Ronit Abramson, Naomi Schwartz, Diana Tordoff

Risk Factors for Maternal Group B Streptococcus Colonization: A Population-Based Case-Control Study in Washington State 2003-2014

Preceptor: Rachel Winer

Background: Maternal colonization with Group B streptococcus (GBS) occurs in 10 to 30% of pregnant women and is the leading infectious cause of mortality in newborns in the United States. This is the first population-based study of multiple risk factors for GBS since the implementation of universal screening guidelines, which recommend screening for all pregnant women between 35-37 weeks gestation.

Methods: We conducted a population-based case-control study to evaluate risk factors for GBS colonization. Washington State birth certificate data (2003-2014) were linked to hospital discharge data obtained from the Washington State Comprehensive Hospital Abstract Reporting System to identify 16,148 cases with GBS colonization and 64,121 controls, frequency matched by year of delivery. We excluded pre-term births and women who received no prenatal care. We identified potential risk factors (race, obesity, pre-eclampsia, genital herpes, bacterial sexually transmitted infections (STIs), and diabetes) and evaluated if obesity modified the association between those risk factors and GBS.

Results: Compared to White women, Black (OR:1.47, 95%CI:1.35-1.58) and Native Hawaiian/Pacific Islander (OR:1.23, 95%CI:1.10-1.38) women were at increased risk while Asian (OR:0.92, 95%CI:0.86-0.98), Hispanic (OR:0.70, 95%CI:0.66-0.75), and Native American (OR:0.86, 95%CI:0.76-0.97) women were at decreased risk. Compared to women with normal BMI, obese (OR:1.17, 95%CI:1.12-1.23) and severely obese women (OR:1.39, 95%CI:1.28-1.51) were at increased risk. Established diabetes (OR:1.34, 95%CI:1.08-1.66), genital herpes (OR:1.26, 95%CI:1.15-1.38), and bacterial STI (OR:1.25, 95%CI:1.08-1.45) also increased risk. Obesity did not modify any observed associations.

Conclusions: We identified several demographic and health factors associated with increased risk for GBS. Knowledge of these potential risk factors for may help identify women at greatest risk of giving birth to babies with GBS infections and inform an eventual targeted vaccine or alternative therapeutic for high-risk pregnancies to protect mothers and infants.

Pre-Pregnancy Body Mass Index and Success of Returning to Pre-Pregnancy BMI Prior to Pregnancy of Second Child, Washington State, 2003-2014

Nicolas Dundas, Tyler Ketterl, Steven Roncaioli

Preceptor: Amanda Phipps

INTRODUCTION: Maternal overweight and obesity is the most common high-risk obstetric condition associated with negative birth outcomes. Although maternal post-partum weight retention may be attributed to a variety of factors, little is known about the association between pre-pregnancy body mass index (BMI) and post-partum weight retention.

OBJECTIVES: The primary objective of this study was to examine the associations between pre-pregnancy BMI status and maternal weight retention.

METHODS: We conducted a population-based retrospective cohort study using Washington State birth certificate data from 2003-2013. We included mothers who had two sequential births during this time period, with the second birth occurring within 18-36 months of the first singleton delivery date. Pre-pregnancy BMI before a mother’s first pregnancy was categorized as normal (18.5-24.9kg/m2), overweight (25-29.9), or obese (30-40). Maternal weight retention was classified as having returned to first pre-pregnancy BMI +≤1 kg/m2 by the start of a second pregnancy. Analyses relating pre-pregnancy BMI to weight retention were stratified by gestational weight gain (GWG) during a first pregnancy (below, met, exceeded recommended GWG).

RESULTS: A total of 49,132 mothers were included in the study. Among mothers who met their recommended GWG, compared to mothers with a normal BMI, obese and overweight mothers were less likely to return to their pre-pregnancy BMI (72.3% vs 68.7%; RRObese = 0.91; CI: 0.87-0.94) (72.3% vs 65.6; RROverweight = 0.95; CI: 0.92-0.99) respectively. Among mothers who exceeded their recommended GWG, compared to mothers with a normal BMI, obese and overweight mothers were less likely to return to their pre-pregnancy BMI (53.2% vs 48.0%; RRObese = 0.90, CI: 0.87-0.92) (53.2% vs 50.5; RROverweight = 0.95, CI: 0.92-0.97) respectively.

CONCLUSION: Above normal pre-pregnancy BMI is associated with a decreased likelihood of returning to pre-pregnancy BMI. Obese mothers and their physicians should regularly be monitored for post-partum weight retention.

