Sickle Cell Disease and Lead Poisoning in New York City, 2005–2019

We conducted a cross-sectional study assessing children with lead poisoning in NYC during 2005 to 2019. We analyzed data from the NYC DOHMH lead registry, which includes BLL results for all NYC children aged 0 to 17 years tested for lead exposure. All of the NYC laboratories that conduct BLL testing are certified, registered, or permitted by New York State Department of Health’s Wadsworth Center.¹³ New York state law requires health care providers to conduct a BLL test for all children at ages 1 and 2 years, and interview caregivers to assess children for risk of lead exposure until age 6 years; the BLL testing rate in NYC is ∼80% among children aged <3 years. During 2005 to 2019, DOHMH was required by law to investigate cases of lead poisoning among children aged <18 years with venous BLLs ≥15 mcg/dL. Investigations included home inspections, risk assessment interviews with the child’s guardian, and coordination of care with the child’s health care providers by public health nurses. Information collected as part of routine surveillance and case investigations was documented in the lead registry.

For this investigation, we abstracted deidentified data from the lead registry for all NYC children aged 0 to 17 years during 2005 to 2019 with venous BLLs ≥15 mcg/dL into SQL Studio. We identified children with SCD by performing a keyword search of the health care provider reports documented by public health nurses. Each child’s medical information was then manually reviewed by a DOHMH physician to confirm SCD diagnosis (Hb SS, Hb SC, or Hb S-beta thalassemia disease); no children with sickle cell trait were classified as having SCD. Variables assessed included SCD status, BLLs, demographic characteristics (age, gender, Medicaid enrollment status, race and ethnicity, and maternal and child countries of birth), and lead exposure risk assessment data (mouthing and chewing nonfood items, pica, pervasive developmental delay, travel outside United States, use of imported products, occupations and hobbies of household members, paint inspection results, and housing age).

Although SCD can affect people of all ethnicities, it mainly persists in people with African, Mediterranean, Middle Eastern, and South Asian ancestry,⁶ with >90% of people with SCD in the United States identifying as non-Hispanic Black.¹⁴ Similarly, in our data set, the majority of children with SCD (n = 37) self-identified as, or were identified by their parents or caregivers as, non-Hispanic Black; the remainder identified as Hispanic (n = 4) or were missing race and ethnicity information (n = 2). Given that the number of Hispanic children with SCD was too small for meaningful statistical analysis, and that there were no children with SCD from other racial or ethnic backgrounds, we limited our analyses to children who identified as non-Hispanic Black. This was necessary to ensure the most conservative and statistically robust estimates of differences in lead exposure risk factors among children with and without SCD in our data set.

Data analyses were performed using SAS 9.4 (SAS Institute, Cary, NC). Descriptive statistics were calculated for demographic characteristics and SCD status. Two-sample t tests and χ² tests were used to compare demographic characteristics, BLLs, and lead exposure risk factors in Black children with and without SCD. A 1-sample t test was used to compare the observed prevalence of children with SCD in Black children with BLLs ≥15 mcg/dL to the expected prevalence of children with SCD in Black children in the general NYC population.

NYC DOHMH institutional review board approval was obtained for this study. This study was reviewed by the Centers for Disease Control and Prevention and was conducted consistent with federal law and Centers for Disease Control and Prevention policy (45 Code of Federal Regulations, Part 46; 21 Code of Federal Regulations, Part 56; 42 US Code [U.S.C.], Section [Sect.] 241(d); 5 U.S.C., Sect. 552a; 44 U.S.C., Sect. 3501, et seq).

