Parental Race/Ethnicity and Adverse Birth Outcomes in New York City: 2000–2010

Abstract

Objectives. To examine the association of maternal race/ethnicity only and parental race/ethnicity jointly with adverse birth outcomes (low birth weight, small for gestational age, preterm birth, and infant mortality) among New York City women.

Methods. We used Bureau of Vital Statistics, New York City Department of Health and Mental Hygiene birth- and death-linked data from 2000 to 2010 (n = 984 807) to quantify the association of maternal race/ethnicity and parental race/ethnicity concordance or discordance with each outcome.

Results. By maternal race/ethnicity, infants of non-Hispanic Black, Hispanic, and Asian women had risks of adverse birth outcomes between 10% and 210% greater than infants of non-Hispanic White women. Infants of non-Hispanic Black, Asian, and Hispanic couples exhibited higher risk of adverse birth outcomes than infants of non-Hispanic White couples. Moreover, parental racial/ethnic discordance was associated with an increased risk of adverse birth outcomes, with highest risks for pairings of Asian men with non-Hispanic White, non-Hispanic Black, and Hispanic women, and of Asian women with non-Hispanic Black and Hispanic men.

Conclusions. Parental race/ethnicity discordance may add stress to women during pregnancy, affecting birth outcomes. Thus, parental race/ethnicity should be considered when examining such outcomes.

Racial/ethnic disparities in birth outcomes in the United States are well documented, with non-Hispanic Black women exhibiting the worst outcomes. For example, recent national data show that babies of non-Hispanic Black women are more likely than babies of non-Hispanic White women to have low birth weight (LBW; < 2500 g) or to be preterm (< 37 weeks)^1,2 and to have higher rates of infant mortality (IM).¹ Compared with non-Hispanic White babies, babies of Hispanic women have the same probabilities of being LBW and dying, whereas babies of Asian women have a higher probability of being LBW and a lower probability of dying.^1,2 Hispanic babies have a higher probability of being born preterm than non-Hispanic White babies.² In New York City (NYC), a place with a more diverse population than the United States as a whole,³ non-Hispanic Black and Hispanic women have higher probabilities of LBW and preterm babies as well as a higher IM rate than non-Hispanic White women.^4,5 For Asians, these estimates are higher for LBW and preterm birth but lower for IM relative to non-Hispanic Whites.^4,5 Thus, it is important to examine these disparities.

To explain these disparities and building upon the weathering hypothesis proposal of cumulative experiences of social inequality and racism,⁶ recent studies suggest that chronic stress associated with everyday interpersonal and institutional racism may have an effect on birth outcomes among non-Hispanic Black women.^7,8 However, this approach only accounts for the stress burden of the mother, as her race/ethnicity has been the sole information used to determine the child’s race/ethnicity,⁹ and ignores possible stress associated with the father’s race/ethnicity. For example, non-Hispanic Blacks,¹⁰ Hispanics,¹¹ and Asians¹² are more likely to report higher rates of perceived racial/ethnic discrimination than non-Hispanic Whites, with men reporting higher rates than women. Moreover, partnering with a man outside a woman’s race/ethnicity may lead to stress because of discriminatory treatment, family disapproval, and reduced social support.¹³ Thus, it is possible that a partner’s race/ethnicity may increase the stress experiences of a woman during pregnancy. In studies that examined the effect of Black–White intermarriage on birth outcomes, infants of non-Hispanic Black mothers had increased odds of LBW, small for gestational age (SGA), preterm birth, and IM, but there was also a substantial increase in adverse birth outcomes associated with having a Black father regardless of ethnicity^14–19 or a non-Hispanic Black father.^20,21 For example, Fulda et al. found that for children of non-Hispanic White and Hispanic women, the odds of very LBW were at least 52% higher if the father was non-Hispanic Black.²⁰ By contrast, Nystrom et al. reported that babies from Asian couples were twice as likely to be LBW than babies from mixed White–Asian couples or White couples.²² This evidence of both parents’ race/ethnicity affecting birth outcomes, together with the increase in racial/ethnic intermarriage and partnering over recent years,^23,24 suggests that adding the father’s race/ethnicity may better capture the stress associated with racial/ethnic discrimination in US society.²⁵

To investigate this issue, we used NYC birth- and death-linked data from 2000 to 2010 to (1) examine the association of maternal race/ethnicity with adverse birth outcomes (LBW, SGA, preterm births, and IM) among NYC women and (2) determine the added effect of paternal race/ethnicity on these outcomes.

