Sickle cell disease (SCD) is an inherited blood disorder causing lifelong anemia and multiorgan complications that contribute to premature death. SCD is among the most prevalent hereditary disorders in humans. It is estimated that more than 3 million Americans are genetic carriers of SCD and that 80,000–100,000 Americans have SCD.^1–3 Four SCD variants—sickle cell anemia (HbSS), sickle-hemoglobin C disease (HbSC), and two types of sickle β-thalassemia (Sβ⁺ thalassemia and Sβ⁰ thalassemia)—account for most cases of SCD.

SCD is most common among people whose ancestors came from sub-Saharan Africa; SCD is also found in India and parts of the Middle East and to a lesser extent in other Mediterranean countries. Consequently, African-Americans are at increased risk for SCD. The average frequency of SCD among African-American births is ~1 in 360 live births and is substantially lower among US Hispanics, ~1 in 16,300 live births based on national newborn screening (NBS) records pooled from states reporting such data.³ Studies estimating incidence or birth prevalence of SCD stratified by demographic characteristics such as race/ethnicity have been conducted using NBS data.^3–6 However, no published information on the frequency of SCD among US births stratified by both race and ethnicity distinguishes children of US-born and foreign-born parents. The primary objectives of this study were to estimate the incidence of SCD among New York State (NYS) newborns classified by maternal race/ethnicity and nativity and to document the contribution of immigration to SCD among newborns in NYS.

From 2000 to 2008, a total of 1,916 newborns were diagnosed with SCD among ~2.2 million NYS live births or 1 in 1,146 live births during this time period. Of these identified newborns, 1,911 (99.7%) were matched to birth certificates. Only matched SCD cases were included in the study cohort. The SCD incidence rates stratified by maternal race/ethnicity, nativity, residence, and birth period, and other selected birth and maternal risk factors are presented in Table 1.

Only minimal changes in overall SCD incidence were observed over this time period. A 1% decrease in SCD incidence among all NYS newborns was observed, from 1 in 1,115 in 2000–2002 to 1 in 1,128 in 2006–2008. A 2% increase in incidence was observed among newborns of US-born mothers, from 1 in 1,693 to 1 in 1,664, offset by an 8% decrease in incidence among newborns of foreign-born mothers, from 1 in 655 during 2000–2002 to 1 in 714 during 2006–2008.

An important predictor of SCD incidence is maternal race/ethnicity (Table 1). Newborns of non-Hispanic black mothers accounted for 86.0% of all SCD cases, with an average rate of 1 in 230 births. Hispanic mothers accounted for 11.5% of all newborns with SCD, for an average incidence of 1 in 2,583. About 2.5% of newborns with SCD had non-Hispanic white mothers or mothers of other maternal race/ethnicity groups. NYC accounted for 69.3% of all newborns with SCD, with an incidence of 1 in 779 live births, as compared with an average incidence of 1 in 1,975 live births in the rest of NYS.

Within every race/ethnicity group, the incidence of SCD was higher among newborns with foreign-born vs. US-born mothers (Table 1). These differences were statistically significant for Hispanics and non-Hispanic blacks and whites (P < 0.05). The overall annual incidence of SCD was 1 in 1,678 newborns of US-born mothers and 1 in 716 newborns of foreign-born mothers. More than half (55.2%) of the 1,911 newborns with SCD in NYS had foreign-born mothers, in comparison with 34.5% of all births in NYS. Nativity varied by race/ethnicity, with foreign-born mothers accounting for 12.2% of non-Hispanic white, 37.6% of non-Hispanic black, 58.0% of Hispanic, and 91.9% of Asian/Pacific Islander births in NYS.

The predominance of foreign-born mothers among NYS newborns with SCD was specific to NYC births; 60.5% of 1,325 newborns with SCD in NYC had foreign-born mothers. In comparison, 43.2% of 586 newborns born in NYS excluding NYC had foreign-born mothers. Of 1,055 newborns with SCD of foreign-born mothers, 802 (76%) were born in NYC and 253 (24%) were born in the rest of NYS.

Differences in SCD incidence by maternal nativity were found within each maternal race/ethnic group. The largest proportional difference in rates was found among non-Hispanic black mothers; the estimated annual incidence rate was 1 in 315 newborns of US-born non-Hispanic black mothers and 1 in 160 newborns of foreign-born non-Hispanic black mothers, a twofold higher rate among infants born to foreign-born non-Hispanic black mothers (P < 0.001). Among infants born to Hispanic mothers, the rate was 1 in 2,320 for foreign-born mothers vs. 1 in 3,173 for US-born mothers (P = 0.03) (Table 1). Finally, the six Asian/Pacific infants with SCD were born solely to mothers born outside the United States, resulting in an incidence of 1 in 27,673 live births.

