Copyright 2002 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2002
East Asians have inherently higher bilirubin levels at birth than whites. The potential for unnecessary treatment makes jaundice a problem of public health and clinical significance.
To report the occurrence of jaundice diagnoses in East Asian and mixed East Asian/white infants in Washington State in recent years, and to compare the risk of diagnosis with neonatal jaundice among these infants, relative to white infants.
Population-based cohort study in Washington state. Participants were infants of full East Asian parentage (n = 3000), maternal Asian parentage (n = 2997), paternal Asian parentage (n = 2048), and white parentage (n = 3000). Diagnoses of jaundice and "severe jaundice" were identified using International Classification of Diseases, Ninth Revision (ICD-9) diagnosis and procedure codes from hospital discharge records.
Infants of full East Asian parentage were more likely to be diagnosed with jaundice than were white infants (relative risk [RR], 1.37; 95% confidence interval [CI], 1.16-1.62). For infants with Asian mothers and white fathers, the RR was 1.09 (95% CI, 0.91-1.30). Infants with Asian fathers and white mothers had an RR of 1.26 (95% CI, 1.05-1.52). The risk of severe jaundice requiring phototherapy, blood transfusion, or rehospitalization, however, was significantly elevated only for infants of full East Asian parentage (RR, 1.7; 95% CI, 1.12-2.58).
Diagnoses of neonatal jaundice occurred more often among East Asian and mixed Asian/white infants than among white infants. However, the risk of jaundice requiring extended hospital stay, rehospitalization, phototherapy, or blood transfusion was elevated only for infants of full East Asian parentage.
APPROXIMATELY 60% to 70% of the 4 million infants born annually in the United States become clinically jaundiced.1 East Asians have higher bilirubin levels at birth than whites.2- 10 Previous reports from the United States indicate that 31% of East Asian infants meet the standard criteria for nonphysiologic hyperbilirubinemia11 and have an approximately 3-fold increased risk of jaundice.12,13
We compared the rates of diagnosis with jaundice in infants born to East Asian, mixed East Asian/white, and white parents in Washington State. We also assessed differences in the proportion of severe jaundice among these cohorts, using procedures performed and duration of hospital stay as indicators of severity.
We conducted a population-based cohort study of infants born in Washington state from 1987-1995. Data were obtained from the Washington State Birth Events Records Database. This database, created by the Washington State Department of Health Office of Hospital and Patient Data, links birth certificates to hospital discharge information for the birth hospitalizations of mother and child. Four cohorts of infants defined by parental race/ethnicity as indicated on the birth certificate (based on prenatal record or self-reported) were identified: those born to 2 white parents, those born to 2 East Asian (hereafter called "Asian") parents, those born to an Asian mother and white father, and those born to a white mother and Asian father. Asian infants included those of Chinese, Japanese, or Filipino descent. We selected these infants because much of the existing literature on jaundice focuses on these ethnicities and because they were the largest Asian cohorts. We excluded other ethnicities such as Korean and Vietnamese to create more homogeneity and to increase our power to examine associations within groups. We also excluded groups such as "Samoan" and "Pacific Islander" because of their small numbers and the "other Asian" category because it was not well defined. Infants with parents identified as Native American or other nonwhite classifications were not included.
A random sample of 3000 infants born from 1987-1995 to white parents was the reference group. The 3 comparison groups included a random sample of 3000 infants born to 2 Asian parents, a random sample of 3000 infants born to an Asian mother and white father, and all 2048 infants born to a white mother and Asian father.
Infants with neonatal jaundice were identified by screening all available International Classification of Diseases, Ninth Revision (ICD-9)14 diagnosis fields in the child's hospital discharge record for codes indicating jaundice (774.1, 774.2, 774.39, 774.4, and 774.6). Information concerning rehospitalization 28 or fewer days after birth was obtained by linking subjects' records with Comprehensive Hospital Abstract Reporting System (CHARS) records for 1987-1996. Created by the Washington State Department of Health, CHARS contains discharge data for all hospitalizations in nonmilitary hospitals.