METHODS

The linked birth and infant death files were provided by the Bureau of Vital Statistics of the NYC Department of Health and Mental Hygiene (DOHMH). These data are linked following National Center for Health Statistics procedures. Specifically, death certificates are linked to information from the corresponding birth certificates for each infant aged younger than 1 year who dies in NYC. The accuracy of the matching for the birth and infant death records is above 99% for the years included in this study (2000–2010). In addition, the data have been collected fairly consistently for these years. These data include all infants born alive and those who died within the first year of life in NYC, resulting in 1 377 404 records.

Outcomes

The outcomes were LBW, SGA, preterm births, and IM. We defined LBW as less than 2500 grams. Using a US national reference,²⁶ we defined SGA as gestational-age and sex-specific birth weight below the 10th percentile. This measure reflects the size of infants relative to their peers regardless of mother’s race/ethnicity. We defined preterm infants as those born before 37 weeks of gestational age and IM as death during the first year of life.

Exposure

We determined race/ethnicity through the birth certificate items pertaining to race and ethnicity. We combined the race and ethnicity variables to obtain the Office of Management and Budget’s standard categories used by the US Census²⁷: non-Hispanic Black, non-Hispanic White, Asian, American Indian/Alaska Native, Native Hawaiian/Pacific Islander, other race, and Hispanic. For analytical purposes, we used non-Hispanic Black, non-Hispanic White, Hispanic, and Asian as racial/ethnic categories for the infant’s mother and father. (Hereafter, Black and White will be used to denote non-Hispanic Black and non-Hispanic White, respectively.) To account for concordance (parents of same race/ethnicity) and discordance (parents of different race/ethnicity), we created a variable combining mother’s and father’s race/ethnicity. We used Whites as the reference, as they have the lowest incidence for the study outcomes in the US population.

Covariates

We included in the analysis maternal and infant characteristics that were considered as confounders or covariates in previous studies.^14,15,28,29 We determined maternal age using the mother’s date of birth; for analytical purposes, we considered age as continuous. We denoted marital status (married vs unmarried) as provided in the data.

Mother’s education at the time she gave birth was collected as continuous (0–17 years) for the years 2000 to 2007 and as categorical (≤ 8th grade; 9th–12th grade, no diploma; high school graduate or general equivalency diploma; some college credit but not degree; associate’s degree; bachelor’s degree; master’s degree; doctorate or professional degree; unknown) for the years 2008 to 2010. For this analysis, we specified education as less than high school, high school or GED, some college, or college graduate or more. We collected mother’s country of birth or nativity status from the birth certificate as NYC, the rest of New York State, other US states and territories (including Puerto Rico), and foreign-born; for analytical purposes, we specified nativity status as US or foreign-born. Insurance coverage of mothers’ medical care and delivery expenses was classified by the DOHMH into categories of health insurance coverage (Medicaid, HMO, other third party, self-pay) and defined as insured and uninsured.

Clinical gestational age was determined by the medical provider and recorded in weeks. We used infant's sex (male or female) and Apgar score at 5 minutes (continuous) as provided by the DOHMH. We calculated parity using data on previous deliveries and defined it as nulliparous or first born versus multiparous. In addition, to identify diabetes and preeclampsia during pregnancy, we used information on self-reported clinical diagnoses of chronic diabetes or gestational diabetes and preeclampsia or eclampsia, respectively. Finally, we denoted maternal use of tobacco and alcohol during pregnancy (yes or no) as provided by the DOHMH.

The analysis was limited to singleton live births to White, Black, Hispanic, and Asian women aged 15 to 45 years and residing in of 1 of the 5 boroughs of NYC (n = 1 206 202) between 2000 and 2010. We excluded records if infants were born with congenital anomalies (n = 20 993), were missing information on SGA, or had out-of-range data on birth weight (< 500 and > 6000 g; n = 2828); if mothers were missing information on educational attainment (n = 7899), nativity status (n = 3408), or health insurance (n = 3263); or if fathers were missing race/ethnicity (n = 182 984). These exclusions yielded an analytical sample of 984 807 records.

