You are on page 1of 18

Review Article

433

Ethnic and Racial Disparities in the Risk of
Preterm Birth: A Systematic Review and
Meta-Analysis
Jelle M. Schaaf, MD 1,2 Sophie M.S. Liem, MD 2,3 Ben Willem J. Mol, MD, PhD 2
Ameen Abu-Hanna, PhD 1 Anita C.J. Ravelli, PhD 1

Amsterdam, The Netherlands
2 Department of Obstetrics and Gynaecology, Academic Medical
Centre, Amsterdam, The Netherlands
3 Department of Obstetrics and Gynaecology, Onze Lieve Vrouwe
Gasthuis, Amsterdam, The Netherlands

Address for correspondence J.M. Schaaf, MD, Academic Medical
Centre, Department of Medical Informatics, Room J1B-227, Postbox
22770, 1100 DE Amsterdam, The Netherlands
(e-mail: j.m.schaaf@amc.uva.nl).

Am J Perinatol 2013;30:433–450.

Abstract

Keywords







preterm birth
race
ethnicity
systematic review
meta-analysis
perinatal mortality
risk factors

Objectives The aim of this study is to present a systematic review of available literature
on the effect of maternal ethnicity (Africans/blacks, Asians, Hispanics, others) on the
risk of preterm birth (PTB).
Study Design Studies investigating ethnicity (or race) as a risk factor for PTB were
included if performing adjustments for confounders. A meta-analysis was performed,
and data were synthesized using a random effects model.
Results Forty-five studies met the inclusion criteria. Black ethnicity was associated with an
increased risk of PTB when compared with whites (range of adjusted odds ratios [ORs] 0.6 to
2.8, pooled OR 2.0; 95% confidence interval [CI] 1.8 to 2.2). For Asian ethnicity, there was no
significant association (range of adjusted ORs 0.6 to 2.3). For Hispanic ethnicity, there also
was no significant association (range of adjusted ORs 0.7 to 1.5).
Conclusions Ethnic disparities in the risk of PTB were clearly pronounced among black
women. Future research should focus on preventative strategies for ethnic groups at
high risk for PTB. Information on ethnic disparities in risk of PTB-related neonatal
morbidity and mortality is lacking and is also a topic of interest for future research.

Preterm delivery is one of the most important factors contributing to perinatal morbidity and mortality in obstetric
practice.1 Preterm deliveries are those that occur at less than
37 completed weeks of gestation. The preterm birth risk has
been reported as 11% in the United States, between 5% and
7% in Europe, and 6.5% in Canada.2 The major impact of
preterm birth on public health has led to broad attention to
the topic in scientific research. Many studies have reported
increasing incidence of preterm birth during the last decades,
mainly caused by an increase in medically indicated (iatrogenic) preterm delivery.3–7 Unfortunately, it appears that

efforts to reduce the risk of preterm birth have not resulted
in lower incidence figures.
Preterm delivery results from several disorders, including
known and unknown maternal and fetal disease.8,9 Risk factors
include specific pregnancy characteristics, obstetric history, and
basic maternal characteristics like maternal age, socioeconomic
status, and ethnicity.10 Despite the identification of all these risk
factors, the way that the risk factors interact in and contribute to
the etiology of preterm birth remains largely unknown.11
Numerous publications have shown profound ethnic disparities in many areas of health and health care.12 Ethnic

received
February 6, 2012
accepted after revision
May 24, 2012
published online
October 11, 2012

Copyright © 2013 by Thieme Medical
Publishers, Inc., 333 Seventh Avenue,
New York, NY 10001, USA.
Tel: +1(212) 584-4662.

DOI http://dx.doi.org/
10.1055/s-0032-1326988.
ISSN 0735-1631.

Downloaded by: IP-Proxy Columbia University, Columbia University. Copyrighted material.

1 Department of Medical Informatics, Academic Medical Centre,

We included all ways of determining maternal ethnicity. Gray literature—including unpublished abstracts. Data Sources and Search Strategy We searched the electronic databases of MEDLINE (U. we specified search terms for the three Schaaf et al.. Therefore. several studies have reported contradictory findings on the relationship between ethnicity and preterm birth. parity. Columbia University. and dissertations —and comments. In this review the results will be described per ethnic group. To the best of our knowledge. Hispanics.A Systematic Review on Ethnic Disparities in Preterm Birth Risk disparities in perinatal health care outcomes. Definition of Preterm Birth Preterm birth is defined in the broadest sense as birth before 37 completed weeks of gestation. North Africans or aboriginals). and others (e. We only included information available from the publications and did not seek additional information by contacting primary authors. The review included studies that had ethnic disparity as their main theme.3 This predisposition could not be explained by differences in medical.S. Asians (South Asia. If the primary reviewers could not reach consensus. namely blacks. Studies that focused on subgroups of preterm birth (e. and marital status as the most important confounders. most of which compare African Americans/blacks to Caucasians/whites. the aim of this study is to present a systematic review of available literature on racial or ethnic disparities in the risk of preterm birth. most frequently investigated ethnic groups.14 However. the expertise of the three remaining team members was used to reach agreement. Methods We used the Cochrane Collaboration’s Systematic Reviews of Health Promotion and Public Health Interventions Handbook15 as a guideline for performing this review and meta-analysis.. There were two master’s-level researchers who acted as primary reviewers. The Netherlands) from inception till August 1. Bethesda. The search strategy is summarized in ►Fig. but also Africans). National Library of Medicine. such as preterm birth.g. and Hispanics. and letters were excluded. Data Collection and Quality Assessment Fig. mostly focusing on the ethnic groups living in the United States. Studies reporting on only paternal ethnicity were excluded. The definition of ethnicity is not straightforward when comparing international literature. 1 and available on request.16 Often ethnicity is determined by the doctor’s report or by self-reporting. and candidate gene approaches have suggested an ethnic predisposition of blacks to preterm birth. we will only use the term ethnicity in this article. and behavioral risk factors. the investigated determinant in most studies is often a mixture of ethnicity and race. 30 No. Furthermore. but if results were available for the maternal/paternal ethnicity combination. there were three doctoral-level researchers. Whites (Caucasians) will be used as the reference group. We restricted our inclusion to studies that reported on primary data and adjusted for confounders. Study Selection Each of the initially identified articles was screened by two independent reviewers on title and abstract to determine its suitability for inclusion. social. 434 . Besides general terms for ethnicity or race. It combined terms for ethnicity with terms for preterm birth. American Journal of Perinatology Vol. Central Asia). no systematic review of the available evidence has been published on the impact of race and ethnicity on preterm birth. All English literature was included. 1 Flow diagram of search strategy and results. 6/2013 The systematic review team consisted of five members. The terms ethnicity and race are used interchangeably in the included studies for the systematic review. these studies were included in the present review. To improve readability. maternal age. Southeast Asia. editorials. technical reports. very preterm birth at less than 32 weeks of gestation) were also included and reported separately. In fact. have been relatively intractable over the past decade.g. which are blacks (mostly African Americans. Amsterdam. Downloaded by: IP-Proxy Columbia University. 2011. for English-language articles published in peer-reviewed journals. We included studies that had preterm birth as their primary or secondary outcome and where the aim was to describe ethnic differences in preterm birth risk. Copyrighted material. Asians. Definition of Maternal Ethnicity The main determinant of the included articles is the ethnicity or race of the pregnant women. We considered socioeconomic status. MD) and EMBASE (Elsevier. Studies considering spontaneous as well as iatrogenic preterm births were included.13 Comparisons of groups in the United States. all with extensive experience in social epidemiology in general and specifically in obstetrics.

