Preterm newborns account for more than 70% of the neonatal morbidity and mortality in the United States and are 40 times more likely to die in the neonatal period than their term counterparts. More than 12% of children born in the United States are delivered preterm, defined as less than 37 weeks of gestation. However, preterm birth is substantially more common in African Americans (17.8% compared with 11.5% in European Americans) even when controlling for social and other confounding factors.1–6 The risk of spontaneous preterm birth is increased in interracial couples; the risk is higher in couples in which the mother is African American and the father white (compared with couples in which the mother is white and the father is African American), suggesting an increased contribution of maternal genes.7
Numerous studies have demonstrated that the susceptibility to spontaneous preterm birth is inherited. It is frequently clustered in families.8 Association studies have identified a number of genes with increased genetic variation among women delivering prematurely.9–12 Other researchers have specifically investigated potential genetic explanations for the overrepresentation of preterm birth in African American women and found that the genetic contribution to prematurity varies with race, particularly in pathways related to inflammation and infection.13–15
Despite evidence regarding the heritability of spontaneous preterm birth, association studies and family studies have identified very few genes associated with this phenotype. Admixture mapping is an efficient and potentially powerful method to scan the genome and identify genes that cause diseases whose prevalence rates vary among populations.16 This technique is suitable when the frequency of a disease is markedly different in two different populations and when genetic mixture (admixture) has occurred between the populations. The underlying rationale is that one or more disease-associated loci are more common in one ancestral population (in this case, African Americans) than in the other (Europeans). If so, a region of the genome that contains such a locus is likely to be of recent African (rather than European) origin.17,18 In a sample of admixed (African American) individuals with spontaneous preterm birth, admixture mapping attempts to pinpoint such a region or regions. This is accomplished by assaying single nucleotide polymorphisms (SNPs) whose frequencies are known to differ strongly in Africans and Europeans. In every admixed individual, some regions of DNA will be of African origin, whereas others will be of European origin. Regions of potential disease loci may have higher frequencies of “African” SNPs compared with regions that do not contain a spontaneous preterm birth locus.16,18
Because European and African populations mixed only recently in the United States (within approximately the last 15 generations), stretches of DNA with contiguous European and African ancestry have not had time to undergo extensive recombination. Thus, a relatively small number of polymorphisms are needed to isolate a disease gene-containing region; SNPs covering several million base pairs of DNA are likely to be strongly associated with a disease-causing gene. This is in contrast to traditional linkage disequilibrium mapping, which requires analysis of SNPs every few thousand base pairs. A traditional control group is not needed; the “control” in this case is the average proportion of African ancestry across the entire genome.16,17,19–21
Admixture mapping has been successfully used to identify genes that contribute to other complex diseases, including prostate cancer, type II diabetes, autoimmune diseases, and obesity.20,22–24 For example, a strong association between a 3.8-Mb chromosomal region (8q24) and susceptibility to prostate cancer was recently found using this method.25
We hypothesized that the disparity between rates of preterm birth in African American and European American women is attributable, at least in part, to genetic differences and that these differences can be detected using admixture mapping. Finding a disease-associated region through admixture mapping will lead to more in-depth, focused studies aimed at uncovering direct genetic causes of spontaneous preterm birth.
MATERIALS AND METHODS
This is a secondary analysis of women enrolled in a multicenter, prospective, double-blind randomized controlled trial of 17-α hydroxyprogesterone caproate for the prevention of recurrent preterm birth conducted from 1999 to 2002 by the Eunice Kennedy Shriver National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network.26 As a part of the original trial protocol, maternal saliva samples were collected for future analyses. Saliva samples were originally labeled with unique, deidentified study codes and frozen at −20°C. DNA was extracted and whole-genome amplified from samples using established methods in July and August 2008. Institutional review board approval and patient consent were obtained at each of the 19 participating network sites for the original study; institutional review board approval for this secondary analysis was also obtained.
All women admitted to the original trial had at least one documented spontaneous preterm birth at less than 37 weeks of gestation. Samples from self-identified African American women were included. There were no other exclusion criteria. Self-identified African American women were genotyped with the 1,509 SNP Illumina African American Admixture Panel. Single nucleotide polymorphisms included on this panel have high minor allele frequency differences between African and European American populations; allele frequency differences of greater than 0.60 are seen for the majority of SNPs.