Statistical Analysis

We present descriptive statistics of the study population for selected variables and according to race/ethnicity, using means for continuous variables and frequencies and proportions for categorical variables. To determine statistical significance, we used χ² statistics (categorical variables) and analysis of variance (continuous variables).

We used log-binomial regression to quantify the association between mother’s race/ethnicity and each outcome (LBW, SGA, preterm birth, and IM) before and after, controlling for maternal age, education, nativity status, marital status, health insurance, and tobacco use during pregnancy; gestational age, diabetes, and preeclampsia; infant’s sex; and parity. However, the model for LBW was not adjusted for preeclampsia, SGA was not adjusted for gestational age and infant’s sex, preterm birth was not adjusted for gestational age and preeclampsia, and IM was not adjusted for gestational age and infant’s sex but was adjusted for SGA.

We conducted data management and statistical analysis with SAS version 9.4 for Windows (SAS Institute, Cary, NC).

RESULTS

Table 1 shows descriptive statistics for selected maternal and infant characteristics according to mother’s race/ethnicity. The mean age of NYC women in the sample was 28.8 years. A third of the population was of Hispanic ethnicity (33.2% of women and 32.0% of their partners), more than half were born outside the United States, almost half had more than a high school education, almost two thirds were married, and almost all had health insurance. The average birth weight among live births was 3291.4 grams, with a gestational age of 38.8 weeks and an Apgar score at 5 minutes of 9.0. A little more than half of infants were boys and 41% were first-born babies. The prevalence of gestational diabetes was 4.3% and the prevalence of preeclampsia was 1.6%. Less than 2% of women reported consuming tobacco and alcohol during pregnancy.

TABLE 1—

Characteristics of Mothers and Infants According to Mother’s Race/Ethnicity: New York City, 2000–2010

Characteristic	White, No. (%) or Mean ±SD	Black, No. (%) or Mean ±SD	Hispanic, No. (%) or Mean ±SD	Asian, No. (%) or Mean ±SD	Total, No. (%) or Mean ±SD
Total	303 668 (30.8)	206 017 (20.9)	327 305 (33.2)	147 817 (15.0)	984 807
Maternal or parent characteristics
Maternal age, y	30.6 ±5.7	28.3 ±6.3	27.0 ±6.1	29.9 ±5.2	28.8 ±6.1
Married	277 183 (91.3)	88 581 (43.0)	129 556 (39.6)	129 431 (86.9)	623 751 (63.3)
Education
< high school	20 333 (6.7)	40 187 (19.5)	123 930 (37.9)	33 898 (22.9)	218 348 (22.2)
High school or GED	79 533 (26.2)	67 401 (32.7)	98 206 (30.0)	41 497 (28.1)	286 637 (29.1)
Some college	50 440 (16.6)	60 399 (29.3)	67 908 (20.7)	21 795 (14.7)	200 542 (20.4)
≥ college graduate	153 362 (50.5)	38 030 (18.5)	37 261 (11.4)	50 627 (34.3)	279 280 (28.4)
Foreign-born mother	97 926 (32.2)	99 895 (48.5)	212 886 (65.0)	138 566 (93.7)	549 273 (55.8)
Health insurance	297 404 (98.0)	201 211 (97.7)	321 817 (98.3)	144 496 (97.7)	964 928 (98.0)
Father’s race/ethnicity
White	276 260 (91.0)	4 142 (2.0)	18 035 (5.5)	10 840 (7.3)	309 277 (31.4)
Black	5 690 (1.9)	188 235 (91.4)	21 996 (6.7)	2 392 (1.6)	218 313 (22.2)
Hispanic	16 053 (5.3)	11 302 (5.5)	284 594 (86.9)	2 770 (1.9)	314 719 (32.0)
Asian	4 532 (1.5)	1 264 (0.6)	1 849 (0.6)	131 373 (88.9)	139 018 (14.1)
Infant or pregnancy characteristics
Birth weight, g	3 375.4 ±503.8	3 197.6 ±606.4	3 302.9 ±545.3	3 224.1 ±494.4	3 291.4 ±543.6
Gestational age, wk	39.1 ±1.6	38.5 ±2.3	38.8 ±1.9	38.8 ±1.7	38.8 ±1.9
Male infant	155 919 (51.3)	105 109 (51.0)	167 170 (51.1)	76 853 (52.0)	505 051 (51.3)
First-born infant	133 560 (44.0)	79 320 (38.5)	122 886 (37.5)	67 501 (45.7)	403 267 (40.9)
Apgar score at 5 min	9.0 ±0.4	8.9 ±0.6	9.0 ±0.5	9.0 ±0.4	9.0 ±0.5
Gestational diabetes	7 823 (2.6)	9 404 (4.6)	14 107 (4.3)	11 328 (7.7)	42 662 (4.3)
Preeclampsia	3 089 (1.0)	5 204 (2.5)	6 279 (1.9)	1 350 (0.9)	15 922 (1.6)
Tobacco use during pregnancy	5 953 (2.0)	5 769 (2.8)	5 742 (1.7)	582 (0.4)	18 046 (1.8)
Alcohol use during pregnancy	995 (0.3)	640 (0.3)	795 (0.2)	356 (0.2)	2 786 (0.3)