Patel et al23 stratified their results for maternal BMI where the aOR for preterm birth was highest in women with a normal BMI between 18. the outcomes examined. Quality assessment was performed by using an adaptation of the Quality Assessment Tool for Quantitative Studies of the Effective Public Health Practice Project. Most (24 of 30. Blacks Thirty studies reported results considering black ethnicity and preterm birth.born blacks and 1. and the direction. the incidence of preterm birth among blacks ranged from 4. A summary of the results is presented in ►Table 1. We used the raw data from each individual study to calculate the crude odds ratios (ORs) and 95% confidence intervals (CI) before pooling the data.2 to 1.0 to 1.791 were duplicates or were excluded on the basis of title and abstract.. Twenty-seven (90%) of the selected studies reported an increased adjusted risk of preterm birth within the investigated black ethnic group when compared with whites. The remaining six studies focused on Sub-Saharan African blacks. aORs for the total of preterm birth < 37 weeks. Then we performed a meta-analysis of 22 studies that provided crude data that allowed us to pool and extract an average (unadjusted) estimate.6%) for inclusion. Among these 25 studies. If other types of comparisons (e. Copyrighted material. we pooled crude data of the separate selected articles. leading to the exclusion of a further 30 studies. 2.. <32 weeks) are reported. based on the description of data sources. and the two reviewers abstracted the data separately. Data Synthesis and Statistical Analysis We designed a data abstraction form. We entered and analyzed the data using Review Manager 5 (Cochrane Collaboration. visual inspection of American Journal of Perinatology Vol. Two studies achieved pregnancy dating by using a combination of LMP and ultrasonic measurement of crown-rump length (CRL).8%. and data validity and reliability. All included articles were scored for all four areas resulting in an overall quality score (strong. A random effects model was used (because of statistical heterogeneity in the outcome data) to calculate combined OR and 95% CI. The scoring algorithm is available on request. A Systematic Review on Ethnic Disparities in Preterm Birth Risk . Study characteristics and quality. Schaaf et al. Most studies reported on the dichotomous outcome of preterm birth. To provide a general estimate of the risk of preterm birth within each ethnic group. Most (n ¼ 32) studies were performed in the United States. We recorded data for each article on the ethnicities under investigation. potential for bias. hazard ratios. namely delivery before 37 completed weeks of gestation. considering the appropriateness of analytic methods and of the presentation and interpretation of the results. moderate. Visual inspection of graphical data and I2 test for heterogeneity were performed before pooling the data. or weak).8) for U. based on epidemiological design and the appropriateness and clear characterization of outcome measures. (2) study design. All but one study reported adjusted estimates after controlling for at least one important confounder. Oxford. The aOR was 1. The incidence ratio appeared to be highest for the idiopathic preterm births. resulting in 45 studies (1. However. (3) data collection. and Brazilian blacks.S.6 (95% CI 1. respectively.18 Study quality was examined in four areas: (1) size and representativeness of the sample.g. iatrogenic. ►Figure 2 shows the results.6 to 2. and spontaneous preterm birth following premature prelabor rupture of membranes (pPROM). Schieve and Handler21 presented results after stratifying for the type of medical insurance. 41 showed a significant association between at least one ethnic group and preterm birth. Cervantes et al22 subdivided their results for blacks by country of birth. For this meta-analysis. whereas another two studies used CRL data only. and significance of each association. meta-analysis results. 6/2013 435 Downloaded by: IP-Proxy Columbia University.3 (95% CI 1. especially those occurring before 34 weeks of gestation. 80%) of the selected studies focused solely on blacks living in the United States in their investigation. Results Of 2. The remaining 23 studies did not report which technique was used for determining gestational age.4 to 1. The full text of the remaining 75 publications was evaluated. Only the Zhang and Savitz24 study made a distinction between idiopathic.The two reviewers scored the articles that were selected after screening title and abstract. and aORs for subgroups of preterm birth (e. 60%) of these included studies were scored as strong after quality assessment. Surinamese Creole blacks. 30 No. and the reported aOR ranged from 0. The majority (18 of 30.866 articles identified by our search strategy. The measurements of comparison consisted of mostly adjusted odds ratios (aORs). Only six studies also reported on the mean gestational age per ethnic group. (4) analysis. magnitude.9 to 39.8. both before and after statistical adjustment for confounders (when possible). All abstracted data are available upon request. 1.17 which was modified by Blumenshine et al for their systematic review on socioeconomic disparities in adverse birth outcomes. Most studies reported only the statistically adjusted results.g. We collected the published raw data for meta-analysis.7) for immigrant blacks. 95% CI 1. only data on the most generic definition of preterm birth (<37 weeks) were pooled and the results are presented in a forest plot. The selected articles (►Tables 1–4) were published between 1983 and 2011. Of the 45 included studies.20 also included multiple pregnancies instead of singleton pregnancies only. Columbia University.9 (1.5 and 24. The included observational studies appeared to be very heterogeneous (I2 ¼ 100%).3. Twenty-five studies published results considering the most generic definition of preterm birth. then this is noted in the results. Reasons for exclusion are summarized in ►Fig. which provided conservative estimates of the associations of ethnicity with preterm birth. There were differences in the way gestational age was calculated.6). risk ratios) were performed. Whites were used as the reference group. UK). Two studies19. Seventeen studies performed pregnancy dating by using date of last menstrual period (LMP).

350 7604 8424 10. 6/2013 3104 5305 6060 9026 38.4–1.7 5.9) 12.0–1.0 (0.6 Prevalence Whites (%) OR 1.4) OR 1.0 (1.9 OR 1.8) OR 1.7–1.0 7.9 25.6) OR 1.6 9.1–2.2–2.6–1.2–1.4) Surinamese 8.4–2.4 25.330 33.5 NR OR 1.607 4.2–1.9 (0.3 OR 1.0) 14.3 7.7) black/black (maternal/paternal race) Adjusted Association (with 95% CI) 14.7) maternal age 10–14 y OR 1.1 (1.9) 9.7–4.0 4.1 12.0–1.033 79.6 (1.0) OR 1.7 (1.8 (1.1) black/black (maternal/paternal race) OR 1.5–1.776 1.7 14 12 19 NR NR 18.5 (1.1–1.5) BMI 30. Columbia University.0–29.8 (2.746 52. 30 No.7) immigrant blacks 12.5–1.786 Sample Size c Table 1 Summary of Studies Reporting on Association between Black Ethnicitya and Risk of Preterm Birth 436 Schaaf et al.9 (1.7 (1.3–1.0 5.415 1.1) OR 1.1 9.2 OR 1.4 (1.0 NR NR 11.5 OR 1.1 14.6 17.1) Sub-Saharan Africa OR 1.1–3.7) OR 2.449 21.005.9 (1.039 292.8–2.6 (0.8) HR 2.975.3 (2.3–2.8 OR 0.755 þ  þ þ þ  þ þ þ þ þ þ þ þ þ  þ ? þ þ þ Maternal Age þ þ þ  þ   þ þ  þ þ þ þ þ þ þ ? þ þ þ SES þ  þ  þ   þ þ  þ þ  þ þ  þ ? þ þ þ Marital Status b þ  þ  þ þ  þ þ þ þ þ þ þ  þ þ ? þ þ þ Other Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Reference Group NR LMP LMP NR NR LMP þ US NR NR LMP NR LMP NR NR NR NR NR LMP NR NR LMP LMP Pregnancy Dating NR NR NR 271 NR NR NR NR NR NR NR NR NR NR NR NR NR 276 NR 274 NR Mean gestational Age Whites (d) NR NR NR 265 NR NR NR NR NR NR NR NR NR NR NR NR NR 274 273 NR 269 NR Mean Gestational Age Blacks (d) <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk Definition of Preterm Birth 5.6) Sub-Saharan Africa OR 1.S.6) Ghanaian 4.030.4 7.5) black/white OR 0.6) BMI 18.0–1.4 (1.6 7.542 48.1 (1.0 OR 1.4–3.2) BMI 25.9 OR 1.4–1.4 (1.2 NR 19.   þ  þ   þ þ  þ  þ þ þ   ? þ þ  Parity Adjustments Yes No Yes Yes No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Yes No Yes Yes Yes Data Available for MetaAnalysis Moderate Moderate Moderate Moderate Moderate Moderate Moderate Strong Strong Strong Strong Strong Strong Strong Strong Strong Strong Strong Strong Strong Strong Quality Rating A Systematic Review on Ethnic Disparities in Preterm Birth Risk 28.8 Prevalence Blacks (%) Downloaded by: IP-Proxy Columbia University.8) OR 2.3) BMI < 18.American Journal of Perinatology USA USA USA USA USA USA USA France USA USA USA USA NL USA UK USA Italy USA Brazil USA Singh (1996)27 Gold (2010) 28 Cox (2009) 29 Howard (2006) 30 Collins (1996) 31 Cervantes (1999) 22 Zeitlin (2004) 32 Lu (2004) 33 Shiono (1986) 34 Brown (2007) 35 Frisbie (1997) 36 Goedhart (2008) 37 Shen (2005) 38 Patel (2004) 23 Leland (1995)39 Zanconato (2011) 40 Blackmore-Prince (1999) 41 Barros (2001) 42 Collins (2007) 43 USA Setting Alexander (2007) 26 Getahun (2005) 25 <37 wk Study Vol.112.7–1.2 (0.608 168.4 NR 2.6 (1.9 OR 1. .1–1.6 (1.7–1.9–1.600.6 8.6–2.1) black/white OR 1.2 6.2–1.0 12 NR 8.6 (1.551 122. Copyrighted material.7 6.000 2.7 6.5–24.8 (1.3 (1.8 OR 2.5–1.1 NR 17.3 (1.5 15.7 (1.7 16.9 NR 39.2 (1.7 (1.-born blacks OR 1.3) OR 1.3) OR 1.8) U.

