OBJECTIVE: We investigated the relationship between maternal race and stillbirth among singletons, twins, and triplets.
METHODS: We conducted a retrospective cohort study on 14,348,318 singletons, 387,419 twins, and 20,953 triplets delivered in the United States from 1995 through 1998. We compared the risk of stillbirth between pregnancies of black and those of white mothers using the generalized estimating equations framework to adjust for intracluster correlation in multiples.
RESULTS: The proportion of black infants was 16%, 18%, and 8% among singletons, twins, and triplets, respectively. Crude stillbirth rate among singletons was 6.6 per 1,000 and 3.5 per 1,000 for black and white fetuses, respectively. Among twins, 796 stillbirths (11.6 per 1,000) were recorded for black mothers versus 3,209 stillbirths (10.1 per 1,000) among white mothers, whereas among triplets there were 233 stillbirths, of which 39 stillbirths were black fetuses (24.6 per 1,000) and 194 stillbirths were white fetuses (10.0 per 1,000). Black singletons, twins, and triplets weighed 278 g, 186 g, and 216 g less than white fetuses, respectively (P < .001). Risk of stillbirth was elevated in black fetuses compared with white fetuses among singletons (adjusted odds ratio [OR] 2.9, 95% confidence interval [CI] 2.8–3.0) and twins (OR 1.3. 95% CI 1.2–1.4) but comparable among triplets (OR 1.2, 95% CI 0.7–2.1). This decreasing trend was significant (P for trend < .001).
CONCLUSION: The disparity of stillbirths between black and white fetuses still persists among singletons and twins. Among triplet gestations, however, the 2 racial groups have a comparable risk level. Our findings highlight the need for a rigorous research agenda to elucidate causes of stillbirth across racial/ethnic entities in the United States.
LEVEL OF EVIDENCE: II-2
Black mothers have a significantly higher risk of stillbirth among singletons and twins but comparable risk levels among triplets compared with white mothers.
From the *Department of Maternal and Child Health and †Department of Epidemiology, University of Alabama at Birmingham.
Partly supported through a Young Clinical Scientist Award to the first author (H.M.S.) by the Flight Attendant Medical Research Institute (FAMRI).
Address reprint requests to: Hamisu Salihu, MD, PhD, Department of Maternal and Child Health, University of Alabama at Birmingham, 1665 University Boulevard, Room 320, Birmingham, AL 35294; e-mail: firstname.lastname@example.org.
Received May 23, 2004. Received in revised form June 29, 2004. Accepted July 1, 2004.
Racial disparity in health status between blacks and whites in the United States is a priority area and focus for Healthy People 2010.1 Black infants have consistently higher rates of low birth weight, premature birth, and mortality compared with white infants.2,3 Although the rates of infant mortality have decreased during the course of time for the entire population, black infants remain at increased risk in part because the rate of decline for this outcome has been slower for black than white infants.2 Previous studies have demonstrated that the rate of stillbirths among black singletons is approximately twice that of white singletons.3–8 However, the magnitude of the racial gap in stillbirth among multiples is poorly understood. The estimation of the risk among multiples is indicated and timely because multiples have increasingly become an important risk entity as a result of the current epidemic of multiple gestations in the United States.9 Extending our research focus to encompass multiples has the potential of improving our understanding of the determinants of stillbirth across population subgroups. To this effect, we investigated disparities between black and white mothers in the occurrence of stillbirth across singleton, twin, and triplet gestations.
MATERIALS AND METHODS
We used 2 data files for the analysis in this study. The first was the “matched multiple birth file” assembled by the National Center for Health Statistics covering the period 1995–1998, which contains the most recent matched and linked data for multiple deliveries in the United States.10 The file contains individual records of live births and fetal deaths involving multiple deliveries that occurred in the United States during the stated period. For analyses involving singletons, we used the natality and fetal death data files for 1995–1998. The procedures for quality control of the data are explained in detail elsewhere.10,11 Perfect matching of the multiple files was achieved for approximately 99% of the records, and the record linkage process has been adequately validated.10 The data source forms the basis for official U.S. birth and death statistics.