Note. GED = general equivalency diploma. “White” and “Black” denote “non-Hispanic White” and “non-Hispanic Black,” respectively. All χ² and analysis of variance results were significant at P < .01 level except for Apgar score.

Source. Bureau of Vital Statistics of the New York City Department of Health and Mental Hygiene.

When we examined these characteristics according to mother’s race/ethnicity, Hispanic women were the youngest (27.0 years; Table 1), Asian women had the highest proportion who were foreign-born (93.7%), and Black and Hispanic women were more likely to have less than a high school education and to be married (all P values < .01). Black women had babies with the lowest birth weight (3197.6 g) and gestational age (38.5 weeks) but had the highest prevalence of preeclampsia (2.5%) and tobacco use during pregnancy (2.8%; all P values < .01). Hispanic women had the lowest proportion of first-born babies (37.5%) and Asian women had the highest prevalence of gestational diabetes (7.7%; P values < .01).

Among NYC women, the proportion of adverse birth outcomes was 6.0% for LBW, 11.9% for SGA, 7.1% for preterm births, and 2.7 of 1000 for IM (Table A, available as a supplement to the online version of this article at http://www.ajph.org). When we considered only mother’s race/ethnicity, infants of Black, Hispanic, and Asian women had at least 50% higher risks of being LBW and at least 20% higher risks of being SGA, preterm, and dying before a year old than infants of White women (Table 2). These risks remained significant after adjustment, although of smaller magnitude. For example, compared with White infants, Black (95% confidence interval [CI] = 1.3, 1.4) and Asian (95% CI = 1.2, 1.3) infants had a 30% greater risk of LBW after we controlled for maternal age, education, nativity status, marital status, health insurance, and tobacco use during pregnancy; gestational age and diabetes; parity; and infant’s sex. We observed a similar pattern, but with greater risks, for SGA. Compared with infants of White women, infants of Black women had the highest risks of preterm birth and IM, followed by those of Hispanic and Asian women.

TABLE 2—

Unadjusted and Adjusted Relative Risk for Birth Outcomes According to Mother’s and Parents’ Race/Ethnicity Concordance or Discordance: New York City, 2000–2010