1 15.330 48.6) Sub-Saharan Africa OR 2.7 (1.1 (2.786 32.7–3.8 2.5 5.9 <33 wk <33 wk 3.2 32–36 wk <34 wk <33 wk 1.9) black/white 2.5) black/black (maternal/paternal race) 3.9–2.0 2.2 (1.1–2.7 2. 30 USA Blackmore (1995) 50 USA USA Shiono (1986) 34 Shen (2008)51 France USA Stein (2009) 48 USA USA Simhan (2008)47 Zeitlin (2004) 32 USA Alexander (2007) 26 Kistka (2007) 49 USA Getahun (2005)25 Subgroups of Preterm Birth Setting Study Table 1 (Continued) 666.0 (1.3 (2.6–1.6 1.0 (1.2 20–33 wk 0.5 5 Prevalence Whites (%) <34 wk <37 wk <37 wk <37 wk <37 wk Definition of Preterm Birth N/A Yes N/A N/A N/A N/A N/A N/A N/A Yes Yes Yes Yes Data Available for MetaAnalysis (Continued) Moderate Strong Strong Strong Strong Strong Strong Strong Strong Weak Moderate Moderate Moderate Quality Rating Schaaf et al.0–2.2 (2.7 2.1) Sub-Saharan Africa OR 1.6–5.5–1.0–2.449 21.5 10.7) OR 1. Columbia University.0–1.686 4.7–1.0 19.8 1.1) African American OR 1.5) after pPROM OR 2.4 (1.005.1 <32 wk 2.3) black/black (maternal/paternal race) RR 2.8 (0.9–2.1) black/white 3.3) African American NR OR 1.017 1868 2063 2072 Sample Size  þ þ þ þ þ þ þ þ    þ Maternal Age  þ þ þ þ þ þ þ þ   þ þ SES  þ þ þ  þ  þ      Parity b þ þ þ  þ  þ þ þ     Marital Status Adjustments þ þ þ þ þ þ þ þ þ  þ  þ Other Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Reference Group NR LMP LMP NR LMP NR NR LMP LMP LMP US LMP LMP Pregnancy Dating 275 NR NR NR NR NR NR 274 NR NR NR NR NR Mean gestational Age Whites (d) 272 NR NR NR NR NR NR 269 NR NR NR NR NR Mean Gestational Age Blacks (d) 1. No.8 (1. c A Systematic Review on Ethnic Disparities in Preterm Birth Risk 437 .9 (4.845.5–1.4 (1.975.9) black/white OR 2.3–2.7–2.3 (2.6 (1.5) OR 2.6 OR 2.746 711.7 33–36 wk 3.0 0.6) Sub-Saharan Africa NR 8.1 (2.4 (1.7 (1.7–2.2) private insurance RR 1.0 36 wk 1.8 NR OR 2.2 3.7–2.3) OR 3.2–1.1 OR 2.462 4916 28. Copyrighted material.7) OR 1.9 (1.9 OR 2.015 949.3 (1.1–2.8 (1.2) OR 2.7 (2.NL Brazil USA USA Verkerk (1994) 44 Silva (2007) 45 Adams (1993) 46 Schieve (1996) 21 American Journal of Perinatology Vol.1) Surinamese/Antillean Adjusted Association (with 95% CI) 5.1 7.5 (1.8–3. 6/2013 Downloaded by: IP-Proxy Columbia University.210 2.9) OR 2.7–3.1 OR 2.1 5.8) OR 2.5 9.0–2.7) black/black (maternal/paternal race) 3.7 16 Prevalence Blacks (%) 11.0 NR 22–31 wk <34 wk NR 1.6 (1.3–2.1 5.9 3.0–6.8 4.4 OR 1.8–4.4 (2.8 OR 4.8 13.9) Medicaid or self-pay blacks HR 1.9–3.5) black/Hispanic 3.8 (2.7) OR 1.7) Sub-Saharan Africa NR OR 2.9 34–35 wk <35 wk 2.

9 12. c SES defined by maternal education. and Children) recipient.1 (1.0 4.5 Prevalence Whites (%) 33–36 wk <33 wk <32 wk 20–33 wk 34–36 wk Definition of Preterm Birth RR 2. unless mentioned otherwise.9) HR 3. socioeconomic status.6 (2.5–1. Copyrighted material.7) idiopathic PR 1.5 (3.551 185.017 1868 6060 9026 38.9–2.3 29–32 wk 3.6 (1.9 3.5 (2.4) private insurance RR 2.0–2.American Journal of Perinatology Vol.9 (1.3 (2.5) Medicaid or self-pay blacks HR 2.3–6. body mass index.0 <29 wk 5.0 0.1) idiopathic Adjusted Association (with 95% CI) N/A N/A N/A N/A N/A N/A Data Available for MetaAnalysis Weak Moderate Moderate Moderate Moderate Moderate Quality Rating Downloaded by: IP-Proxy Columbia University.0 8.2 2.5 0.7–2. 30 USA USA Italy USA USA USA Zhang (1992) 24 No.4–3.0 (0.7–2. BMI. not reported.3 1.1 (2. WIC (Women. if not applied (). if not specified (?).1) OR 2.8 10. ultrasound.4) HR 2.4 7. relative risk.244 Sample Size N/A  þ þ þ þ Maternal Age N/A  þ  þ þ SES c N/A  þ  þ þ Parity b N/A  þ  þ  Marital Status Adjustments N/A þ þ  þ þ Other Whites Whites Whites Whites Whites Whites Reference Group LMP US LMP NR NR LMP Pregnancy Dating NR NR NR 271 NR NR Mean gestational Age Whites (d) NR NR NR 265 NR NR Mean Gestational Age Blacks (d) 2.4 20–28 wk 2. RR.7) after pPROM PR 3. family income. a Ethnic group under investigation are African Americans.7 5.9 1. LMP.0–3.9 1.4 1.9 1. OR. Infants. premature prelabor rupture of membranes. Columbia University. Setting Study Table 1 (Continued) 438 Schaaf et al.0) Sub-Saharan Africa OR 1.3–4.8 (3.6) HR 4.5–1. 6/2013 Leland (1995)39 Zanconato (2011) 40 Blackmore-Prince (1999) 41 Adams (1993) 46 Schieve (1996) 21 32.3 2.9–3. hazard ratio. PR.0 (1.0) HR 1.5 6. odds ratio.8) after pPROM PR 2.7 0. SES.6 2.1–3. b Adjustment for confounders if applied (þ).4 <32 wk <37 wk 3.2 0. A Systematic Review on Ethnic Disparities in Preterm Birth Risk Studies are ordered by quality rating and sample size.6 1.6 Prevalence Blacks (%) 4.1) iatrogenic PR 1.2–3. prenatal care utilization and/or maternal insurance status.8 0.9–4. food stamp recipient.6 (1.0 (1.5 0. last menstrual period. if no adjustment applied in general (N/A).4 1.8) iatrogenic PR 3.2–2. NR.6 (1.8) maternal age 10–14 y PR 2. pPROM. US. . HR. incidence ratio.6 (2.

Patel et al23 stratified their analyses into smoking versus nonsmoking women and supported versus unsupported mothers. Eleven studies were included. Mantel-Haenszel. Five studies reported results for preterm birth before 34 completed weeks of gestation and four studies had delivery before 32 weeks as their outcome measure. unsupported women of Asian origin when compared with smoking. All included studies reported the outcome of preterm birth <37 weeks. After quality assessment. 82% of the studies were scored as strong. The reported relative measures of association of Hispanics compared with whites ranged from 0. Three studies show a significant decreased risk of preterm birth within the Hispanic ethnic group. 6/2013 439 Downloaded by: IP-Proxy Columbia University. of which 10 were performed in the United States. the broadest definition of preterm birth. The highest risk of preterm birth was found within the Asian ethnic subgroup of Filipinos. All studies reported adjusted estimates after controlling for at least one important confounder.3%. The included studies appeared to be very heterogeneous (I2 ¼ 99%) and visual inspection did not allow us to present a pooled estimate. Six studies reported on Hispanics without defining this group in more detail. Only two of these studies used whites as their reference group. The risk of preterm birth among Asians varied between 2.2) for a woman of black ethnicity when compared with whites. the included studies showed that the results are comparable. Reported adjusted relative measure of association of blacks compared with whites varied from 1.975. Hispanics ►Table 3 shows a summary of the studies reporting on Hispanic ethnicity as a risk factor for preterm birth. results were comparable as can be seen in ►Table 2.3% and 16. and Filipino women separately instead of considering all Asians as one ethnic group. CI. M-H. unsupported white women. Five studies specified the investigated Asian groups in more detail. they divided the population into U. Between these two studies.7 (<34 weeks) and from 2. Schempf et al52 and Singh and Yu27 reported results for Chinese. confidence interval. Seven studies showed a significant increased risk of preterm birth for the Asian ethnic group compared with whites or at least one of the reported Asian subgroups. The remaining five studies using whites as their reference group showed no significant effect. Asians Seventeen studies reported on the effect of Asian ethnicity on the risk of preterm birth. A Systematic Review on Ethnic Disparities in Preterm Birth Risk . seven studies American Journal of Perinatology Vol. 2 Forest plot of studies on preterm birth among black women.-born women and immigrants. The reported relative measures of association of Asians compared with whites ranged from 0. Various subgroups of preterm birth were investigated.1 to 1.779 and 4. All but one used whites as their reference group. The odds of delivering a child preterm were 2.8 to 2. Fig. The sample size of the included studies varied between 3.9 (<32 weeks). Li et al53 and Yi et al54 presented data on Chinese and Korean women.9 to 2. Japanese. Furthermore. Twelve studies provided crude data and were included for meta-analysis.S.3. 30 No. In addition. Most of these studies (71%) scored as strong after quality assessment. A summary of the published results is presented in ►Table 2.0 to 4. and the reported results show great variation. The aOR appeared to be highest for smoking. ►Figure 3 shows the results. whereas three other studies reported the opposite effect. Columbia University.449 women.5.Schaaf et al. Copyrighted material.7 to 3. respectively.0 (95% CI 1.