Women were analyzed as a group (spontaneous preterm birth at less than 37.0 weeks of gestation) and also stratified by preterm birth severity. Women with more severe phenotypes, those with one or more spontaneous preterm birth at less than 35 and less than 32 weeks of gestation, were also analyzed separately. Women with an early spontaneous preterm birth, less than 35 or less than 32 weeks of gestation, either before entry into or during their participation in the original trial were included in the severe phenotype groups.26
The cohort of women in this analysis was compared with the original cohort using chi square and Student t test. Genotypes were analyzed using ANCESTRYMAP software 2.0.16 A risk parameter of 1.5 was used, because the risk of preterm birth among African American women is approximately 1.5 times the risk in European Americans; this is a conservative estimate.1,4,5 Markers were tested for Hardy-Weinberg equilibrium and those with P<.01 were excluded. A logarithm of odds locus-genome score was calculated for each SNP marker. A logarithm of odds score is based on a likelihood ratio statistic that compares, for each point on the genome, the likelihood of data being from a disease locus compared with the likelihood of data being from a locus unrelated to disease.16 A higher logarithm of odds score (2 or higher) represents a higher probability that the point is associated with a disease locus.
Initial results were confirmed with a larger number of burn-in iterations (500) and follow-on iterations (500) per ANCESTRYMAP guidelines. Those areas with logarithm of odds scores 1.50–1.99 were considered suggestive of a disease locus, and those areas with scores 2 or greater were considered significant for a disease locus.16,27
There were 177 of 271 (65%) self-identified African American women from the original trial who had DNA available. These 177 women included in our study were similar to the original cohort of African American women with respect to mean maternal age (25.1±5.0 compared with 24.9±5.0 years, P=.48), randomization to 17-α hydroxyprogesterone caproate (68.9% compared with 66.8%, P=.31), history of more than one prior spontaneous preterm birth (33.3% compared with 33.6%, P=.91), tobacco use (24.9% compared with 21.8%, P=.09), and mean prepregnancy body mass index (28.3±8.4 compared with 27.6±8.1, P=.06; body mass index is calculated as weight (kg)/[height (m)]2). However, women in the cohort with DNA available were less likely to have delivered prematurely during the Meis trial compared with the original cohort (preterm birth at less than 37 weeks of gestation 35.0% compared with 41.0%, P=.006; and preterm birth at less than 32 weeks of gestation 11.3% compared with 15.9%, P=.005).
Nine of 1,509 SNPs (less than 1%) failed and produced no calls; an additional 50 markers (3%) had P<.01 on Hardy-Weinberg exact testing and were excluded. Thus, 1,450 SNPs (96%) were included in this analysis. The average African ancestry across all markers was 82.5% (median 82.3%; range 78.3–87.9%). To illustrate how ancestry proportions are estimated, consider a single SNP, rs10488004. The minor allele, G, has a frequency of 0.14 in African Americans and the major allele, A, has a frequency of 0.86. In contrast, for the same SNP, A is the minor allele in European Americans (frequency 0.28) and G is the major allele (frequency 0.72). An individual carrying two copies of A at rs10488004 would be more likely to have inherited this allele because of recent African ancestry, whereas an individual who carries two copies of G would be more likely to have received it because of recent European ancestry. In practice, thousands of SNPs are used to derive an overall estimate of ancestry proportions. Logarithm of odds locus-genome scores were calculated for each SNP marker. Among the entire cohort of 177 women with a prior spontaneous preterm birth at less than 37 weeks of gestation, three SNPs had suggestive logarithm of odds locus-genome scores (Table 1). All three SNPs correspond to the chromosome region of 7q21. When logarithm of odds scores are graphed by physical location in this region of chromosome 7, a distinctive gradual peak is seen (Fig. 1). This area is suggestive of a preterm birth locus.
One hundred fifty-two women had one or more spontaneous preterm births at less than 35 weeks of gestation. Among these women with earlier preterm birth, the signal strengthened and eight SNPs (all in the 7q21 and 7q22 chromosome region) had logarithm of odds scores greater than 1.5; four of these logarithm of odds scores were greater than 2. Single nucleotide polymorphisms with logarithm of odds scores greater than 1.5 are shown in Table 2 and chromosome location in Figure 2. Similar to the results for the entire cohort, a distinctive gradual peak is seen, suggestive of a preterm birth locus.