Is the extra bedroom worth the extra perinatal stress? Maternal residence change during pregnancy and birth outcomes

Julia Bond, Amanda Mancenido, Divya Patil

Preceptor: Jack Goldberg

Introduction: Psychosocial stress during pregnancy may adversely affect birth outcomes, although the evidence is inconsistent. Studies examining maternal psychosocial stress are often difficult to interpret because maternal stress is treated as a composite exposure of various stressful events occurring over all trimesters. Maternal residence change during pregnancy is an understudied source of maternal stress even though nearly 25% of US women move while pregnant. Our study assessed the effect of maternal residence change during the first trimester on birth outcomes including low birthweight, preterm birth, and any congenital malformation.

Methods: We conducted a population-based cohort study using Washington State birth certificate data from 2007 to 2014 (N=84,252). Women were categorized as first-trimester movers if estimated gestational age at birth subtracted from the estimated time at present residence was less than 12 weeks (n=4,252). Unexposed women were women in which estimated time at residence was longer than estimated gestational age. Analyses examined the cumulative incidence for each outcome and the associated relative risks and 95% confidence intervals. We used Mantel-Haenszel methods to assess potential confounding by demographic and maternal factors.

Results: We found that moving in the first trimester was associated with an increased risk of low birthweight (6.9% vs 4.7%, RR 1.46 [95% CI 1.30, 1.63]) and preterm birth (10.7% vs 6.9%, RR 1.55 [95% CI 1.42, 1.70]). There was no effect of moving on rates of congenital malformations. There was no evidence of confounding by any of the potential confounding factors.

Discussion: Our results suggest that moving during the first trimester of pregnancy may be a risk factor for adverse birth outcomes in United States women. Future studies should attempt to examine whether this association persists for women who move after the first trimester.

Effects of hepatitis B infection on pregnancy outcomes in Washington State, 1992 – 2014

Kristina Bajema, Helen Stankiewicz Karita, Mark Tenforde

Preceptor: Renee Heffron

Introduction: Hepatitis B virus (HBV) infection in pregnancy has been associated with risk of adverse maternal and infant outcomes in highly-endemic Asian settings. However, the association between HBV infection and adverse outcomes is not well characterized in the United States.

Methods: We conducted a retrospective population-based cohort study in Washington State using linked birth certificate and hospital delivery discharge records from 1992-2014. We examined the risk of gestational diabetes (GDM), pre-term delivery (PTD), low birthweight (LBW), and small for gestational age (SGA) comparing pregnant women with HBV identified from either database (n=4,391) to randomly-selected HBV-negative pregnant women (n=27,220) using stratified analysis and Mantel-Haenszel methods.

Results: HBV-infected pregnant women were more likely to identify as Asian (61% vs. 9%) and to be foreign-born (76% vs. 24%) and less likely to be overweight or obese (33% vs. 51%). Adjusting for age, race, country of birth, and pre-pregnancy BMI, HBV-infected women had a similar risk of GDM compared to HBV-negative women (aOR 0.99, 95% CI 0.85-1.15). There was also a similar risk of PTD in adjusted models (aOR 1.01, 95% CI 0.89-1.16). No significant difference was observed between groups for risk of LBW (aOR 0.92, 95% CI 0.79-1.07) or SGA (aOR 0.98, 95% CI 0.89-1.08).

Discussion: In a HBV low-burden setting in the United States, HBV infection was not associated with adverse maternal or infant outcomes.

Maternal chorioamnionitis and neonatal sepsis: A population based case-control study in WA State

Hang Yin, Monalisa Penumetsa, Stephanie Liu

Preceptor: Susan Reed

Background: Maternal chorioamnionitis is a risk factor for neonatal sepsis, an important cause of morbidity and mortality among infants, especially low birth weight babies. This study evaluated the association of chorioamnionitis and neonatal sepsis.

Methods: We conducted a population-based case-control study using data from Washington State birth certificates from 2003-2013 linked to the Washington Comprehensive Hospital Abstract Reporting System (CHARS). Cases and controls were randomly selected from all singleton livebirths four weeks of age or younger. Cases had neonatal sepsis (ICD-9 codes 771.81/995.91/995.92, n=7165). Controls were without CHARS sepsis codes and did not receive antibiotics as indicated on birth certificate (n=65,100). Exposure was maternal chorioamnionitis during labor (ICD-9 codes 762.7/658.4 and/or birth certificate). The odds ratio (OR) and 95% CIs were calculated by stratification analysis, with adjustment for birth year and parity. We also stratified by birth weight and time of sepsis diagnosis (birth hospitalization vs. readmission).