2) U.2) Indian OR 0.9–1. Daniels (2006)58 USA Howard (2006) 30 Wong (2008) 55 USA USA Li (2010) 53 USA Singh (1996) 27 Setting Alexander (2007) 26 <37 wk Study 3779 6511 28.112.6) Asian Indian OR 1.1 8.1 (1.2) South/East Asia 6.0 8.9 (0.1) Japanese OR 0.5 7. Cheng (2006) 57 Rao.6 (1.9–1.4–1.330 30.4 (1.449 Sample Size Table 2 Summary of Studies Reporting on Association between Asian Ethnicitya and Risk of Preterm Birth 440 Schaaf et al.7 OR 1.3–1.6–1.9) Chinese 12.6–1.1 5.2) Korean 7.1) OR 1.3 12. 30 No.0 (0.5 OR 1.7–1.7 13.975.4 (1.0) Mainland Chinese OR 0.4 (1.607 4.9–1.9 OR 0.8 OR 1.1 OR 0.6 6.4 OR 1.8 (0.6) Vietnamese 10.6) Filipino 10.1 (0.4 (1.4–1.0–1.3–2.2–1.9–1.1 OR 1.6) Chinese 8. .7 11.2–1.1 OR 1.5 OR 1.7 (1.S.5 (1.280 2.6–1.9–1.9 8.4–1.8 (1.4 6.5) Filipino OR 1.0) foreign-born Korean OR 1.8 (0.7 Prevalence Asians (%) Downloaded by: IP-Proxy Columbia University.2–1.686 316.751 48.0–1. Parity Adjustments SES No No Yes No Yes No Yes No Yes Yes Yes Data available for MetaAnalysis Strong Strong Strong Strong Strong Strong Strong Strong Strong Strong Strong Quality Rating A Systematic Review on Ethnic Disparities in Preterm Birth Risk 37.6 5.-born Chinese Americans 7.S.2 7.5) Japanese 12.2 OR 1.2 4.3–1.0 (0.7 (1.5–2.7) OR 0.4) Adjusted Association (with 95% CI) 8.6 6.6 OR 1.3 (1.039 202.8) Filipina 10.9 (0.9–1.2) Japanese 11.6 9.5 (1.1) Korean 11.6 (1.6) Guamanian 14.9) Hawaiian OR 1.6 (0. Copyrighted material.1 (0. 6/2013 USA France USA Taiwan USA USA USA Zeitlin (2004)32 Yi (2011) 54 Liu (2008)56 Shiono (1986) 34 Rao.4–1.3) Filipino OR 0.2 (0.770 þ þ þ þ þ þ  þ þ þ þ Maternal Age þ þ þ  þ þ þ þ þ þ þ c b  þ    þ  þ þ þ þ   þ     þ þ þ þ Marital Status þ þ þ þ   þ þ þ þ þ Other Other Asian Japanese Whites Nonaboriginal Taiwanese Whites Whites Blacks Chinese Whites Whites Whites Reference Group NR NR LMP NR NR NR NR NR LMP NR LMP Pregnancy Dating NR NR NR NR NR NR NR NR NR NR 273 Mean Gestational Age Whites (d) NR NR NR NR NR NR NR NR NR NR 272 Mean gestational Age Asians (d) <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk Definition of Preterm Birth 9.5 7.4) Vietnamese OR 1.-born Korean 7.5 OR 1.7 (0.0 (0.4–1.5 (1.1) Indonesian 6.0–1.1) aboriginal Taiwanese 7.8–1. Columbia University.3 OR 0.7–1.9 (0.2 8.0) foreign-born Chinese Americans OR 1.5 (1.0) Chinese OR 1.1 (1.7 8.8 6.746 168.3 OR 1.9) Samoan OR 1.1 (1.1–2.6 (0.9 OR 1.2 7.6 Prevalence Whites (%) OR 1.8 9.2) U.4–1.3 7.5–0.7–1.3 OR 1.American Journal of Perinatology Vol.6 11.0 7.5) Other Asian 9.

6) South Central Asia NR NR OR 1.030.6 OR 1. Columbia University.6 (1.3 OR 0.5 12.7 271 275 NR NR NR NR Mean Gestational Age Whites (d) NR 272.6–1.6) Pakistani 9.3–1.6 (1.1–1.2 7. unsupported mother NR NR OR 0.0) Laotian OR 1.9–1.5–1.1–1.2) OR 1.975.1 2.7 OR 1.5 OR 1.9) smoking.4 (1.0–1.9 (0.4 14.2) OR 1.1) OR 1.3 (0.8–1. supported mother OR 2.5–1.2 (1.1–1.9 12.8–1.01–1.1) Pacific Islander 9.9 (1.0) East Asia OR 1. 30  þ   þ   þ Marital Status þ þ   þ þ  þ Other Whites Whites Whites Whites Whites Whites Whites Whites Reference Group NR LMP NR NR LMP LMP þ US NR LMP Pregnancy Dating NR 273.1 Adjusted Association (with 95% CI) Prevalence Asians (%) Downloaded by: IP-Proxy Columbia University.2) Japanese 7.3–2.USA UK UK USA Shen (2005) 38 Patel (2004)23 Aveyard (2002) 59 Nystrom (2008) 60 USA USA Alexander (2007) 26 Stein (2009) 48 Subgroups of preterm birth Italy USA Schempf (2010) 52 Zanconato (2011) 40 Setting Study Table 2 (Continued) 949.2) Vietnamese 9.449 9026 9669 36.9–1.6 5.7 (1.2 OR 1.2 16.1 (0.4–1.7 NR 2.7 (1.9 (0.1 (1.0) South Central Asia NR OR 1. Parity Adjustments N/A N/A Yes Yes Yes Yes Yes Yes Data available for MetaAnalysis (Continued) Strong Strong Moderate Moderate Moderate Moderate Moderate Strong Quality Rating A Systematic Review on Ethnic Disparities in Preterm Birth Risk Schaaf et al.9) Cambodian OR 1.2) Asian/Asian (maternal/paternal race) NR 1.4) Vietnamese 12.257 122.350 NR Sample Size þ þ þ þ þ  þ þ Maternal Age þ þ  þ þ   þ SES c b þ þ  þ þ   þ American Journal of Perinatology Vol.3 (1.0) East Asia OR 1.0) Korean OR 1.0) Southeast Asia OR 1.2) smoking.6) nonsmoking unsupported mother NR 8.3–2.0) Chinese 6.3 12.1 (0.7 7.6–1. Copyrighted material.6) Thai OR 1.8) Southeast Asia NR NR OR 0.2) Asian Indian OR 1.1 (1.9 (0.6 7.9 (0.1–1.4 8.5 (1. 6/2013 441 .8) Indonesian 9.5) OR 0.3 (1.1–1.7) Filipino OR 1.210 4. No.7–1.9–1.1 (1.7–2.6) nonsmoking.5) Asian/white (maternal/paternal race) OR 1.4 (1.7 (1.0 OR 0.5 11.9–1.2 OR 1.8 OR 1.7 (0.9 13.7 (1.9 (1.2–1.5–2. supported mother NR OR 0.3–4.4–2.0 (0.3–1.5 Prevalence Whites (%) OR 1.6 (1.415 1.3 (1.5 (1.3 16.4) Hmong 10.2 271 274 274 NR NR NR NR Mean gestational Age Asians (d) <32 wk 32–36 wk <33 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk Definition of Preterm Birth NR 1.1 OR 1.5–2.9 (0.8–1.

1 (0.2 OR 4.7–1.0) Indian 3.9 1. if no adjustment applied in general (N/A).7–1.3 OR 1.5) Asian/white (maternal/paternal race) OR 1.6–1.3) OR 0.0 3.6 NR OR 0.2 0.686 Sample Size N/A N/A N/A N/A N/A N/A N/A N/A Data available for MetaAnalysis Moderate Moderate Moderate Strong Strong Strong Strong Strong Quality Rating A Systematic Review on Ethnic Disparities in Preterm Birth Risk 40 France USA Wong (2008) 55 Zeitlin (2004) 32 Setting Study Table 2 (Continued) 442 Schaaf et al. OR. 6/2013 USA UK USA Rao.3) Filipino 2.9–1.2 (0.American Journal of Perinatology Vol.1–1. socioeconomic status.3 (0. food stamp recipient. US.2 (1. not reported.8) OR 1.2 3.7 (1.8–1.2–0.6 (0.1 4.4–3.8 0. Cheng (2006) 57 Zanconato (2011) USA Shiono (1986) 34 48.7 33–36 wk <33 wk 0. . and/or maternal insurance status. and Children) recipient.3 OR 0.4 (1.1 (0.4) South/East Asia OR 1.6 OR 1. last menstrual period.7 (2. if not specified (?).8 OR 1.746 202.6) Samoan OR 1. WIC (Women.5 1.7 Prevalence Asians (%) Downloaded by: IP-Proxy Columbia University.0) Filipino 4.7) Chinese OR 1. family income.7) Guamanian 2.7 0.0–1.6 <28 wk <32 wk 2.1–1.3 (1.4–2.4 (1.330 þ þ þ þ þ þ þ þ Maternal Age  þ þ þ þ þ þ þ c b  þ þ  þ  þ þ Parity Adjustments SES   þ   þ  þ Marital Status   þ þ þ þ  þ Other Whites Whites Whites Other Asian Japanese Whites Whites Chinese Reference Group NR NR LMP NR NR LMP NR NR Pregnancy Dating 271 275 NR NR NR NR NR NR Mean Gestational Age Whites (d) 271 274 274 NR NR NR NR NR NR Mean gestational Age Asians (d) <32 wk 3.8) Asian/Asian (maternal/paternal race) OR 1.6–1.1 (0.5) Japanese OR 1.0 OR 0.2 (1.0–1. Columbia University. Infants.7 2.6 OR 1.2 Adjusted Association (with 95% CI) 1.2 OR 1.9–2.6) Vietnamese 2.0 Prevalence Whites (%) <33 wk <32 wk Definition of Preterm Birth OR 1. c SES defined by maternal education.9) Asian 1. b Adjustment for confounders if applied (þ).1 <34 wk <34 wk <34 wk 1.2 0.3 (0.9) OR 1.7 (0.8 OR 1.3) Hawaiian OR 2. SES. Copyrighted material.2–3. NR.4) Vietnamese 2.9 4.5) Pacific Islander 1.9) Korean 1.3 0.6 (1.8) Chinese OR 1.0–3.4 (0. if not applied ().7) OR 1.7–2. odds ratio.1 (0.5 1. prenatal care utilization.5–5.0 (0.1 (0.2 1. Italy USA Rao.3 2.257 3779 6511 28.3–1.9 (0.6 (1.7–8. 30 No.3 (0. Daniels (2006)58 Aveyard (2002) 59 Nystrom (2008) 60 9026 9669 36.4) South/East Asia 3.1–0. Studies are ordered by quality rating and sample size.2 (1.4 1.9–1. unless specified in more detail.8 2.2 OR 1. LMP. ultrasound.9–1.0) Korean 1.0–1.0) Filipina 1.9 3. a Ethnic group under investigation are Asians.