The primary aim of this study was to investigate the influence of maternal race on stillbirth in singleton, twin, and triplet gestations. We defined maternal race as black or white regardless of ethnicity. We compared the following sociodemographic characteristics between white and black mothers: maternal age, educational level attained, parity (nulliparous, multiparous), marital status (married, single), reported use of tobacco during pregnancy (yes, no), and adequacy of prenatal care (adequate and inadequate). Parity as used here refers to the total number of live deliveries the mother had experienced. Adequacy of prenatal care was assessed using the revised graduated index algorithm,12,13 which has been found to be more accurate than several others, especially in describing the level of prenatal care use among groups that are at high risk and therefore exposed to intense care (eg, multiple pregnancies).14 The revised graduated index assesses the adequacy of care based on 3 variables (trimester prenatal care began, number of visits, and the gestational age of the infant at birth). In this study, inadequate prenatal care use refers to women who either had missing prenatal care information, had prenatal care but the level was considered suboptimal, or had no prenatal care at all. The accuracy of all these aforementioned sociodemographic variables on the birth certificate has been validated in previous studies.15,16
The main outcome of interest was occurrence of stillbirth, which we defined as intrauterine fetal death at 24 weeks of gestation or greater. We chose this gestational age cutoff point because after 24 weeks, deaths are generally well reported.17,18 We computed gestational age and birth weight-specific stillbirth rates by dividing the number of stillbirths by the sum of live births and stillbirths in the given gestational age or birth weight category. For singletons and twins, we created a priori 4 gestational age categories (24–27, 28–31, 32–35, and 36 weeks of gestation and greater) whereas for birth weight, 6 categories were constructed (< 1,000, 1,000–1,499, 1,500–1,999, 2,000–2,499, 2,500–2,999, and 3,000 g and greater). For triplets, however, the number of gestational age and birth weight categories was limited to 3 and 4, respectively, because of the paucity of data beyond 35 weeks of gestation and 2,000 g, especially among black infants.
Adjusted estimates for stillbirth were generated by using the logistic regression model for singletons and multiples. Furthermore, we used the methodology of generalized estimating equations19 to adjust for the presence of intracluster correlation among twins and triplets. PROC GENMOD in SAS 9.0 (SAS Institute, Cary, NC) was used to model all outcomes of interest. We determined the level of excess stillbirths associated with maternal race (%) among black singletons, twins, and triplets as compared with whites by using the following formula:20
where RR = adjusted relative risk of stillbirth.
We used adjusted odds ratios to approximate the RR. The excess stillbirths denote the magnitude of stillbirths among blacks over and above what would have been the expected level. The χ2 statistic was applied to assess linear trend.20 For continuous variables, the student t test was applied to compare racial differences in means. All tests of hypothesis were 2-tailed with a type 1 error rate fixed at 5%. This study was approved by the Institutional Review Board at the University of Alabama at Birmingham.
We analyzed a total of 14,348,318 singletons, 387,419 individual twins, and 20,953 triplets. The proportion of black infants was 16.3%, 17.6%, and 7.6% among singletons, twins, and triplets, respectively. Differences in selected sociodemographic characteristics between mothers of black and mothers of white infants are summarized in Table 1. Black mothers were more likely to be younger, less educated, multiparous, and unmarried and to have received inadequate level of prenatal care compared with their white counterparts irrespective of plurality. White mothers had a higher prevalence for prenatal smoking except among triplets, for whom the level of prenatal smoking was significantly higher among blacks.
For singletons, a total of 57,282 and 24,396 stillbirths were recorded for white and black mothers, respectively. Among twins, 4,005 stillbirths were observed, of which 796 were among black mothers and 3,209 were among white mothers. Among triplet pregnancies, there were 233 stillbirths consisting of 39 stillbirths among black mothers and 194 stillbirths among white mothers. The corresponding crude stillbirth rates for the 2 racial groups are illustrated in Figure 1.