One hundred six women had one or more spontaneous preterm births at less than 32 weeks of gestation. These women with the earliest prior spontaneous preterm birth had the strongest signal with 15 SNPs with logarithm of odds scores greater than 1.5; six of these logarithm of odds scores were greater than 2.0 (Table 3). In this group with the most severe phenotype (less than 32 weeks of gestation), the majority of the significant results were again seen in the 7q21 and 7q22 region; however, four significant results corresponded to the region of 7q31, located directly adjacent to 7q22 (Fig. 3). Again, the peak in Figure 3 represents an area of increased African ancestry in this group of women with very early preterm birth and is suggestive of one or more preterm birth loci.
No other significant logarithm of odds scores were noted on any other chromosomes for the entire cohort or either of the earlier preterm birth groups.
In this admixture mapping study of 177 African American women with well-documented spontaneous preterm birth, we found chromosome 7q21-7q22 to be an area of potential preterm birth susceptibility loci. Importantly, the number of significant results increased as the clinical phenotype worsened; women with the earliest spontaneous preterm birth had the strongest signal in this region (Figs. 1–3).
Emerging evidence supports the contribution of genetics to the pathogenesis and predisposition to spontaneous preterm birth. Spontaneous preterm birth is a complex phenotype and is unlikely to be a “single gene” condition. Given the variation in rates of preterm birth with self-reported race or ethnicity, it is reasonable to conclude that some of these differences may be the result of genetic factors.5 Although genetic differences between African American and European American women with spontaneous preterm birth have been pronounced in other studies, our current knowledge of the exact genetic variants that contribute to preterm birth is limited.14,15 Typically, studies have focused on examining variation in specific candidate genes thought to be involved in the pathogenesis of preterm birth. However, it is clear that a significant knowledge gap exists, because even the strongest case–control studies report population attributable risks of less than 30%.28 One previous small study (examining only 61 ancestry informative markers) of preterm birth among African American mothers found significant associations between preterm birth as a whole and preterm birth resulting from maternal hypertensive disorders.1 The authors concluded that more intensive investigations of admixture are needed to identify novel preterm birth susceptibility genes. Thus, studies such as this one have the potential to make significant contributions to the overall understanding of preterm birth pathogenesis by identifying different, understudied genetic areas of interest.
Chromosome 7 in the region of 7q21 and 7q22 is a gene-rich area. Although admixture mapping is unable to identify specific genes in this area that may contribute to spontaneous preterm birth among African American women, several potential candidate genes are located in this region, including: metabolic genes such as CYP51A1 and CYP3A (cytochrome P450 family 51A1 and 3A); inflammatory and immunomodulatory genes such as STEAP4 (tumor necrosis factor-α-induced protein 9), TFPI2 (tissue factor pathway inhibitor 2), PILRA and PILRB (paired immunoglobulin-like type 2 receptor α and β), calcium regulation genes such as CALCR (calcitonin receptor), CACNA2D1 (voltage-dependent calcium channel α), MYLC2PL (myosin light chain 2 precursor), collagen genes such as COL1A2 (α-2 type 1 collagen), and PCOLCE (procollagen C-endopeptidase enhancer).
Numerous other regulatory and signaling genes are encoded in this region, including zinc-finger proteins and adenosine triphosphate synthases.
Our population, with a median of approximately 82% African ancestry, is consistent with prior reports of other groups of admixed African Americans.1,29 Furthermore, the accuracy of self-reported race in our cohort was high; all of the self-identified African Americans had at least 78.3% African ancestry. All women in this cohort had prospectively collected clinical outcomes and all “index” preterm births permitting entry into the initial trial were documented and verified by trained researchers. We used a “cutoff” locus-genome score of 1.5 to consider a marker “‘suggestive” of a disease locus, more stringent than the cutoff of 1.0 used in other studies.27 Interestingly, because the analysis was restricted to include only the most “severe” women (preterm birth at less than 32 weeks of gestation), we noted the strongest signals and the most suggestive and significant results, as illustrated in Figure 3. Additionally, we did not note any “spurious” results or single SNP “hits” in other areas of the genome.