Five studies used subgroups of preterm birth (e. 95% CI 1. the American Journal of Perinatology Vol. which makes it difficult to compare the results. when looking at preterm birth rates in blacks before 37 weeks of gestation.62. <34 weeks) as their main outcome measure. some studies did not control for important confounders. 443 provided crude data that allowed us to pool and extract an average unadjusted estimate. with ORs ranging from 0.63 Taipale and Hiilesmaa also stated that they found no difference in the number of preterm deliveries when using CRL measurement instead of LMP. A summary of all the results is presented in ►Table 4. <32 weeks) there also appeared to be no significant effect.6 to 2.g. The studies that did report these data mostly calculated gestational age by using LMP.868 to 21. For women of Asian ethnicity. However.38. For women of Hispanic ethnicity.23. Less adjustment for confounding will lead to an overestimation of the effect of maternal ethnicity on preterm birth. and therefore the lack of information on technique of pregnancy dating has not influenced our results to a large degree. 30 No. Currently recognized confounders do not appear to explain the increased risk of preterm birth among black women. Forty-four of the included studies (all but one) only reported data on preterm birth as a single outcome. pooled odds ratio 2.3 (17 included studies).5 days more accurate than the use of LMP. Most of the studies were scored as strong (55%) or moderate (45%) after quality assessment. .6 to 2.40. 3 Forest plot of studies on preterm birth among Asian women. Taipale and Hiilesmaa61 showed that prediction of day of delivery by ultrasonically measuring CRL between 11 and 16 weeks of gestation is 1. Discussion Principal Findings We found 45 studies on the association between maternal ethnicity and the risk of preterm birth.46 Because various factors other than maternal ethnicity are associated with preterm birth. Copyrighted material. confidence interval. there was no significant association.605).. Nevertheless. 6/2013 Downloaded by: IP-Proxy Columbia University. showing the greatest racial disparities in the subtype of spontaneous preterm birth without pPROM.0..7 to 3. and iatrogenic) is lacking. maternal age. Five studies specified Hispanic ethnicity in more detail. Fig. Additional information on racial disparities in the three subtypes of preterm birth (idiopathic. they have less statistical power.8 to 2. Only the study of Zhang and Savitz24 made this distinction.45. The included studies appeared to be very heterogeneous (I2 ¼ 99%). Most studies performed adjustments for the most relevant possible confounders like socioeconomic status. The included studies show great variation in the reported incidence of preterm birth. M-H.g.21. Unfortunately. the included observational studies were scored as strong or moderate after quality assessment. Mantel-Haenszel. For instance. focusing on country of birth or including paternal ethnicity. For the subgroups of preterm birth (e. there was no significant difference in the risk of preterm birth when compared with whites. which in turn has an impact on the generalizability of the smaller studies and.A Systematic Review on Ethnic Disparities in Preterm Birth Risk Schaaf et al. Strengths and Weaknesses of the Included Studies In general. Other Ethnicities We included 11 studies that investigated the effect of other ethnicities on the risk of preterm birth. spontaneous after pPROM. There was great variation in the sample size of the included studies (range 1. Various reference groups were used. CI. as in other studies. Blacks appear to have a significantly increased (range of adjusted ORs 0. of which 41 reported a significant positive association between at least one ethnic group and preterm birth risk. 4). Three of these studies showed no significant increased risk of preterm birth when compared with whites. inevitably. We included four studies investigating ethnic groups from Mediterranean countries including Turkish and Moroccan women and Middle East or North African women in general.8. Columbia University.012. and parity. it is difficult to compare and combine the results of individual outcomes across studies because of varying degrees of control for potential confounders. and visual inspection did not allow us to present a pooled estimate (►Fig. the majority of studies did not report how pregnancy dating was achieved.2) risk of preterm birth when compared with whites (30 included studies).

Columbia University.9) immigrant Puerto Rican 12.1 (1.2–1.8) OR 1.0–1.5 8.3) Haiti OR 1.7 17.030.3) other Caribbean country NR OR 1.0 8.0–1.210 2.3) other Caribbean country NR NR OR 4.8–1.7 (0.4) other Caribbean country NR OR 1.5–1.5–4.039 21.542 52.8 (0.3–1. Copyrighted material.6) 1.3 NR 2.6 NR 2.4 5.0) Puerto Rican 10.424 1.605 Sample Size þ þ þ þ þ þ þ þ þ þ  þ Maternal Age þ þ þ þ þ   þ þ þ þ þ c b þ þ  þ þ   þ þ þ  þ  þ þ þ þ   þ þ þ  þ Marital Status þ þ þ þ þ  þ þ þ þ þ þ Other Whites Whites Whites Whites Whites Whites Whites Whites Whites Whites Blacks Whites Reference Group NR US NR LMP LMP NR NR LMP NR NR NR LMP Pregnancy Dating NR NR NR 273 NR NR NR NR NR NR NR 273 Mean Gestational Age Whites (d) NR NR NR 272 NR NR NR NR NR NR NR 272 Mean Gestational Age Hispanics (d) 32–36 wk  27 wk 28–31 wk 32–36 wk <34 wk <33 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk Definition of Preterm Birth NR NR 1.4 9.7 OR 1.605 8.3 OR 1.4 (1.4) OR 0.3 (1.1 OR 1.0–1.8) Hispanic Caribbean OR 1.0–1.350 10.134 2. 6/2013 Brown (2007)35 Shen (2005) 38 Frisbie (1997) 36 USA USA Canada USA Alexander (2008) 64 Simhan (2008) 47 Auger (2011)65 Stein (2009) 48 949.8 (1. .3 8.012.American Journal of Perinatology Vol.7) Central America NR NR OR 3.1) West-Indian/ Brazilian OR 1.6) Mexico OR 1.2–1.0 (3.5–0.2) Hispanic/white (maternal/paternal race) 1.8.2–1.6–2.3 Prevalence Whites (%) OR 1.6 6.7 NR OR 1.5 (2.-born Puerto Rican 10.012.033 168.6 1.330 33.5 (1.1) Hispanic/Hispanic (maternal/paternal race) OR 1.686 21.3 OR 1.4 8.0 7.8 OR 0.143.6 (1.1 (1.3 (1.8–1.0–1.845.9 (1.2 (1.3–1.1.S.5 (1.5 (1.9) Hispanic/black (maternal/paternal race) 2.7 6.-born Mexican OR 0.8) OR 0.7–0.2) OR 1.3) Haiti OR 1.9 (0.1 9.7 NR OR 1.0 (0. 30 USA USA USA USA USA Lu (2004) 33 Shiono (1986) 34 No.0 (1.0) OR 0.9) immigrant Mexican OR 1. Parity Adjustments SES N/A Yes N/A N/A No Yes Yes Yes Yes Yes Yes Yes Data Available for Meta-Analysis Strong Strong Strong Strong Moderate Moderate Strong Strong Strong Strong Strong Strong Quality Rating A Systematic Review on Ethnic Disparities in Preterm Birth Risk Subgroups of preterm birth USA USA Cervantes (1999) 22 USA Howard (2006) 30 Setting Alexander (2008) 64 <37 wk Study Table 3 Summary of Studies Reporting on Association between Hispanic Ethnicitya and Risk of Preterm Birth 444 Schaaf et al.0 (0.3) Adjusted Association (with 95% CI) 10.7–0.8) South/Central American 7.7–1.7–2.9 Prevalence Hispanics (%) Downloaded by: IP-Proxy Columbia University.6) Haiti NR OR 2.5) Cuban 11.9–4.8–1.9 1.4–1.2–1.3 (1.5) U.5 8.7–1.9 (0.755 28.1 (0.S.2–1.3 (1.2) U.8 (0.1 (1.8 OR 0.1 (0.