Results: Analyses included 72,265 neonates. Infants with neonatal sepsis were more likely to: be male, have low birthweight, have low Apgar score, be premature, and be born by cesarean. Infants with neonatal sepsis were more likely to have a mother who: was young (<25 years), was single, had high school education or less, had inadequate or intensive prenatal care, had an STI, had antibiotics during pregnancy, was a smoker, and was nulliparous. Infants with neonatal sepsis had a 9-fold increased risk of having a mother diagnosed with chorioamnionitis. The association was higher for normal birth weight infants [aOR: 9.3, 95%CI: 8.3, 10.4] than it was for low birth weight infants [aOR: 5.3, 95%CI: 3.7, 7.7]. The association was higher for neonates diagnosed with sepsis at birth compared to neonates who were readmitted to hospital with sepsis [aOR: 8.1, 95%CI: 7.3, 8.9 vs. 2.6, 95%CI: 2.0, 3.5].

Conclusions: Neonates hospitalized with sepsis were more likely to have mothers diagnosed with chorioamnionitis, particularly normal birth weight infants, and infants with birth hospitalization neonatal sepsis compared to readmitted infants. This finding could be explained by misclassification, or is real - suggesting a need to improve treatment of in utero infections to positively impact neonatal outcomes.

Dads worth the weight: Examining the healthy immigrant effect among fathers on maternal and newborn weight outcomes

Alexandra Akhunova, Katherine Garcia-Rosales, Michelle Passater

Preceptor: Babette Saltzman

Background: With the growing Asian immigrant population in the United States, it is becoming increasingly important to assess population-specific perinatal outcomes. The healthy immigrant effect (HIE) among women is known to impact these outcomes, yet few studies have evaluated the paternal component of HIE. This study assessed the association between paternal birthplace on maternal gestational weight gain (GWG) and small for gestational age (SGA) offspring among Asian immigrants in Washington State.

Methods: A retrospective cohort study using 2003-2014 birth certificate data from Washington state hospitals was conducted evaluating birth outcomes among singleton full-term live births to Asian couples. Due to smaller stature of this population relative to average stature for whom the norms regarding GWG and SGA were calculated in the US, population-specific GWG and SGA scales were constructed using 10th percentile cut-offs. Relative risks and 95% confidence intervals were calculated comparing the risk of low and high GWG and SGA among US-born father and Asian-born mother couples to these risks among Asian-born couples.

Results: Among Asian born mothers- those paired with US born fathers gained more weight than Asian born fathers (RR = 0.72; 95% CI = 0.61-0.86 and RR = 1.2; 95% CI = 1.0-1.4 of low and high GWG, respectively). When assessing SGA, stratified analysis showed that paternal race modified the risk for SGA in this population. Among Chinese and Vietnamese fathers, US-born fathers had a higher risk of SGA when compared to Asian-born couples (RR = 2.3; 95% CI = 1.4-3.8 and RR = 1.9; 95% CI = 1.8-3.0). The opposite association was observed among fathers of other Asian races (RR = 0.85; 95% CI = 0.72-1.0).

Discussion: Compared to Asian-born couples, having a US-born father was associated with higher maternal GWG. The risk of SGA varied among Asian race subgroups for couples with US-born versus Asian-born fathers. Because Asian immigrants comprise a large and diverse population, further research is needed to determine race-specific influences of paternal country of birth and risk of SGA. These findings may reflect differential pressures influencing immigration among Asians from different countries in Asia.

Maternal Exposure to Animal/Crop Agriculture and Pediatric Hospitalizations Due to Acute Respiratory Illnesses in Washington State

BreAnna Kinghorn, Kate McConnell, Orly Stampfer

Preceptor: Anjum Hajat

Introduction

Associations between exposure to animal and crop agriculture and pediatric acute respiratory illness are poorly understood. Studies suggest both protective and harmful effects of early life exposure to animal agriculture on pediatric asthma and wheeze. Growing research indicates that environmental exposures in utero may play a role in later respiratory health. We assessed the association between maternal exposure to animal and crop vs. crop only agriculture and pediatric respiratory outcomes.

Methods

We conducted a retrospective cohort study using data from the Comprehensive Hospital Abstract Reporting System. We used zip-code of residence listed on the Washington State birth certificate to define exposure to animal and crop agriculture. Eight zip-codes were classified as containing animal and crop agriculture. As comparison areas, we selected eight zip codes that contained only crop agriculture that were relatively comparable to the animal and crop agriculture areas in terms of ethnicity, education, income, and amount of crop agriculture. We analyzed 10,503 children from the animal and crop areas and 21,009 children from the crop only areas. All births occurred between 1989-2013. We used stratified analysis to assess associations between maternal exposure to animal and crop vs. crop only agriculture and respiratory hospitalizations within the first five years of life.