Stratified analyses were conducted to calculate Mantel-Haenszel relative risk estimates and to evaluate the presence of confounding and/or effect modification. Variables considered for their potential effects included maternal age (<20, 20-24, 25-29, 30-34, or ≥35 years), sex, gravidity, parity (0, 1, 2, or ≥3 prior births), duration of gestation (20-36, 37-42, or >42 weeks), maternal established or gestational diabetes, prenatal smoking or alcohol use (yes/no), birth weight (<2500, 2500-4500, or >4500 g), and preeclampsia (ICD-9 code 642.4 or 642.5). Factors that altered risk estimates more than 10% were considered confounders. Other factors possibly related to jaundice, such as maternal hepatitis (ICD-9 code 070), congenital anemia (ICD-9 code 776.5), and newborn sepsis (per the birth certificate), were also considered.
Initially, we evaluated jaundice from any cause as a single outcome. However, we were concerned that infants with jaundice might differ across cohorts with respect to short gestational duration (20-36 weeks), preterm delivery, hepatitis, hemolysis/bruising, maternal hepatitis, or congenital anemia. To isolate relationships of interest, we excluded infants with these potential causes to identify infants with presumed physiologic jaundice.
Infants with hospital stays of more than 5 days, procedure codes indicating phototherapy or blood transfusion during birth hospitalization, or rehospitalization for jaundice 28 or fewer days after birth were classified as having "severe" jaundice.
Asian mothers were older and less likely to smoke than white mothers (Table 1). The most common ICD-9 code was "neonatal jaundice due to unspecified causes" (85%). "Neonatal jaundice due to pre-term delivery" was the second most common jaundice diagnosis (13.5%). A jaundice diagnosis occurred less frequently among white infants (7.4 per 100 infants) and most often among infants with 2 Asian parents (10.1 per 100 infants) (Table 2). For infants of mixed parentage, the diagnosis of jaundice differed slightly by whether Asian heritage was maternal (8 per 100 infants) or paternal (9.3 per 100 infants).
Infants of full Asian parentage were 37% more likely to be diagnosed with jaundice than white infants (relative risk [RR], 1.37; 95% confidence interval [CI], 1.16-1.62). Infants with Asian fathers and white mothers had a 26% greater risk (RR, 1.26; 95% CI, 1.05-1.52), whereas no increased risk was observed among infants with Asian mothers and white fathers (RR, 1.09; 95% CI, 0.91-1.30). These findings did not change when analyses were restricted to infants classified as having physiologic jaundice.
Adjustment for maternal age, infant sex, parity, duration of gestation, diabetes, smoking and alcohol consumption during pregnancy, birth weight, and preeclampsia did not appreciably change the estimates, nor were marked differences in risk observed with respect to these variables.
Infants with 2 Asian parents were more likely to be diagnosed with jaundice regardless of their parents' country of ethnic origin. Relative to white infants, the RR of diagnosis with neonatal jaundice for infants identified as being born to Chinese parents was 1.25 (95% CI, 1.00-1.57). For infants of Japanese parents, the RR was 1.85 (95% CI, 1.34-2.55), and for infants of Filipino parents, the RR was 1.34 (95% CI, 1.10-1.63). The RR among infants of mixed-heritage Asian parents was 1.26 (95% CI, 0.81-1.97). When infants with other known causes of jaundice were excluded, the risks of diagnosis with physiologic jaundice increased even more for infants in all Asian subgroups, except those born to Filipino parents.
The risk of severe jaundice significantly increased among infants with 2 Asian parents (RR, 1.70; 95% CI, 1.12-2.58) (Table 3). Infants with Asian mothers/white fathers and white mothers/Asian fathers had RRs of 1.36 (95% CI, 0.87-2.11) and 1.15 (95% CI, 0.69-1.91), respectively. Among specific Asian groups, significantly increased risk of severe jaundice was observed for Japanese (RR, 2.64; 95% CI, 1.27-5.51) and Filipino (RR, 1.68; 95% CI, 1.02-2.76) infants.