5–3. The search strategy was broad and thorough and designed to capture all available relevant literature. odds ratio. Furthermore.1. However. Infants.9% reported by Shen et al38 to almost 40% reported by Leland et al. this is the first review focusing solely on the subject of ethnicity or race as a risk factor for preterm birth. Strengths and Weaknesses of This Review To the best of our knowledge.6–1. socioeconomic status. SES. OR 2. ultrasound. WIC (Women.6–2. incidence ranges from 4.7–3.7 (1. Leland et al only included teenage pregnancies with a maternal age between 10 and 14 years. the way in which the composite variable socioeconomic status was determined varied between studies.8 (1. In most studies. OR.7) other Hispanic NR OR 2.5–1. Furthermore. On the other hand. if not specified (?). family income. if no adjustment applied in general (N/A). American Journal of Perinatology Vol. c SES defined by maternal education. We used general synonyms for ethnicity in our search terms. b Adjustment for confounders if applied (þ).3 OR 1.330 USA Shiono (1986) 34 Schaaf et al. a Ethnic group under investigation are Hispanics.0 <33 wk NR NR LMP Whites þ þ þ þ þ 28. 30 No. a narrative review (Dominguez69) together with all of our included studies were checked for relevant references. 6/2013 445 Downloaded by: IP-Proxy Columbia University. and Children) recipient. The problem of definition also holds. ethnicity was defined by the caregiver. Most included studies that were performed in the United States used a classification of ethnicity into nonHispanic whites.9) OR 2. which is a composite measurement of maternal education and family income. US. a recently published systematic review by Shah et al68 on perinatal outcomes among Aboriginal women contained studies on Aboriginal women that were not included in this review. defining maternal ethnicity is not straightforward. Asians. “Inuit”) in their title or abstract without the words ethnicity or race might thus have been missed.6 (1. unless specified in more detail.0–2.67 many studies that were included in this review performed a complete case analysis and did not use imputation techniques for missing data. The results presented in the studies are less applicable or difficult to translate to countries with another or mixed composition of maternal ethnicities. NR.3 (0.9–1. Our search strategy was designed to retrieve all studies with ethnicity as their main theme when investigating the risk of preterm birth.1) Mexico NR OR 1. The number of studies included in the “others” section is influenced by this limitation.8) South America NR OR 1. not reported. Others used self-reported ethnicity data provided by the participating women or studies classified ethnicity by country of origin or by skin color. Columbia University.1) Hispanic Caribbean OR 1. and Hispanics as they are the three most frequently investigated ethnic groups. and Hispanics.39 These large differences can mostly be attributed to varying inclusion and exclusion criteria of the different studies.16 The main determinant investigated in the included studies was often a mixture of ethnicity and race. This specific subgroup is at high risk for various adverse perinatal outcomes compared with the total population. if not applied (). Copyrighted material. unfortunately not all studies reported their results in a way that allowed us to pool them for metaanalysis.8 (2. incidence figures differ and as a consequence we did not include these data for metaanalysis.66 Therefore.7) South America NR NR Adjusted Association (with 95% CI) Pregnancy Dating Parity Setting Study Table 3 (Continued) Sample Size Maternal Age SES c Adjustments b Marital Status Other Reference Group Mean Gestational Age Whites (d) Mean Gestational Age Hispanics (d) Definition of Preterm Birth Prevalence Whites (%) Prevalence Hispanics (%) N/A Data Available for Meta-Analysis Strong Quality Rating A Systematic Review on Ethnic Disparities in Preterm Birth Risk . For instance.8) other Hispanic NR 22–31 wk Studies are ordered by quality rating and sample size. OR 2. prenatal care utilization and/or maternal insurance status.3–2. This might have led to biased estimates.2) Central America NR 1. For instance. food stamp recipient. last menstrual period. Finally. We discussed the difficulties of defining ethnicity in the Methods section.5 (2.9 (2. LMP. Despite recent advances in the handling of missing data. but to a lesser extent.4 (2. American blacks. Articles with only a specific ethnicity name (for instance. for socioeconomic status. we did specify our search terms for blacks.

5 (1. family income.0 NR 8.2 (0. last menstrual period. SES.0–1.3 (1. OR.1 OR 1.164 Israel Italy NL Brasil Melamed (2000)71 Zanconato (2011) 40 Verkerk (1994) 44 Silva (2007) 45 USA Italy Stein (2009) 48 Zanconato (2011) 40 9026 949.3 (0.0 7.3) Bedouins OR 0.5–0.3–1.1 9.3 (1. ultrasound.210 21.605 NL Australia Sample Size Goedhart (2008) 37 USA Alexander (2008) 64 Setting Langridge (2010) 70 <37 wk Study Table 4 Summary of Studies Reporting on Association between Various Ethnicitiesa and Risk of Preterm Birth 446 Schaaf et al.3 Prevalence Whites (%) OR 1.2–1.5–1.2) Moroccan 4.1 OR 0. 30 No.9–1.3) American Indian Adjusted Association (with 95% CI) 5.5) Native Hawaiian 11.0 5.0 9.8) North Africa OR 1.04–1.3–1.0 Prevalence Other (%) Downloaded by: IP-Proxy Columbia University.8 13. if not applied ().5 3 17. a Ethnic group under investigation is specified in more detail.1 OR 1.8 (0.5) Mulatto OR 0. if no adjustment applied in general (N/A).8 2.2–1.5 5 16. US.605 þ þ þ  þ  þ þ þ þ   þ þ b  þ þ    þ þ þ þ þ þ þ Parity   þ     þ þ þ þ  þ Marital Status  þ þ  þ  þ þ þ þ þ þ þ Other Whites Whites Whites Whites Whites Whites Jewish Whites Whites Whites Whites Nonaboriginals Whites Reference Group NR NR LMP LMP LMP NR LMP þ US NR LMP NR NR NR LMP Pregnancy Dating 271 NR 273 NR NR 271 NR NR NR NR NR NR 273 Mean gestational Age Whites 270 272 NR 266 NR NR 270 272 NR NR NR NR NR NR 266 Mean Gestational Age Other <33 wk 22–31 wk 32–36 wk <33 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk <37 wk Definition of Preterm Birth 3. odds ratio.7–0. food stamp recipient. .0 9.3–1.1–1.3) American Indian OR 1. 6/2013 69.5 5.9) Middle East and North Africa OR 1.5) Central and South America OR 1. Copyrighted material.7 (0.9–1. USA Alexander (2008) 64 2063 2072 9026 þ þ þ þ þ þ c Adjustments SES þ þ þ þ þ þ Maternal Age N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A Data available for MetaAnalysis Moderate Strong Strong Moderate Moderate Moderate Moderate Moderate Strong Strong Strong Strong Strong Quality Rating A Systematic Review on Ethnic Disparities in Preterm Birth Risk Subgroups of preterm birth 1.012.5 (1.3 5.5) Central and South America OR 0. c SES defined by maternal education.0 (0.7–1.8–2.5 (1.4 (1.3) North Africa OR 1.6 (0.5 OR 2.8) Inuit OR 1.8 NR 1.8 OR 1.5–1. not reported.1–1. WIC (Women.6) Marshallese 18.125. Columbia University.9) Indian OR 1.1–1.2 (0.0) Middle East and North Africa OR 1.1 (1.0 NR OR 0.7 (0.1) American Indian OR 1.012. socioeconomic status. prenatal care utilization.8 (0. LMP.2–2.7–2.5 8.542 Schempf (2010) 52 USA Lu (2004) 33 567.American Journal of Perinatology Vol.6–1. and/or maternal insurance status. if not specified (?).1) Turkish OR 1.8 (1. NR. Studies are ordered by quality rating and sample size.3–1.3 4.3 (1.8 NR 1. and Children) recipient.7) Aboriginal OR 1.2 (0.6) Samoan 12.462 Canada Luo (2004) 20 NR USA 7604 33.5) Turkish OR 0.8 10.8 9.7 NR 11.468 21.8) Guamanian 11.2 (1.3–1.1 (1. b Adjustment for confounders if applied (þ).5) Asian Pacific Island OR 0.9 (0. Infant.3 (1.9 NR 6.6) Tongan 10.

30 No. Copyrighted material. and (3) self-selection of the healthiest immigrants. This is likely to be incorrect. In the future. 4 Forest plot of studies on preterm birth among Hispanic women. As presented in this review. The effect of Asian and Hispanic ethnicity on the risk of preterm birth is less pronounced. Subsequently. one should investigate the heterogeneity of the different studies. As the majority of studies in this review focus on ethnic groups living in the United States. For Hispanics. This individual risk profile should be determined using all specific maternal. 6/2013 Downloaded by: IP-Proxy Columbia University.16 Also. This is done by eyeballing as well as by performing the I2 test for heterogeneity. Despite the major effort of much scientific research. Future research should further focus on the genetic or epigenetic components leading to the increased incidence of preterm birth.and long-term health care costs. In literature. Future research should focus on the question of whether there are ethnic disparities in optimal gestational length.. especially among blacks and Asians. . Instead. In this review.g. The perinatal condition after preterm birth has a great impact on shortand long-term morbidity and short. When investigating optimal gestational length. CI. M-H. Risk of preterm birth appeared only to be increased in some Asian subgroups. However. Thus we implicitly assume that mean gestational length is similar for all individuals. Despite tending to be less educated. Preterm birth is defined by using the rigid cutoff of birth before 37 completed weeks. Therefore epidemiological research should be performed to investigate disparities in. and the late entry into prenatal care. and the strongly related risk of perinatal morbidity and mortality.3–7 To significantly reduce the risk of preterm birth. we have demonstrated the ethnic disparities in risk of preterm birth independent of other risk factors. especially black ethnicity. Fig. Meaning of the Results and Proposal for Future Research This review emphasizes the independent effect of ethnicity.1 the results presented here highlight the need for thoughtful conceptualization of likely pathways through which ethnicity affects preterm birth risk and. (2) protective factors in the immigrant culture. having high rates of being uninsured and low socioeconomic status. This effect of black ethnicity is even more pronounced in the subgroups of preterm birth (e. the ethnic diversity of other countries is not well represented. black and Asian maternal ethnicity appears to be an important factor in determining such an individual risk profile. confidence interval. we found no significant increased risk for preterm birth when compared with whites.A Systematic Review on Ethnic Disparities in Preterm Birth Risk Schaaf et al. <34 of <32 weeks). we should adjust the perinatal care provided to an individual woman’s risk profile. Before pooling the results. However. Mantel-Haenszel. we should also incorporate perinatal outcome in the methodology of research. maternal and infant health. The latter showed poor results in our analyses.77 Information on the risk of preterm birth among ethnic minorities from NorthAfrican or Middle East origin is scarce. the risk has been increasing in most developed countries. The causal pathway of this phenomenon is likely to be of epigenetic origin. it is known that the I2 heterogeneity tends to overestimate heterogeneity in studies performed on large databases. this cutoff is identical for all ethnic groups. in meta-analyses of observational studies.”73 The possible explanations for this phenomenon are: (1) strong social support within the immigrant community. In current clinical practice. we should think of possible ways to intervene in those pathways. in turn. This issue is an important topic for future research within perinatal care.72 The heterogeneity might be caused by the usage of slightly varying definitions of ethnicity or the inclusion of slightly differing subgroups of a specific ethnic group. Columbia University. introduced through the individual studies as well as by the process of selecting studies for a systematic review. 447 Meta-analyses are limited by biases. irrespective of maternal ethnicity.74–76 These relatively favorable perinatal outcomes are especially reported for the first-generation immigrants. ethnic groups living in Europe. the subject of our review does not lend itself to experimental studies such as randomized trials. Optimal gestational length should namely be defined as the gestational age at which the risks of perinatal and maternal morbidity and mortality are the lowest. biomarkers might help us to assess the individual risk profile or provide an incentive to investigate preventive treatment strategies. there is no significant reduction of the risk of preterm birth over the last decades. secondary researchers are unable to adjust for possible confounders. including paternal and fetal characteristics known to contribute to preterm birth pathogenesis. Goedhart et al37 and Verkerk et al44 showed no significant risk difference when comparing Turkish and Moroccan women to the Caucasian American Journal of Perinatology Vol. on the risk of preterm birth before 37 weeks. for instance. this phenomenon is often referred to as the “Hispanic paradox. Hispanics have relatively low rates of preterm birth. As preterm birth is the most important cause of perinatal morbidity and mortality.