For other birth and maternal variables examined, higher SCD incidence was observed among female children, children with low birth weight (<2,500 g), pre-term births (<37 weeks), and children of teenage mothers (Table 1), as compared with male children, children with higher weight (≥2500 g), full-term births (≥37 weeks), and children of older mothers (≥20 years), respectively. The incidence of SCD was higher among newborns with foreign-born vs. US-born mothers within each variable group examined.

Table 2 shows the numbers, proportions, and incidence rates of SCD newborns (overall and by subtype) born to foreign-born non-Hispanic black mothers by maternal country of birth among the 20 countries with the largest numbers of SCD-affected births recorded. The largest absolute numbers of infants with SCD born to foreign-born non-Hispanic black mothers were seen among women born in Jamaica, followed by Haiti, Ghana, and Nigeria. However, the highest rates of SCD relative to numbers of births were seen among mothers from West Africa, including specifically Togo (1 in 43), Sierra Leone (1 in 50), Ghana (1 in 65), Nigeria (1 in 69), Guinea (1 in 84), and Mali (1 in 88) (Table 2). Lower SCD rates were observed among births in non-Hispanic black foreign-born mothers from the Caribbean region, including specifically Barbados (1 in 157), Antigua and Barbuda (1 in 162), Jamaica (1 in 180), and Haiti (1 in 187) (Table 2). The differences among Caribbean countries could have been due to chance (P = 0.11), but the differences among African countries were statistically significant (P < 0.00001).

Among foreign-born mothers from West Africa, marked differences were seen in the relative frequencies of HbSS and HbSC among countries. A large majority of children with SCD born to mothers from Sierra Leone (79%), Nigeria (78%), Guinea (76%), and Mali (72%) had the HbSS phenotype; only 41 and 27% of children with SCD born to Ghanaian and Togolese mothers, respectively had HbSS (Table 2).

Table 3 summarizes the estimated annual incidence rates by country of birth and SCD subtype for foreign-born Hispanic mothers. Although the overall incidence rate in that group was 1 in 2,320, it varied for infants born to Hispanic mothers from different countries: Puerto Rico (1 in 3,681), Central American countries (Belize, Guatemala, Honduras, El Salvador, Nicaragua, Costa Rica, and Panama) (1 in 1,526), and the Dominican Republic (1 in 906). The largest number (n = 87,801) of foreign-born Hispanic mothers originated from Mexico. No births in this group were affected by SCD. For all other foreign-born Hispanic mothers, the overall incidence was 1 in 12,728. Infants with SCD whose mothers were born in the Dominican Republic had a higher proportion with HbSC than other Hispanic children with foreign-born mothers (Table 3).

In this report, we present data on the incidence of SCD stratified by both maternal race/ethnicity and maternal place of birth. Previous state-based estimates of SCD incidence among newborn infants of African-American and Hispanic mothers were 1 in 396 and 1 in 36,000, respectively, in California⁵ and 1 in 334 and 1 in 34,500, respectively, in Texas.⁴ Hassell³ published a tabulation of information taken from the National Newborn Screening Information System for the years 2005–2007, according to which the average incidence of SCD was 1 in 365 births among children classified as black or African-American and 1 in 16,305 for Hispanic children. Our data show that the incidence of SCD is much higher among infants born to non-Hispanic black or Hispanic mothers in New York, 1 in 230 and 1 in 2,583, respectively. Among NYS infants of US-born non-Hispanic black mothers, the rate of live births affected by SCD, 1 in 315, was at the low end of the range of previous reports cited above. The relatively high rate of SCD among NYS infants of US-born non-Hispanic black mothers could reflect an influence of the previous generation of immigrants, but we were not able to determine nativity of grandparents. The lower rates of SCD among non-Hispanic blacks in the United States relative to countries in West or Central Africa or the Caribbean could reflect a number of factors, including degree of European admixture and lesser degree of selective pressure by malaria in the United States over the course of multiple generations.

The incidence of SCD among Hispanics living in the northeastern United States is substantially higher than that among Hispanics living elsewhere in the United States, reflecting geographic differences in origins. In much of the country, Hispanics have primarily Mexican ancestry, with relatively little African ancestry or admixture. In contrast, a large proportion of Hispanics in the Northeast have Caribbean origins, especially Puerto Rico and the Dominican Republic, and many either report black race or have African admixture. An often-cited 1993 estimate of the incidence of SCD among Hispanics in the Eastern United States, including New York, was 1:1,114.¹² It should be noted that that estimate was not based on NBS data, unlike our finding of a rate of 1 in 2,583 for infants of Hispanic mothers in New York. The latter is half the rate of 1 in 1,243 recently reported for Hispanics in northern Manhattan, where most are of Dominican descent.¹³ However, the latter estimate is close to the estimate reported in this study for Hispanic mothers from the Dominican Republic. As shown in Table 3, the incidence of SCD among Hispanics in New York varies greatly by country of origin.