	Low Birth Weight		Small for Gestational Age		Preterm Birth		Infant Mortality
Race/Ethnicity	RR	ARR (95% CI)	RR	ARR (95% CI)	RR	ARR (95% CI)	RR	ARR (95% CI)
Mother’s race/ethnicity only
White (Ref)	1	1	1	1	1	1	1	1
Black	2.4	1.3 (1.3, 1.4)	1.6	1.5 (1.4, 1.5)	2.0	2.0 (1.9, 2.1)	3.4	3.1 (2.7, 3.5)
Hispanic	1.5	1.1 (1.0, 1.1)	1.2	1.1 (1.0, 1.1)	1.5	1.6 (1.5, 1.6)	1.9	1.8 (1.6, 2.1)
Asian	1.5	1.3 (1.2, 1.3)	1.5	1.6 (1.5, 1.6)	1.2	1.4 (1.3, 1.4)	1.2	1.3 (1.1, 1.6)
Mother’s–father’s race/ethnicity
White–White (Ref)	1	1	1	1	1	1	1	1
White–Black	1.5	1.1 (1.0, 1.1)	1.1	1.0 (0.9, 1.1)	1.4	1.3 (1.2, 1.5)	2.7	2.5 (1.7, 4.0)
White–Hispanic	1.5	1.1 (1.0, 1.1)	1.0	0.9 (0.9, 1.0)	1.4	1.3 (1.3, 1.4)	1.2a	1.1 (0.7, 1.6)
White–Asian	1.4	1.2 (1.1, 1.4)	1.4	1.4 (1.3, 1.5)	1.0	1.0 (0.9, 1.1)	1.1a	1.1 (0.5, 2.4)
Black–Black	2.5	1.4 (1.3, 1.4)	1.6	1.5 (1.5, 1.6)	2.2	2.1 (2.0, 2.2)	3.6	3.3 (2.9, 3.8)
Black–White	1.9	1.2 (1.1, 1.3)	1.3	1.2 (1.1, 1.4)	1.7	1.6 (1.5, 1.8)	2.9	2.9 (1.8, 4.8)
Black–Hispanic	2.6	1.3 (1.3, 1.4)	1.7	1.6 (1.4, 1.6)	2.1	2.0 (1.9, 2.1)	2.9	2.2 (1.8, 3.3)
Black–Asian	2.7	1.5 (1.3, 1.8)	2.0	1.9 (1.7, 2.2)	2.1	2.1 (1.8, 2.5)	3.9	4.0 (1.9, 8.4)
Hispanic–Hispanic	1.5	1.1 (1.1, 1.1)	1.2	1.1 (1.1, 1.1)	1.6	1.6 (1.6, 1.7)	1.9	1.8 (1.6, 2.1)
Hispanic–White	1.4	1.0 (0.9, 1.1)	1.0	1.0 (0.9, 1.0)	1.5	1.4 (1.3, 1.5)	1.4	1.4 (1.0, 2.0)
Hispanic–Black	2.2	1.1 (1.1, 1.2)	1.4	1.2 (1.2, 1.3)	2.0	1.9 (1.8, 2.0)	3.4	2.9 (2.3, 3.6)
Hispanic–Asian	1.7	1.3 (1.1, 1.5)	1.5	1.5 (1.3, 1.7)	1.3	1.3 (1.1, 1.6)	2.3	2.3 (1.0, 5.1)
Asian–Asian	1.6	1.3 (1.3, 1.4)	1.6	1.6 (1.6, 1.7)	1.3	1.4 (1.3, 1.4)	1.2	1.3 (1.1, 1.6)
Asian–White	1.3	1.1 (1.0, 1.1)	1.1	1.1 (1.1, 1.2)	1.3	1.3 (1.2, 1.4)	0.8a	1.1 (0.6, 1.8)
Asian–Black	1.9	1.2 (1.1, 1.4)	1.6	1.6 (1.4, 1.7)	1.8	1.9 (1.7, 2.2)	2.7	3.0 (1.5, 5.7)
Asian–Hispanic	2.2	1.2 (1.1, 1.3)	1.3	1.2 (1.1, 1.4)	2.1	2.1 (1.9, 2.4)	2.8	3.2 (1.7, 5.8)
Other combinations	2.0	1.3 (1.2, 1.4)	1.4	1.3 (1.2, 1.4)	1.6	1.6 (1.4, 1.7)	2.2	2.2 (1.2, 4.1)

Note. ARR = adjusted relative risk; CI = confidence interval; RR = relative risk;. “White” and “Black” denote “non-Hispanic White” and “non-Hispanic Black,” respectively. Adjustments were as follows: low birth weight: maternal age, maternal education, maternal nativity status, marital status, health insurance, tobacco use during pregnancy, gestational diabetes, gestational age, infant’s sex, and parity; small for gestational age: maternal age, maternal education, maternal nativity status, marital status, health insurance, tobacco use during pregnancy, gestational diabetes, preeclampsia, and parity; preterm birth: maternal age, maternal education, maternal nativity status, marital status, health insurance, tobacco use during pregnancy, gestational diabetes, infant’s sex, and parity; infant mortality: maternal age, maternal education, maternal nativity status, marital status, health insurance, tobacco use during pregnancy, gestational diabetes, preeclampsia, parity, and small for gestational age.

Source. Bureau of Vital Statistics of the New York City Department of Health and Mental Hygiene.

^aNonsignificant at .05.