The Illumina admixture panel used for this study is an established set of SNPs developed by leaders in the field of admixture mapping (Reich and Patterson).30,31 The targeted loci on this African American admixture panel yields approximately 75% of the power for admixture mapping, equivalent to 300,000–1,000,000 loci that would be required for dense, nonspecific whole genome scans.31
It is possible that a few patients may have been excluded from the severe (preterm birth at less than 35 and less than 32 weeks) cohorts resulting from the beneficial effects of 17-α hydroxyprogesterone caproate, provided to two-thirds of women in the original trial. However, allowing the gestational age of the qualifying preterm birth (before enrollment in the Meis trial) to also classify a women as “severe” likely minimized this effect.
These results are encouraging and help to isolate the general region of the genome that is different in African Americans with spontaneous preterm birth. However, admixture mapping is not able to identify the specific gene or genes in this region that may contribute to spontaneous preterm birth. All, some, or none of the genes listed may or may not be involved in the pathogenesis of spontaneous preterm birth in African American women. Areas of the genome identified by admixture mapping require more in-depth analysis to isolate specific gene(s) of interest. Although we only found results on chromosome 7, as a result of the sample size limitations of this study, we cannot definitively exclude the contribution of genes on other chromosomes to the disparity in preterm birth rates between African and European Americans. Although a post hoc power analysis would not be appropriate, Patterson et al16 estimates that for a population with 80% African ancestry (like in this study), approximately 200 patients are necessary to detect loci in which the relative risk of disease resulting from an allele is 2. Additionally, because specific genetic variants cannot be identified from admixture mapping, the study design does not permit us to directly assess gene–environment interactions, which may confer additional risk to some women.
Future studies should confirm these results in other cohorts of women and should further examine this region of the chromosome using case–control methods to identify specific genetic variants associated with spontaneous preterm birth.
1. Tsai HJ, Yu Y, Zhang S, Pearson C, Ortiz K, Xu X, et al. Association of genetic ancestry with preterm delivery and related traits among African American mothers. Am J Obstet Gynecol 2009;201:94.e1–10.
2. Blackmore-Prince C, Kieke B Jr, Kugaraj KA, Ferre C, Elam-Evans LD, Krulewitch CJ, et al. Racial differences in the patterns of singleton preterm delivery in the 1988 National Maternal and Infant Health Survey. Matern Child Health J 1999;3:189–97.
3. Behrman RE, Butler AS, editors. Preterm birth: cause, consequence, and prevention. Washington (DC): Institute of Medicine of the National Academies (Committee on Understanding Premature Birth and Assuring Healthy Outcomes Board on Health Science Policy); 2006.
4. Schieve LA, Handler A. Preterm delivery and perinatal death among black and white infants in a Chicago-area perinatal registry. Obstet Gynecol 1996;88:356–63.
5. Ananth CV, Misra DP, Demissie K, Smulian JC. Rates of preterm delivery among Black women and White women in the United States over two decades: an age-period-cohort analysis. Am J Epidemiol 2001;154:657–65.
6. Aveyard P, Cheng KK, Manaseki S, Gardosi J. The risk of preterm delivery in women from different ethnic groups. BJOG 2002;109:894–9.
7. Palomar L, DeFranco EA, Lee KA, Allsworth JE, Muglia LJ. Paternal race is a risk factor for preterm birth. Am J Obstet Gynecol 2007;197:152.e1–7.
8. Esplin MS. Preterm birth: a review of genetic factors and future directions for genetic study. Obstet Gynecol Surv 2006;61:800–6.
9. Crider KS, Whitehead N, Buus RM. Genetic variation associated with preterm birth: a HuGE review. Genet Med 2005;7:593–604.
10. Gibson CS, MacLennan AH, Dekker GA, Goldwater PN, Dambrosia JM, Munroe DJ, et al. Genetic polymorphisms and spontaneous preterm birth. Obstet Gynecol 2007;109:384–91.
11. Murtha AP, Nieves A, Hauser ER, Swamy GK, Yonish BA, Sinclair TR, et al. Association of maternal IL-1 receptor antagonist intron 2 gene polymorphism and preterm birth. Am J Obstet Gynecol 2006;195:1249–53.
12. Ryckman KK, Morken NH, White MJ, Velez DR, Menon R, Fortunato SJ, et al. Maternal and fetal genetic associations of PTGER3 and PON1 with preterm birth. PLoS One 2010;5:e9040.