Perinatal factors and preterm delivery in an Irish obstetric population. 30 No. The authors modified that protocol for the quality assessment in this review. Iams JD. Ananth CV. et al. Future research on ethnic disparities in preterm birth risk should systematically adjust for important confounders. et al. Washington AE.309:327–330 Effective Public Health Practice Project. Demissie K. Barclay CJ. 175:1317–1324 Jackson N. Australia. Copyrighted material. Romero R. 2002. Systematic reviews of health promotion and public health interventions handbook. 1998–2002.39:263–272 de Costa C. Birth outcomes of AsianIndian-Americans.9:429–435 3 Demissie K. Little MP. Morrison JJ. Preterm delivery and perinatal death among black and white infants in a Chicago-area perinatal registry. Adverse perinatal outcomes among interracial couples in the United States. Rivera L.. Morris C. Laws P. As ethnic compositions of societies differ greatly. The effect of changing patterns of obstetric care in Scotland (1980–2004) on rates of preterm birth and its neonatal consequences: perinatal database study. Mor J. Med J Aust 1996. Marshall SS. Tomich PG. Etiology and outcome of extremely low-birth-weight infants.371:75–84 Steer P.3:428–433 Getahun D. 448 .88:356–363 Cervantes A.3:99–109 Patel RR. Available at: http://www. Accessed 1 Feb 2012 Goldenberg RL. Kramer MS. Platt RW. Am J Epidemiol 2001. Am J Obstet Gynecol 1996. who collaborated on the paper by Blumenshine et al. Demonner SM.H. non-low birth weight infants: a population study. Adverse birth outcomes among native-born and immigrant women: replicating national evidence regarding Mexicans at the local level. Does gestation vary by ethnic group? A London-based study of over 122.D. 12 Bryant AS. This merits research on the causal pathways of these differences and on preventative measures to reduce the incidence of preterm birth. 118(5):e1399–e1405 Collins J. Papiernik E. Morrison JJ. Graham J.33:107–113 Zhang J. Obstet Gynecol 2005. Obstet Gynecol 1996. Handler A. Int J Epidemiol 2004. Lancet 2008. Caughey AB.A Systematic Review on Ethnic Disparities in Preterm Birth Risk women. psychosocial. discussion 1900–1902 Goldenberg RL. and C. Culhane JF.154:307–315 Joseph KS. Epidemiology and causes of preterm birth.D. Ph. and history of preterm birth.97(3):215–220 Singh GK. BMJ 1994. Yang Q.161:1409 Norman JE. Carlton South. CMAJ 1999.18:40–50 Schieve LA. Risks of adverse pregnancy outcomes among Inuit and North American Indian women in Quebec. 112(Suppl 1):1–3 American Journal of Perinatology Vol. A.6:e1000153 Tracy SK. Pregnancy outcomes in urban aboriginal women. Am J Public Health 1996. Keith L.106(1):81–88 Alexander GR. Worjoloh A. 6/2013 Schaaf et al. Int J Gynaecol Obstet 2007. Pediatrics 2006.pdf. Kaufman JS. Wilkins R. Accessed 1 Feb 2012 Blumenshine P.P. Smith G. Child A. Gianopoulos JG.and foreign-born women in major US racial and ethnic groups. M. Lancet 2002. Preterm birth and maternal country of birth in a French district with a multiethnic population. Ethnicity as a variable in epidemiological research. Cubbin C. Spontaneous preterm birth of liveborn infants in women at low risk in Australia over 10 years: a population-based study. 18 19 20 21 22 23 References 24 1 Slattery MM. Am J Epidemiol 1996. Prenatal care utilization in Mississippi: racial disparities and implications for unfavorable birth outcomes. Savitz DA. Dean J. This will improve the quality of research and the possibility for comparison.. Determinants of preterm birth rates in Canada from 1981 through 1983 and from 1992 through 1994. with black women being at higher risk. Lantz PM. Healthy people 2010: Midcourse review. such as socioeconomic status. Acknowledgments The authors would like to thank Dr. Yu SM. Bucourt M. Sullivan E. Doyle P. Expanding our knowledge on these and other ethnic groups is important.143(4):333–337 Zeitlin J. Epidemiology 1992. Kramer MS. Trends in preterm birth 4 5 6 7 8 9 10 11 and neonatal mortality among blacks and whites in the United States from 1989 to 1997. Boulet S. Preterm birth subtypes among blacks and whites. Am J Prev Med 2010. Matern Child Health J 1999. Bhopal R.111(8):849–855 Downloaded by: IP-Proxy Columbia University. Prematurity and low birth weight as potential mediators of higher stillbirth risk in mixed black/white race couples. Fetal and Infant Health Study Group of the Canadian Perinatal Surveillance System. Elliott P. et al. Am J Obstet Gynecol 1996. Tracy MB. Egerter. BJOG 2005.86:837–843 Gold KJ. Preterm delivery.000 pregnancies with spontaneous onset of labour. 2006 Senior PA. 2002. Rhoads GG. Recent versus historical trends in preterm birth in Canada. Wyshak G. The epidemiology of preterm labour. future prospective research should focus on ethnic groups living outside the United States. BJOG: An International Journal of Obstetrics and Gynaecology 2004. N Engl J Med 1998. Columbia University. 1985–97. Selvam N.ca/PDF/QATool.B.15:931–942 [epub ahead of print 2009 Nov 27] Howard DL. ethnic disparities clearly exist. Variations in low birth weight and preterm delivery among blacks in relation to ancestry and nativity: New York City.360: 1489–1497 25 2 Wen SW. Adverse pregnancy outcomes: differences between US. Epidemiology of preterm birth and neonatal outcome. Ontario. J Womens Health 2010. Hamilton. Wingate MS.114:731–735 Burke C. Barclay. Steer P. Despite the heterogeneity of the included studies in defining ethnicity and adjustment for confounding. and behavioral risk factors do not explain the increased risk for low birth weight among black women.18 for sharing their protocol for quality assessment of included studies. Kramer MS. Egerter S. Cliver SP.202:335–343 USDHHS. Mulvihill FX. Cochrane Collaboration Health Promotion and Public Health Field. Zhang L. Matern Child Health J 2011. Quality assessment tool for quantitative studies. as these ethnicities form a substantial proportion of women of childbearing age in many European countries. Hammond NA.19(4): 767–773 Cox RG.28:49–53 Iannucci TA. Walker M. Marcoux S. Chalmers J. maternal age.ephpp. PLoS Med 2009. BJOG 2007. Paediatr Perinat Epidemiol 2004.164:523–526 Luo Z-C. Braveman PA.174:1896–1900. Medical.339:1434–1439 Joseph KS.healthypeople. 26 27 28 29 30 31 32 disparities in obstetric outcomes and care: prevalence and determinants. whereas Zanconato et al40 showed a significant decreased risk of preterm birth for women from the Middle East or Northern Africa. Available at: www. Racial/ethnic 13 14 15 16 Conclusions 17 There are clear ethnic disparities in the risk of preterm birth. Topuz B. Am J Obstet Gynecol 2010. Socioeconomic disparities in adverse birth outcomes: a systematic review. Zotti ME. CanadaMcMaster University. J Perinat Med 2000. Hayward RA. Savitz DA. Semin Fetal Neonatal Med 2004. Ananth CV. Relation of maternal race to the risk of preterm.gov/.