Table 2 also presents the risk of adverse birth outcomes that we observed using the mother’s and father’s race/ethnicity before or after adjustment relative to infants of White parents. Compared with infants of White parents, infants with White mothers had higher risks of LBW regardless of the father’s race/ethnicity, for SGA when the father was Asian, for preterm births when the father was Black or Hispanic, and for IM when the father was Black. Black couples had higher risks of adverse birth outcomes than White couples. Moreover, whereas we observed higher risks for all outcomes regardless of a Black woman’s partner’s race/ethnicity, risks were highest for LBW, SGA, and IM when Black women had an Asian partner and for preterm birth when they had either a Hispanic or an Asian partner.

For Hispanic couples, the risk of adverse birth outcomes was higher than for White couples. With the exception of LBW and SGA for Hispanic women with White partners, risks were higher for Hispanic women regardless of the race/ethnicity of their partner, with the highest risks observed for LBW and SGA when the father was Asian and for preterm birth and IM when the father was Black. Compared with White couples, Asian couples had higher risk of adverse birth outcomes. With the exception of no risk of IM when the father was White, Asian women’s risks were higher for all outcomes regardless of the father’s race/ethnicity, with the highest risks observed when the father was either Black or Hispanic.

DISCUSSION

Consistent with previous studies,^1,2,30 we found that when maternal race/ethnicity was solely considered, infants of Black women had higher risks of LBW, preterm births, and IM than infants of White women. Similarly, and in contrast to previous studies,^1,2,30 infants of Hispanic women had higher risks of all outcomes examined relative to infants of White women. The same pattern was observed for Asian women. When we examined parental racial/ethnic concordance, infants of Black, Hispanic, and Asian couples had higher risk of adverse birth outcomes than infants of White couples. Moreover, compared with White couples, parental racial/ethnic discordance was associated with an increased risk of adverse birth outcomes for all racial/ethnic groups. Relative to White couples, White women had highest risk of LBW and SGA when the partner was Asian, of preterm birth with a Black or Hispanic partner, and of IM with a Black partner. For Black women, the risks of an adverse birth outcome were highest with an Asian partner. Hispanic women had greater risks of having a LBW or SGA baby with an Asian partner and higher risks of a preterm birth or IM with a Black partner. Finally, for Asian women, the risks were higher for LBW, preterm birth, and IM with a Black or Hispanic partner and for SGA with a Black partner.

Our findings of higher risks of adverse birth outcomes among Hispanic women relative to White women could be explained by the origins of the Hispanic population in NYC: Puerto Ricans and Dominicans represent the largest Hispanic subgroups there, whereas Mexicans are the largest for the US population.³¹ Puerto Ricans and Dominicans have a stronger African ancestry than most Hispanic subgroups, exposing them to higher levels of racial/ethnic discrimination in the United States.³² Interestingly, although limited data are available for Dominicans, Puerto Rican^1,2,29 and Dominican²⁹ infants tend to have birth outcomes similar to those of non-Hispanic Black infants.

Previous studies examining LBW among Black and White couples in the United States^16,17 and California^15,21 have found that the greatest odds of LBW were among infants whose parents were both Black. However, for Black–White couples, the results were mixed, with some studies showing increased odds of LBW for infants regardless of whether the father or the mother was Black,^15,17 whereas 1 study found no risk of LBW associated with such partnering arrangements.²¹ Our results also show that infants of Black couples have higher risk of LBW than infants of White couples. In addition, this high risk of LBW was observed for infants of Black–White couples regardless of which was Black—the mother or the father. Moreover, Nystrom et al.²² found that infants of Asian parents had higher odds of being LBW than those of White couples or White–Asian couples. We also found an increased risk of LBW among infants of Asian parents compared with those of White parents. Finally, and consistent with a previous study,²⁰ we found that Hispanic women had higher risk of LBW when they had a Black partner.

Getahun et al.,¹⁴ using 1995 to 2001 US birth and infant death files and the same definition of SGA (> 10th percentile) that we used, found that compared with infants of White parents, infants of Black parents had 1.92 times the risk of being SGA, followed by 1.49 times the risk for infants of Black mothers and White fathers. We found that infants of Black parents had a 50% greater risk of SGA, whereas those with Black mothers and White fathers had a 20% greater risk of SGA than those with White parents. For IM, consistent with Getahun et al.,¹⁴ we found that the highest risk was associated with having a Black mother regardless of the father’s race/ethnicity.