Significant differences in gestational age and birth weight were observed between infants of black and those of white mothers. Among singleton gestations, black neonates were born 4 days earlier (mean ± standard deviation for blacks, 38.4 ± 3.0 days versus 39.1 ± 2.3 days for whites) and weighed at least 278 g less (3,104.5 ± 692.9 g for blacks versus 3,382.5 ± 591.6 g for whites). Also, for twin gestations, black infants were born at least 4 days earlier than whites (35.2 ± 3.9 weeks for blacks versus 35.8 ± 3.2 weeks for whites). About the same gestational age difference was also detected among triplets (31.8 ± 3.7 weeks for blacks versus 32.6 ± 3.3 weeks for whites). Twins born to black mothers weighed on average 186 g less than their white counterparts (2,246 ± 767 g for blacks versus 2,439 ± 734 g for whites). Among triplets, the mean difference was 216 g, to the disadvantage of black infants (1,534 ± 553 g for blacks versus 1,750 ± 547 g for whites).
Gestational age–specific stillbirth rates are presented in Table 2. Among singletons and twins younger than 32 gestational weeks, white fetuses had significantly higher stillbirth rates than black fetuses, whereas with an increase in gestational age beyond 32 weeks, the relationship was reversed to the disadvantage of black fetuses. However, a different pattern was noted in triplets. At lower gestational ages, black triplets had greater rates of stillbirth whereas the stillbirth rates for both racial groups were comparable beyond 32 weeks. A similar pattern was observed for birth weight–specific stillbirth rates (Table 3). For singletons and twins, black fetuses were advantaged at lower birth weights, but the pattern was reversed at birth weights greater than 2,000 g in a linear dose-dependent fashion. However, racial disparity for birth weight–specific stillbirth among triplets was U-shaped. Blacks had higher rates for stillbirth at the extremes of birth weights and lower rates in between although these differences reached significant levels only for a birth weight less than 1,000 g.
Crude and adjusted odds ratios for stillbirths among blacks are presented in Table 4. The adjusted odds ratios were obtained by controlling for the following factors: maternal age, parity, educational status, cigarette smoking, and prenatal care. Black fetuses had an elevated risk of intrauterine demise compared with white fetuses among singletons and twins, whereas the risk was comparable among triplets. The level of excess cases of stillbirth among black fetuses reflects the magnitude of the racial disparity in birth outcomes.
We found significant differences among black and white mothers in the risk of stillbirth among singletons and twins, whereas the racial disparity among triplets was not significant. Our results among singletons are in accord with those of previous studies.3,5,7,8,21 Risk estimates in these studies range from 1.6- to 2.4-fold, which is statistically comparable with our adjusted odds ratio of 2.9. The slight difference in the magnitude may be partially attributable to the fact that ours is a population-wide study with sufficient sample size to permit control of several confounding factors. Many of the earlier studies were either clinically or geographically limited in scope or did not provide adjusted estimates.
We also detected a 23% level of excess stillbirth among black twins compared with white twins. Previous studies have shown that the gap between white and black mothers regarding infant mortality was wider among singleton births than in twin gestations.22 Our findings among twins also are in agreement with a population-based report23 that observed a lower level of disparity among black and white mothers concerning fetal death among twins than in singletons. Despite a 17% level of excess stillbirths among black triplets, our results were not statistically significant, probably as the result of insufficient sample size. Another possible reason why the results in triplets did not reach a statistically significant level could be the statistical technique used, which was appropriate albeit conservative; the use of this technique could be considered another point of strength of this study. The use of the generalized estimating equation provided conservative (wider) confidence intervals because of adjustment for intracluster correlations.24 Whatever the reasons, we feel that the 20% higher risk of stillbirth is meaningful from both clinical and public health perspectives.