The overall incidence of SCD among infants of foreign-born non-Hispanic black women in New York was 1 in 160 live births. This is comparable with the rates reported from screening programs in other countries with large immigrant populations from the Caribbean and West Africa.^14,15 The incidence of SCD among NYS births to foreign-born non-Hispanic black mothers varies by country of origin: 1 in 65–70 infants born to mothers from Ghana or Nigeria and 1 in 180–190 infants born to mothers from Jamaica or Haiti. These estimates are consistent with data from maternal countries of origin. In Ghana, a screening study reported that 1 in 52 infants had probable SCD.¹⁶ Studies from Jamaica have reported the birth rate of SCD to be roughly 1 in 170 live births.^17,18

The striking finding that 55% of all infants with SCD born in NYS during 2000–2008 were born to foreign-born mothers has important implications for US NBS programs and hemoglobinopathy centers, particularly those located in major metropolitan areas. Among live births in NYC, 61% of all infants with SCD had foreign-born mothers. Public health needs to target outreach efforts to those at-risk communities with relatively high incidences of SCD to identify their medical and social needs. Primary care and obstetrics providers can also use these data to inform their approaches to screening for SCD among immigrant children and women initially presenting for medical services.

The relative proportions of HbSS and HbSC among newborns with SCD vary by maternal country of birth. In particular, the Hb-C allele is most frequent in Burkina Faso, Ghana, and Togo, and in those countries HbSC is more common than HbSS, unlike in most of Africa, where HbSS is the predominant form of SCD.¹⁹ The finding from our study that the majority of infants with SCD born to mothers from Ghana or Togo had HbSC is consistent with information reported previously.¹⁹ The relatively low frequency of HbSC (<20%) among infants with SCD born to foreign-born Hispanic mothers, with the exception of those from the Dominican Republic, appears to be a novel finding.

This study illustrates the value of linking multiple maternal and child health databases using individual-level linking variables.^20,21 Without linking vital records to NBS records, it would not have been possible to ascertain the effect of maternal nativity on SCD incidence rates; this is the first US study to conduct such a linked analysis.

Limitations of the study include missing or incomplete information on the origins of fathers in vital records. Because this information was missing for 15% of cases, analysis was restricted to maternal country of birth. The implication of only including origin of one parent is likely attenuation of the association of parental origins with SCD incidence. However, the degree of misclassification bias appears to be low. For example, this study reported that SCD incidence among infants whose mothers were born in Jamaica was 1 in 180, whereas screening information from Jamaica indicates the incidence to be 1 in 170 births.^17,18 We were not able to examine the effect of marital status on SCD incidence because that information is not collected on the NYS birth certificate. Another limitation is the ambiguity of recording race/ethnicity for people with mixed ancestries. For example, an investigation of six reported cases of SCD among infants born to foreign-born Asian/Pacific Islander mothers (Table 1) identified mothers born in Caribbean countries who were reported to be of Asian race and Chinese ethnicity. Individuals of mixed black and Asian ancestry might have chosen to report Asian race/ethnicity. In addition, clerical errors in recording race/ethnicity are a potential limitation.

Another limitation is that SCD incidence may be underestimated due to our inability to verify the completeness of screening of all newborns in NYS. Because we did not conduct data linkage of all children in the NBS database born in the study period (2000–2008) to birth certificate records, no information about the completeness of screening was available. It was possible that the completeness of screening was <100%. This may lead to a decreased number of SCD cases being identified through NBS. However, this is unlikely to be quantitatively important. A recent study of linked Michigan birth and NBS records during January 2007 to March 2008 found that 99.2% births were successfully matched to NBS records.²² The matching rate increased to 99.6% following a transition to a Web-based electronic birth certificate system and inclusion of a NBS card identification number on the birth record in 2008. The NYS NBS Program is currently developing a data linkage program for matching children in the NBS database to birth records to estimate the completeness of screening and identify unscreened children.

This study provides the first estimates of NYS annual SCD incidence by maternal race/ethnicity and nativity. It has been known for decades that black infants born in the Caribbean or West or Central Africa had much higher rates of SCD than black infants born in the United States due to higher frequencies of the HbS allele. Consistent with those data, this study has demonstrated that NYS newborns of foreign-born non-Hispanic black mothers from those countries had a substantially higher incidence of SCD as compared with NYS newborns of US-born non-Hispanic black mothers. Such findings identify at-risk populations and inform public health education and community outreach activities that promote ongoing, high-quality medical management to affected children.