13. Anum EA, Springel EH, Shriver MD, Strauss JF 3rd. Genetic contributions to disparities in preterm birth. Pediatr Res 2009;65:1–9.
14. Fortunato SJ, Menon R, Velez DR, Thorsen P, Williams SM. Racial disparity in maternal-fetal genetic epistasis in spontaneous preterm birth. Am J Obstet Gynecol 2008;198:666.e1–9; discussion 666.e9–10.
15. Menon R, Pearce B, Velez DR, Merialdi M, Williams SM, Fortunato SJ, et al. Racial disparity in pathophysiologic pathways of preterm birth based on genetic variants. Reprod Biol Endocrinol 2009;7:62.
16. Patterson N, Hattangadi N, Lane B, Lohmueller KE, Hafler DA, Oksenberg JR, et al. Methods for high-density admixture mapping of disease genes. Am J Hum Genet 2004;74:979–1000.
17. Smith MW, Patterson N, Lautenberger JA, Truelove AL, McDonald GJ, Waliszewska A, et al. A high-density admixture map for disease gene discovery in african americans. Am J Hum Genet 2004;74:1001–13.
18. Nievergelt CM, Schork NJ. Admixture mapping as a gene discovery approach for complex human traits and diseases. Curr Hypertens Rep 2005;7:31–7.
19. Smith MW, Lautenberger JA, Shin HD, Chretien JP, Shrestha S, Gilbert DA, et al. Markers for mapping by admixture linkage disequilibrium in African American and Hispanic populations. Am J Hum Genet 2001;69:1080–94.
20. Reich D, Patterson N. Will admixture mapping work to find disease genes? Philos Trans R Soc Lond B Biol Sci 2005;360:1605–7.
21. Montana G, Pritchard JK. Statistical tests for admixture mapping with case-control and cases-only data. Am J Hum Genet 2004;75:771–89.
22. Basu A, Tang H, Arnett D, Gu CC, Mosley T, Kardia S, et al. Admixture mapping of quantitative trait loci for BMI in African Americans: evidence for loci on chromosomes 3q, 5q, and 15q. Obesity (Silver Spring) 2009;17:1226–31.
23. Cheng CY, Reich D, Coresh J, Boerwinkle E, Patterson N, Li M, et al. Admixture mapping of obesity-related traits in African Americans: the Atherosclerosis Risk in Communities (ARIC) Study. Obesity (Silver Spring) 2010;18:563–72.
24. Deo RC, Patterson N, Tandon A, McDonald GJ, Haiman CA, Ardlie K, et al. A high-density admixture scan in 1,670 African Americans with hypertension. PLoS Genet 2007;3:e196.
25. Freedman ML, Haiman CA, Patterson N, McDonald GJ, Tandon A, Waliszewska A, et al. Admixture mapping identifies 8q24 as a prostate cancer risk locus in African-American men. Proc Natl Acad Sci U S A 2006;103:14068–73.
26. Meis PJ, Klebanoff M, Thom E, Dombrowski MP, Sibai B, Moawad AH, et al. Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate. N Engl J Med 2003;348:2379–85.
27. Reich D, Patterson N, De Jager PL, McDonald GJ, Waliszewska A, Tandon A, et al. A whole-genome admixture scan finds a candidate locus for multiple sclerosis susceptibility. Nat Genet 2005;37:1113–8.
28. Wang H, Parry S, Macones G, Sammel MD, Kuivaniemi H, Tromp G, et al. A functional SNP in the promoter of the SERPINH1 gene increases risk of preterm premature rupture of membranes in African Americans [published erratum appears in Proc Natl Acad Sci U S A 2006;103:19212]. Proc Natl Acad Sci U S A 2006;103:13463–7.
29. Bock CH, Schwartz AG, Ruterbusch JJ, Levin AM, Neslund-Dudas C, Land SJ, et al. Results from a prostate cancer admixture mapping study in African-American men. Hum Genet 2009;126:637–42.
30. Tandon A, Patterson N, Reich D. Ancestry informative marker panels for African Americans based on subsets of commercially available SNP arrays. Genet Epidemiol 2011;35:80–3.
Supplemental Digital Content
© 2011 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.