction with coagulation factor levels and oral contraceptive use.  Thromb Haemost. 2003;89(3):493-498.PubMedGoogle Scholar
14.
Bodnar  LM, Catov  JM, Klebanoff  MA, Ness  RB, Roberts  JM.  Prepregnancy body mass index and the occurrence of severe hypertensive disorders of pregnancy.  Epidemiology. 2007;18(2):234-239.PubMedGoogle ScholarCrossref
15.
Durst  JK, Tuuli  MG, Stout  MJ, Macones  GA, Cahill  AG.  Degree of obesity at delivery and risk of preeclampsia with severe features.  Am J Obstet Gynecol. 2016;214(5):651.e1-651.e5.PubMedGoogle ScholarCrossref
16.
Birth data quality technical notes. Washington State Department of Health website. https://www.doh.wa.gov/Portals/1/Documents/5300/TechnicalNotes.pdf. July 2016. Accessed July 26, 2017.
17.
Comprehensive Hospital Abstract Reporting System (CHARS): hospital inpatient discharge database reports 2010-2016. Washington State Department of Health website. https://www.doh.wa.gov/ForPublicHealthandHealthcareProviders/HealthcareProfessionsandFacilities/DataReportingandRetrieval/HospitalInpatientDatabaseCHARS. Accessed July 26, 2017.
18.
Procedure manual for submitting discharge data UB-40 and 837I 5010. Washington State Department of Health website. https://www.doh.wa.gov/Portals/1/Documents/5300/CHARS-UB04-5010-CompanionGuide-R5.pdf. January 28, 2011 (revised September 21, 2013). Accessed July 26, 2017.
19.
Lydon-Rochelle  MT, Holt  VL, Cárdenas  V,  et al.  The reporting of pre-existing maternal medical conditions and complications of pregnancy on birth certificates and in hospital discharge data.  Am J Obstet Gynecol. 2005;193(1):125-134.PubMedGoogle ScholarCrossref
20.
Lydon-Rochelle  MT, Holt  VL, Nelson  JC,  et al.  Accuracy of reporting maternal in-hospital diagnoses and intrapartum procedures in Washington State linked birth records.  Paediatr Perinat Epidemiol. 2005;19(6):460-471.PubMedGoogle ScholarCrossref
21.
Joseph  KS, Liu  S, Rouleau  J,  et al.  Severe maternal morbidity in Canada, 2003 to 2007: surveillance using routine hospitalization data and ICD-10CA codes.  J Obstet Gynaecol Can. 2010;32(9):837-846.PubMedGoogle ScholarCrossref
22.
Centers for Disease Control and Prevention (CDC). Severe maternal morbidity in the United States. CDC website. https://www.cdc.gov/reproductivehealth/maternalinfanthealth/severematernalmorbidity.html. 2013. Accessed July 26, 2017.
23.
Kramer  MS, Ananth  CV, Platt  RW, Joseph  KS.  US black vs white disparities in foetal growth: physiological or pathological?  Int J Epidemiol. 2006;35(5):1187-1195.PubMedGoogle ScholarCrossref
24.
Martin  JA, Hamilton  BE, Osterman  MJK, Driscoll  AK, Mathews  TJ.  Births: final data for 2015.  Natl Vital Stat Rep. 2017;66(1):1.PubMedGoogle Scholar
25.
VanderWeele  TJ, Mumford  SL, Schisterman  EF.  Conditioning on intermediates in perinatal epidemiology.  Epidemiology. 2012;23(1):1-9.PubMedGoogle ScholarCrossref
26.
Schisterman  EF, Cole  SR, Platt  RW.  Overadjustment bias and unnecessary adjustment in epidemiologic studies.  Epidemiology. 2009;20(4):488-495.PubMedGoogle ScholarCrossref
27.
Joseph  KS.  Incidence-based measures of birth, growth restriction, and death can free perinatal epidemiology from erroneous concepts of risk.  J Clin Epidemiol. 2004;57(9):889-897.PubMedGoogle ScholarCrossref
28.
American College of Obstetricians and Gynecologists.  ACOG Committee opinion no. 548: weight gain during pregnancy.  Obstet Gynecol. 2013;121(1):210-212.PubMedGoogle ScholarCrossref
29.
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.PubMedGoogle ScholarCrossref
30.
Sebire  NJ, Jolly  M, Harris  JP,  et al.  Maternal obesity and pregnancy outcome: a study of 287,213 pregnancies in London.  Int J Obes Relat Metab Disord. 2001;25(8):1175-1182.PubMedGoogle ScholarCrossref
31.
Sattar  N, Clark  P, Holmes  A, Lean  ME, Walker  I, Greer  IA.  Antenatal waist circumference and hypertension risk.  Obstet Gynecol. 2001;97(2):268-271.PubMedGoogle Scholar
32.
WHO, UNICEF, UNFPA, The World Bank and the United Nations Population Division. Trends in maternal mortality: 1990 to 2013: estimates by WHO, UNICEF, UNFPA, The World Bank and the United Nations Population Division. World Health Organization website. http://www.who.int/reproductivehealth/publications/monitoring/maternal-mortality-2013/en/. 2014. Accessed August 1, 2017.
33.
Ogden  CL, Carroll  MD, Fryar  CD, Flegal  KM. Prevalence of obesity among adults and youth: United States, 2011-2014: NCHS data brief no. 219. Centers for Disease Control and Prevention website. https://www.cdc.gov/nchs/data/databriefs/db219.pdf. 2015. Accessed August 1, 2017.
34.
National Center for Chronic Disease Prevention and Health Promotion. BRFSS prevalence and trends data. Centers for Disease Control and Prevention website. https://www.cdc.gov/brfss/brfssprevalence/. Accessed March 15, 2017.
35.
Goldstein  RF, Abell  SK, Ranasinha  S,  et al.  Association of gestational weight gain with maternal and infant outcomes: a systematic review and meta-analysis.  JAMA. 2017;317(21):2207-2225.PubMedGoogle ScholarCrossref
36.
Braekkan  SK, Siegerink  B, Lijfering  WM, Hansen  JB, Cannegieter  SC, Rosendaal  FR.  Role of obesity in the etiology of deep vein thrombosis and pulmonary embolism: current epidemiological insights.  Semin Thromb Hemost. 2013;39(5):533-540.PubMedGoogle ScholarCrossref
×