For preterm births, previous studies using US^14,17,19 or California¹⁵ data reported a higher risk for infants with Black parents than for those with White parents, followed by infants with Black mothers and White fathers. These risks ranged from 1.07¹⁵ to 2.12¹⁷ when both parents were Black and from 1.25¹⁹ to 1.57¹⁷ when the mothers were Black and the fathers were White. Consistent with these studies,^14,15,17,19 our findings suggest that infants of Black parents had 2.1 times the risk of having a preterm baby compared with White couples. This risk was 1.6 for Black women with White partners.

Our study included 11 years of NYC infant birth and death data. This provided us with a large and diverse sample with a significant number of outcomes to examine the associations of interest while controlling for important confounders. However, there are some limitations. One potential limitation is the use of a single locale, NYC. However, NYC data allowed us to have a large and diverse sample of women in order to examine the associations of interests. In fact, previous studies^{14–16,18,19,21} have been limited by focusing only on Black and White women regardless of ethnicity, whereas this study was able to include non-Hispanic Black, Hispanic, Asian, and non-Hispanic White women. Another possible limitation is the self-reported nature of race/ethnicity. However, the positive predictive value of maternal race/ethnicity for the groups included in this analysis—White, Black, Hispanic, and Asian—has been shown to be 92.5% to 99.5% with California birth and US death certificates data. Similarly, in the National Health and Nutrition Examination Survey III, self-reported data on smoking (for both sexes) closely matched (92.5%) serum cotinine levels of higher than 15.0 nanograms per milliliter, a cutpoint used to identify smokers. Thus, it is unlikely that use of self-reported data affected our results.

The number of records excluded because of father’s missing race/ethnicity (n = 182 984) could also have been a limitation. However, we conducted several sensitivity analyses to determine the bias introduced by these exclusions. First, consistent with a previous study,²⁰ we found that records excluded because the father’s race/ethnicity was missing were more likely to be for Black (44.8%) and Hispanic (41.0%) women than for White (7.2%) and Asian (7.1%) women. Second, as previously reported,³³ infants whose father’s race/ethnicity was missing had higher risks of being LBW, SGA, preterm, and dying before a year old than infants whose father’s race/ethnicity was not missing. Third, consistent with a previous study,³⁴ we assigned mother’s race/ethnicity to records with missing father’s race/ethnicity and repeated the analyses for the fully adjusted models for all outcomes. The relative risk estimates obtained were nearly identical to those presented in Table 2. Finally, we performed multiple imputations and generated 3 data sets with paternal race/ethnicity. The results were nearly identical to those in Table 2. Thus, it is very unlikely that these exclusions affected our results. Lack of adjustment for prenatal care may have affected our results. Prenatal care information was not available for the years 2008 to 2010 because of data quality. However, from 2000 through 2007, between 92% and 95% of pregnant women in NYC received prenatal care within the first and second trimester. Moreover, adequate prenatal care is linked to health insurance regardless of whether women have private insurance or Medicaid, and 98% of women in our population had health insurance.

We found that, although there was increased risk of having an adverse birth outcome associated with having parents of different or discordant race/ethnicity, this increase was not consistent across outcomes and women’s race/ethnicity. For example, we found that for White women, the risks of LBW and SGA were higher when they had an Asian partner whereas the risk of preterm birth was higher with a Black or Hispanic partner and the risk of IM was higher with a Black partner. Having an Asian partner carried increased risks of all adverse birth outcomes for Black women and for LBW and SGA for Hispanic women. By contrast, for Asian women, the risks were higher for LBW, preterm birth, and IM when they had a Black or Hispanic partner.

Our findings are consistent with the weathering⁶ and chronic stress^7,8 hypotheses. First, our findings suggest that using the mother’s race/ethnicity only⁹ may not accurately quantify her stress and that the father’s race/ethnicity may add or moderate stress on pregnant women,²⁵ affecting birth outcomes. Both mother’s and father’s race/ethnicity should therefore be considered when examining such outcomes. Second, the risks seem to be higher for the least common pairings—Asian men with White, Black, and Hispanic women and Asian women with Black and Hispanic men.^13,35 Therefore, these interracial partnering and marriages may add stress to women during pregnancy because they then experience a “double minority status” within their immediate social environment. Lastly, our findings call attention to the stress associated with racial/ethnic discrimination and the lack of tolerance or acceptance for interracial relationships in an increasingly diverse US society, where racism appears to be resurfacing as a growing concern.