Results

Risk of pediatric hospitalization for acute respiratory illness was elevated in the group with maternal exposure to animal and crop agriculture relative to the group with maternal exposure to crop agriculture alone among Latinos (8.3% vs. 6.0%, RR=1.4; 95% CI: 1.2-1.4) and non-Latino non-Whites (8.4% vs. 4.2%, RR=2.0; 95% CI: 1.3-3.2), but there was no association among non-Latino Whites (4.8% vs. 5.4%, RR=0.9; 95% CI: 0.8-1.0).

Discussion

These results suggest that maternal exposure to animal agriculture may increase the risk of pediatric respiratory illness, but appears to differ depending on race/ethnicity. Despite limitations due to non-differential exposure misclassification, this study identifies maternal agricultural exposures as a health equity issue potentially affecting pediatric respiratory health.

Association Between Delivery by Cesarean Section and Necrotizing Enterocolitis: A Case Control Study Among Washington State Infants

Danae Black, Rachel Shaffer, and Stephanie Tornberg-Belanger

Preceptor: Noel Weiss

Introduction

Necrotizing enterocolitis (NEC) is the most frequently occurring neonatal gastrointestinal complication and is associated with serious morbidities and mortality. Previous work has documented differences in intestinal colonization in preterm infants based on mode of delivery, and differences in microbial diversity in the gut of infants with NEC compared to controls. We aimed to address the association of mode of delivery with the incidence of NEC in a population-based case-control study.

Methods

Birth certificate data were linked to the Comprehensive Hospital Abstract Reporting System (CHARS) database to identify cases of NEC (n = 755) during the years 1989-2013 in Washington State. Control infants without NEC (n=3775) who were born in Washington during this period were matched to cases on gestational age, but otherwise were selected at random. The odds ratio (OR) and 95% confidence interval (CI) were calculated to determine the association between mode of delivery (cesarean section (C-section) vs. vaginal delivery) and NEC. We evaluated the following factors for possible confounding: maternal age, maternal race, maternal smoking, birthweight, infant sex, and small for gestational age.

Results

A majority of infants with NEC were delivered by C-section (50.9%), in contrast to 41.1% of controls (OR = 1.5, 95% CI 1.3, 1.7). Adjustment for a priori selected confounders did not substantially alter the size of the association. The higher proportion of cases than controls delivered via C-section was present to a similar degree whether the infants’ mothers had diabetes, infection with group B streptococcus, or whether or not they were breast-fed.

Discussion

We detected a 50% increase in NEC incidence among infants delivered via C-section. Prior studies found C-section to be associated with reduced risk of NEC, which suggests cautious interpretations of our findings. Possible explanations for these differences may be explained by study inclusion criteria, study exclusion criteria, or alternative causal pathways.

Maternal and Infant Risk Factors for the Development of Congenital Hemangiomas

Matthew Dellinger, Kathleen O’Connell, Robert Tessler

Preceptor: Eric Chow

Importance: Congenital hemangiomas (CH) are vascular tumors that develop in utero and are present at birth. While evidence strongly supports an etiologic genetic component, few data exist on maternal or infant risk factors for congenital hemangioma (CH).

Objective: Our aim was to determine the association of maternal factors and infant birth characteristics with the development of congenital hemangioma.

Design, Setting, and Participants: Population-based case-control study using Washington state birth certificate data from 2003-2014. Cases were identified as singleton births with a hemangioma diagnosed at birth and were frequency matched 1:4 on birth year to infants without hemangiomas.

Exposures: Maternal exposures included maternal age, maternal race, gestational hypertension, diabetes, smoking, and prior cesarean section. Infant exposures included sex, preterm birth, low birth weight, small for gestational age, and 5-minute Apgar score.

Main Outcome: International Classification of Disease code version 9 for hemangioma diagnosed during the birth hospitalization.

Results: There were a total of 1,456 cases of CH and 5,824 year-matched controls with no CH. Cases were more likely to be female (59.0% v 48.8%), more frequently preterm (15.5% v 8.1%), and less frequently small for gestational age (7.4% v 9.4%). Mothers of CH infants more frequently had a prior cesarean section (12.8% v 7.7%), but had similar proportions of gestational hypertension, diabetes, and smoking during pregnancy. Multivariable logistic regression demonstrated higher odds of CH with maternal white race (OR 1.7, 95% CI 1.46-1.98), prior cesarean section (OR 1.76, 95% CI 1.46-2.14), infant female sex (OR 1.51, 95% CI 1.33-1.71), and low birth weight (OR 2.42, 95% CI 1.77-3.30) Small for gestational age (SGA) was associated with a lower odds of CH (OR 0.61. 95% CI 0.47-0.79).