Infants of East Asian parentage were more likely to be diagnosed with jaundice than white infants. This is consistent with the results of other studies.3,6,10- 13 In our study, it is possible that clinicians had a lower threshold for testing for, and thus diagnosing, jaundice in Asian infants because of awareness of higher jaundice rates in Asians, or possibly because of skin coloration. To the extent that a diagnosis of jaundice in our data accurately indicates jaundice, we found that, among infants with 2 Asian parents, Japanese infants had the greatest risk, whereas risks for Filipino and Chinese infants were elevated to a lesser extent. Ho5 found that not all Asian groups had similar risks of jaundice, and recent investigations provide evidence of elevated mutation levels in the bilirubin uridine diphosphate–glucuronosyltransferase gene associated with jaundice in Japanese infants.15,16
The rate of jaundice diagnosis among infants with Asian mothers and white fathers was not substantially different from that of white infants. However, infants with Asian fathers and white mothers had a 32% greater risk relative to white infants, suggesting a stronger paternal influence in determining an infant's risk of jaundice. At this time, a possible genetic basis for paternal influence is unknown.
Asian infants were more likely to have severe jaundice requiring phototherapy and/or blood transfusion, rehospitalization for jaundice, or birth hospitalization greater than 5 days. The subgroup analysis by Asian ancestry suggests that infants of full Filipino and Japanese ancestry may be contributing to this increased risk.
One strength of this analysis is that it was population-based, so infants are representative of all those born in Washington State from 1987-1995. The growing Asian population in Washington State permitted us to identify samples of sufficient size to examine paternal vs maternal Asian influence on risk.
Limitations of this study include reliance on birth certificate and hospital discharge record coding of race/ethnicity and jaundice, neither of which we could validate. We also lacked data on other factors, such as family history of neonatal jaundice,13 genetic traits that might have varied by race,15- 18 medicinal herbs in the diet, breastfeeding,19 or the use of oxytocin to induce or augment labor.15,16,19 Sepsis, preeclampsia, and preterm delivery are also reportedly associated with jaundice.20,21 However, these variables were not effect modifiers, nor did their frequency differ among the 4 cohorts.
Greater knowledge of characteristics associated with risk of jaundice is helpful, particularly as earlier hospital discharge after birth limits the opportunity for clinicians to detect progression to more serious disease. Racial differences have been observed in the time when peak serum bilirubin concentrations occur,18 with about 6% of Asian infants diagnosed with jaundice more than 3 days after birth,11 so early discharge may be of particular relevance for these children. As more is learned about genetic influences for jaundice, these findings may also be helpful in understanding genetic causes.
Accepted for publication November 27, 2001.
This study was presented as a poster at the American Public Health Association 127th Annual Meeting, Chicago, Ill, November 7-11, 1999.
Ethnic variation in the rates of neonatal jaundice has been recognized, and gene mutations associated with hyperbilirubinemia among some Asian groups have been identified. Greater knowledge of characteristics, including race/ethnicity, that may be associated with an increased risk of jaundice may be helpful, particularly as earlier hospital discharge after birth limits opportunity for clinicians to detect progression to more serious disease.
To our knowledge, this is the first report of levels of jaundice diagnosis from population-based data in the United States for infants of mixed Asian-white descent. These population-based findings of increased risks for infants with Asian parents or one Asian parent and one white parent, may provide useful information to clinicians and enhance our understanding of potential genetic causes of jaundice.
Corresponding author and reprints: Beth A. Mueller, DrPH, 1100 Fairview Ave N, MP-381, PO Box 19024, Seattle, WA 98109-1024.
Setia S, Villaveces A, Dhillon P, Mueller BA. Neonatal Jaundice in Asian, White, and Mixed-Race Infants. Arch Pediatr Adolesc Med. 2002;156(3):276-279. doi:10.1001/archpedi.156.3.276