The disparity among black and white mothers was observed in mean gestational ages and birth weights for all plurality subtypes. The magnitude of these differences among singletons and twins albeit statistically significant, may not have considerable clinical bearing.
We found variation in the degree of disparity among black and white mothers in stillbirth within gestational age and birth weight strata. Among singletons and twins, the rate of stillbirths was higher in white fetuses before 32 weeks of gestation and was shifted to the disadvantage of black fetuses from 32 weeks of gestation and onward. A similar pattern was observed for increasing birth weight. This “crossover effect,” where black infants have better survival at lower birth weights but worse survival at higher weights, has been reported previously.3 On the contrary, black fetuses had higher rates of stillbirth among triplets before 32 weeks of gestation and lower thereafter. The findings among triplets by birth weight were not consistent. One possible explanation for the results among singletons and twins may be the preponderance of cases of late stillbirth among black fetuses compared with white fetuses. Although approximately one third of the causes of stillbirth remain unknown, hypertension in pregnancy, including pregnancy toxemia, has been identified as the leading known cause of stillbirth among black fetuses.5 Compared with white women, the proportion of stillbirths accounted for by hypertension among black women is 10 times that of white women.5 Because hypertension-associated stillbirth is mediated mainly through fetal growth restriction, this could explain the preponderance of late-onset excess stillbirth among black singletons and twins. The reverse finding in triplets is difficult to explain but could be another indication of the heterogeneous nature of the etiologies of stillbirth across plurality subtypes.
Our findings also highlight the importance of adjusting for influential factors of risk in determining an overall RR estimate. For singletons and triplets, the odds ratio changed appreciably (from 1.9 [crude] to 2.9 [adjusted] and from 2.5 [crude] to 1.2 [adjusted] for singletons and triplets, respectively) after controlling for negative risk factors that were preponderant among black women (eg, lower education and suboptimal prenatal care). In contrast, only a slight rise was noted for twins (from 1.1 [crude] to 1.3 [adjusted]). Hence, our ability to control for several confounders across plurality categories as well as the adjustment for intracluster source of bias among multiples represents distinct strengths of this study.
One limitation of this study was our inability to provide information on causes of stillbirths because the relevant data as indicated on birth certificates are not reliable. In a study conducted in 5 states, Kirby25 reported that approximately 30–45% of reports of causes of fetal death from vital records were not valid or useful. In addition, the National Center for Health Statistics does not provide underlying cause of death on its public-use data files. Consequently, we did not analyze causes of stillbirth to avoid disseminating misleading or biased results. Another limitation of the data was the lack of information on the use of assisted reproduction technology, especially for triplet gestations. Despite these limitations, to our knowledge this is the most comprehensive study to date that addresses racial disparities in the occurrence of stillbirth across singletons, twins, and triplets within the same study.
The results in this study have clinical and public health implications. The prevention of stillbirths is closely linked to our understanding of its etiology, which from this study, appears to be influenced by race. In the battle to curtail the occurrence of stillbirth, the identification of high-risk subpopulations is critical so that targeted interventions can be efficiently implemented to enhance birth outcomes.
In summary, we found that the gap between black and white mothers in the incidence of stillbirth still persists among singletons and twins. Although the difference detected among triplets was not statistically significant, it is important in magnitude. Our findings illustrate the need for more research in this direction.
1. U.S. Department of Health and Human Services. Healthy People 2010. 2nd ed. Understanding and improving health and Objectives for improving health. 2 vols. Washington, DC: U.S. Government Printing Office; 2000.
2. Hoyert DL. Perinatal mortality in the United States, 1985–91. National Center for Health Statistics. Vital Health Stat 1995;20(26):1–26.
3. Hsieh H, Lee K, Khoshnood B, Herschel M. Fetal death rate in the United States, 1979-1990: trend and racial disparity. Obstet Gynecol 1997;89:33–9.