van der Wal MF.e1–584. J Immigr Minor Health 2010. Keith LG.e1–131. Gardosi J. Chery M. Forbes D. Copyrighted material. Obstetric risks of pregnancy in women less than 18 years old. Daniel M. Janevic T. Klebanoff MA. Caughey AB. Grøttum P. Franchi M. Tran SH. Ethn Dis 2007. Stanley FJ. Kieke B. 39:932. Moshesh MK. J Clin Epidemiol 2006. Hoffman HJ. Review: a gentle introduction to imputation of missing values. Canada. Bonsel GJ. Stijnen T. Ethnicity and infant health in Southern Brazil. Lyell DJ.e7 Frisbie WP.16:760–766 Wong LF.115(6):710–719 Shen JJ. 6/2013 449 Downloaded by: IP-Proxy Columbia University.72: 157–162 Blackmore-Prince C. El-Sayed YY. A birth cohort study. Social and racial inequalities in preterm births in Western Australia.199(4):382. Savitz DA. Am J Obstet Gynecol 2007.e6 Liu C-Y.e1–644.9(3): 281–295 Shen TT. and other Pacific Islander mothers of single and multiple race/ethnicity: Califor- 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 Schaaf et al. Schiff MA.93(1):41–46 Coory MD. Obstet Gynecol 2001. Zaadstra BM.195(3):834–838 Aveyard P.196(2):131. Herngreen WP. Jallah Y. Howard D. van der Heijden GJ. Chang N-T.e1–373. Ananth CV. Kaufman JS.e1–385. Wingate MS. Taiwan.33 Lu MC. Racial and ethnic differences in determinants of intrauterine growth retardation and other compromised birth outcomes. Comment on: Heterogeneity in meta-analysis should be expected and appropriately quantified. Disparities in maternal outcomes among four ethnic populations. et al.59:1087–1091 Shah PS. Li J. Predicting delivery date by ultrasound and last menstrual period in early gestation. Hackmon-Ram R. Bettiol H. Duration of human singleton pregnancy. Social class. 1984 to 2006. Matern Child Health J 2008. Perinatal outcomes among different Asian-American subgroups. 30(5):1001–1008 Collins J. Public Health Rep 1995. Knowledge Synthesis Group on Determinants of Preterm/LBW Births. The risk of preterm delivery in women from different ethnic groups. MacMullen N. Chou P.e6 Stein CR. Paternal race and preterm birth. Caughey AB.110:53–58 Zanconato G. Eur J Obstet Gynecol Reprod Biol 1994.8:178–185 Alexander GR. Int J Epidemiol 2001. Bowes J. Adverse birth outcomes among Korean women compared to white women in the United States. Differences in preterm delivery rates and outcomes in Jews and Bedouins in southern Israel. Harris J. Herring AH. Eik-Nes SH. Ultrasound Obstet Gynecol 1996. Muglia LJ. Al-Wassia H. Krulewitch CJ. Ethn Dis 2005. Silva RA. Bashiri A. Krohn MA. Racial disparity in the frequency of recurrence of preterm birth. Alexander GR. Hayes DK. Caughey AB.15:492–497 Leland NL.107(10): 782–790 Rao AK.109(8): 894–899 Nystrom MJ. Chireau MV. Biegler M. Cheng KK. Obstet Gynecol 2000. Boutain D.97:189–194 Bergsjø P. et al. Boslaugh SE. Pregnancy outcomes of American Indians: contrasts among regions and with other ethnic groups. and racial disparities in adverse birth outcomes. A population-based study of race-specific risk for preterm premature rupture of membranes.e1–382. Ferre C. Iacovella C.53(2):129–134 Silva LM.e1–197. VerlooveVanhorick SP. Am J Obstet Gynecol 2006. Stamilio DM.e6 Blackmore CA. Ethnic differences in preterm birth and its subtypes: the effect of a cumulative risk profile. Clin Obstet Gynecol 2008. Makuc DM. Bergamini V. Tymkow C. racism. Elam-Evans LD. Prachand NG.87(12):1977–1983 Goedhart G. Rising disparities in severe adverse birth outcomes among Haitians in Quebec. Columbia University. Pullum SG.12(Suppl 1):S5–S11 Auger N. Am J Obstet Gynecol 2009. Pregnancy and neonatal outcomes of aboriginal women: a systematic review and metaanalysis. Perinatal outcomes among Asian-white interracial couples.17(1):113–117 Verkerk PH. Lee KA. Daniels K. Rawlings JS.76(11):1317–1321 Brown HL. 194(5):e39–e41 Rao AK. Am J Obstet Gynecol 2008. Gynecol Obstet Invest 2011. Am J Obstet Gynecol 2008. Native Hawaiian. Petersen DJ. Simon DM. Taiwan Yi Xue Hui Za Zhi 2008. Barbieri MA. Variations in pregnancy outcomes by race among 10–14-year-old mothers in the United States.21:28–39 Dominguez TP.e9 Kistka ZAF. author reply 933 American Journal of Perinatology Vol.. BJOG: An International Journal of Obstetrics and Gynaecology 2002. DeFranco EA. Shoham-Vardi I. J Immigr Minor Health 2012.3(4):189–197 Barros FC. Am J Obstet Gynecol 2008. Edwards LJ. Regan L. Horta BL. Eur J Obstet Gynecol Reprod Biol 2000. Shah V.81:65–71 Simhan HN. Racial differences in the patterns of preterm delivery in central North Carolina. Wangler MF. Preterm delivery among black and white enlisted women in the United States Army.199(4): 373. Sarno AP. A population-based study. Harlow SD. Braz J Med Biol Res 2007. Manaseki S.12:282–289 Yi JD. Moons KG. El-Sayed YY. Am J Public Health 1986. 30 No. Robinson S.40:1187–1194 Adams MM. 1995–2000. Caughey AB. Maternal ethnic ancestry and adverse perinatal outcomes in New York City. Am J Obstet Gynecol 2004.14(2):198–208 [epub ahead of print 2011 March 20] Jolly MC. nia and Hawaii. Palomar L. Matern Child Health J 2012.201(6): 584. Paediatr Perinat Epidemiol 1995.e7 Schempf AH. Chen B. Furman B. Denman DW III. Boulet S. Am J Public Health 2010. Parazzini F. Hiilesmaa V. Testing the “epidemiologic paradox” of birth outcomes among Asian immigrant women in Hsin-Chu County. Perinatal outcomes among foreign-born and US-born Chinese Americans. Racial differences in the patterns of singleton preterm delivery in the 1988 National Maternal and Infant Health Survey. et al.000 examinations. Braddock M. Kirby RS. Nassar N. Cheng YW. Race. Druzin ML.e5 Taipale P. Preterm birth among African American and white women with a lifelong residence in high-income Chicago neighborhoods: an exploratory study. Am J Obstet Gynecol 2008.24(4):352–362 Melamed Y. Ethnic differences in preterm and very preterm delivery.198(6):644. A comparison between ultrasound and a reliable last menstrual period as predictors of the day of delivery in 15. Savitz DA.51:360–370 Langridge AT.191(3): 691–699 Shiono PH. David RJ. 2003–2005. Obstet Gynecol 1993. Perinatal complications among different Asian-American subgroups. Cole FS. The “Hispanic paradox”: an investigation of racial disparity in pregnancy outcomes at a tertiary care medical center. Hamilton BE. Pregnancy outcome of migrant women delivering in a public institution in northern Italy. ethnicity and other risk factors for small for gestational age and preterm delivery in The Netherlands. Sebire N. Racial inequalities and perinatal health in the southeast region of Brazil. Acta Obstet Gynecol Scand 1990. Rhodes PH. USA. Kaimal AJ. Silva AA. Zao J. Paediatr Perinat Epidemiol 2010. A Systematic Review on Ethnic Disparities in Preterm Birth Risk . Womens Health Issues 2011. Matern Child Health J 1999. Racial and ethnic disparities in preterm birth: the 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 role of stressful life events. Int J Epidemiol 2010. Mazor M. Am J Obstet Gynecol 2006. Am J Public Health 1997. Cheng YW.100: 877–887 Li Q. BJOG: An International Journal of Obstetrics and Gynaecology 2008. 1981–2006. Meirik O. Van EM. de Turk P. Harlass FB. Jr. Read JA.199(4):385.197(2): 197. Mendola P.69:197–207 Tunón K. Reerink JD. Kugaraj KA. Perinatal outcomes among Asian American and Pacific Islander women.96:962–966 Donders AR. Am J Obstet Gynecol 2007. Chang JJ. Perinatal outcomes for Asian. Victora CG. Nakagawa S.

Raggio TP. Dahlstrom ME. Copyrighted material. J Epidemiol Community Health 2011. Prena- the “epidemiologic paradox”: birth outcomes among Latinas. Collins N. Saha S. 450 American Journal of Perinatology Vol.65:696–701 Downloaded by: IP-Proxy Columbia University. 73 Rosenberg TJ.10:351–360 75 McGlade MS.87:1022–1026 77 Ravelli AC. J Natl Med Assoc 2005. et al. 6/2013 . Tromp M. Selective migration and infant mortality among Puerto Ricans. Scrimshaw SC. 30 No. Gorman BK. Am J Public Health 2004. Eskes M.A Systematic Review on Ethnic Disparities in Preterm Birth Risk Schaaf et al. Columbia University.94:2062–2065 tal health behaviors and psychosocial risk factors in pregnant women of Mexican origin: the role of acculturation. The Latina paradox: an opportunity for restructuring prenatal care delivery. Ethnic differences in stillbirth and early neonatal mortality in The Netherlands. Chiasson MA. Dunkel-Schetter C. A further examination of 76 Zambrana RE. Am J Public Health 1997.97:550–556 74 Landale NS. Oropesa RS. Matern Child Health J 2006.