4. Copper RL, Goldenberg RL, DuBard MB, Davis RO. Risk factors for fetal death in white, black, and Hispanic women: The Collaborative Group on Preterm Birth Prevention. Obstet Gynecol 1994;84:490–5.
5. Herschel M, Hsieh H, Mittendorf R, Khoshnood B, Covert RF, Lee K. Fetal death in a population of Black women. Am J Prev Med 1995;11:185–91.
6. Herschel M, Hsieh H, Mittendorf R, Khoshnood B, Covert RF, Lee K. Risk factors for fetal death in white, black, and Hispanic women. Obstet Gynecol 1995;85:318–9.
7. Guendelman S, Chavez G, Christianson R. Fetal deaths in Mexican American, black and white non-Hispanic women seeking government-funded prenatal care. J Community Health 1994;19:319–30.
8. Buck GM, Shelton JA, Mahoney MC, Michalek AM, Powell EJ. Racial variation in spontaneous fetal deaths at 20 weeks or older in upstate New York, 1980-86. Public Health Rep 1995;110:587–92.
9. Salihu HM, Aliyu MH, Rouse DJ, Kirby RS, Alexander GR. Potentially preventable excess mortality among higher-order multiples. Obstet Gynecol 2003;102:679–84.
10. Martin J, Curtin S, Saulnier M, Mousavi J. Development of the Matched Multiple Birth File. In: 1995-1998 Matched Multiple Birth dataset. NCHS CD-ROM series 21, no. 13a. Hyattsville, MD: National Center for Health Statistics; 2003.
11. National Center for Health Statistics. 1995-1998 linked birth/infant death data set. Vital Statistics of the United States: Quality Control Procedures. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2000.
12. Alexander GR, Kotelchuck M. Quantifying the adequacy of prenatal care: a comparison of indices. Public Health Rep 1996;3:408–18; discussion 419.
13. Alexander GR, Cornely DA. Prenatal care utilization: its measurement and relationship to pregnancy outcome. Am J Prevent Med 1987;3:243–53.
14. Kogan MD, Martin JA, Alexander GR, Kotelchuck M, Ventura SJ, Frigoletto FD. The changing pattern of prenatal care utilization in the United States, 1981-1995, using different prenatal care indices. JAMA 1998;279:1623–8.
15. Buescher PA, Taylor KP, Davis MH, Bowling JM. The quality of the new birth certificate data: a validation study in North Carolina. Am J Public Health 1993;83:1163–5.
16. DiGiuseppe DL, Aron DC, Ranbom L, Harper DL, Rosenthal GE. Reliability of birth certificate data: a multi-hospital comparison to medical records information. Matern Child Health J 2002;16:169–79.
17. Greb AE, Pauli RM, Kirby RS. Accuracy of fetal death reports: comparison with data from an independent stillbirth assessment program. Am J Public Health 1987;77:1202–6.
18. Zhang J, Meikle S, Grainger D, Trumble A. Multifetal pregnancy in older women and perinatal outcomes. Fertil Steril 2002;78:562–8.
19. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986;42:121–30.
20. Clayton D, Hills M. Statistical models in epidemiology. Oxford: Oxford University Press, 1993.
21. National Center for Health Statistics. Vital statistics of the United States, 1998; II-Mortality. Hyattsville, MD: NCHS; 1991.
22. Powers WF, Kiely JL. The risks confronting twins: a national perspective. Am J Obstet Gynecol 1994;170:456–61.
23. Vital statistics of the United States, 1979-1990. Vols I and II. Hyattsville, MD: National Center for Health Statistics; 1984–1994.
24. Kristensen S, Salihu HM, Alexander GR. Premature rupture of the membranes and early mortality among triplets in the United States. Eur J Obstet Gynecol Reprod Med 2004;112:36–42.
© 2004 by The American College of Obstetricians and Gynecologists.
25. Kirby RS. The coding of underlying cause of death from fetal death certificates: issues and policy considerations. Am J Public Health 1993;83:1088–91.