We identified 1728 non-Hispanic Black children in the lead registry with BLLs ≥15 mcg/dL during 2005 to 2019 (Table 1). Thirty-seven (2.14%) children had SCD. Among Black children with BLLs ≥15 mcg/dL, we compared demographic characteristics, BLLs, and lead exposure risk factors of children with SCD with those of children without SCD (Table 2). No significant difference was found in the percentage of children enrolled in Medicaid or living in pre-1960 housing. Notably, the mean age at peak BLL was higher for children with SCD (62.8 months [95% confidence interval (CI) 50.3–75.3] vs 42.7 months [95% CI 41.1–44.2]; P = .003). Although initial BLLs for children with SCD and children without SCD were not significantly different (11.30 mcg/dL [95% CI 5.5–17.1] vs 11.79 mcg/dL [95% CI 11.3–12.3]; P = .86), children with SCD had significantly higher peak BLLs (42.59 mcg/dL [95% CI 28.6–56.6] vs 23.06 mcg/dL [95% CI 22.5–23.6]; P = .008). Additionally, a significantly higher percentage of children with BLLs ≥45 mcg/dL, the level at which hospitalization and chelation is typically recommended, was found among those with SCD (27% [95% CI 13.8%–44.1%] vs 4.7% [95% CI 3.7%–5.8%]; P < .001). Among risk factors for lead exposure, significant differences were found when comparing Black children with and without SCD for the proportion of children reported in the parental questionnaire to be chewing nonfood items (41.4% [95% CI 23.5%–61.1%] vs 10.3% [95% CI 8.5%–12.5%]; P < .001); children reported in the parental questionnaire to be eating soil, paint, or clay (44.8% [95% CI 26.5%–64.3%] vs 15.4% [95% CI 13.1%–17.8%]; P < .001); and children with nursing notes documenting pica in their discussions with parents or providers (60% [95% CI 40.6%–77.3%] vs 18.6% [95% CI 16.3%–21.2%]; P < .001).

When comparing the observed prevalence of SCD in Black children with BLLs ≥15 mcg/dL (2.14%) with the estimated prevalence of SCD among Black children in NYC overall (0.43%),¹⁵ the difference was statistically significant (P < .001; 95% CI 1.5%–2.9%). This difference was more pronounced when limiting the analysis to Black children with BLLs ≥15 mcg/dL born to foreign-born mothers. In this subsample, 19 of 520 children (3.65%) had SCD, whereas estimated prevalence of SCD among children born to foreign-born Black mothers in NYC is 0.63%¹³; this difference was statistically significant (P < .001; 95% CI 2.2%–5.7%).

Our analyses with ∼15 years of NYC lead registry data found a potential association between SCD and lead poisoning in children. Among Black children with BLLs ≥15 mcg/dL, children with SCD had significantly higher peak BLLs, as well as a higher percentage of severe lead poisoning (BLL ≥45 mcg/dL), compared with those without SCD. Furthermore, SCD was significantly more prevalent among Black children with BLLs ≥15 mcg/dL compared with the SCD prevalence estimate in Black children in NYC overall.

Children with SCD were significantly more likely to engage in both pica behavior and chewing nonfood items, compared with those without SCD. This highlights a potentially modifiable risk factor for lead poisoning in children with SCD, because pica can be treated with nutritional and behavioral interventions. Pica refers to the ingestion of nonfood items, which is similar to but distinguishable from chewing and mouthing nonfood items; all 3 behaviors are considered risk factors for lead exposure. Although no studies have assessed the prevalence of chewing or mouthing nonfood items in children with SCD, previous literature has documented a high prevalence of pica behavior in children with SCD; however, the cause remains unclear.¹⁶ 

The reason for the potential association between SCD and lead poisoning among Black children in NYC is unknown. One possible explanation could be a nutritional deficiency contributing to the higher prevalence of pica, and in turn risk for lead exposure, because children with SCD are at risk for micronutrient deficiencies¹⁷; for instance, US children with SCD have been found to have lower mean zinc levels, although not generally in the zinc deficiency range, when compared with children without SCD,¹⁵ and zinc deficiency has been linked to pica.¹⁸ 

Regardless of cause, our finding has significant public health and health policy implications. If a higher risk for lead poisoning in children with SCD is present, as our data suggest, then it would be important to consider raising awareness among both health care providers and caregivers of children with SCD about the potential need for more frequent BLL testing and about addressing behaviors that increase risk for lead exposure. Additionally, we found that the mean age at peak BLL in Black children with SCD was 5 years; this raises the question of whether children with SCD might need to be tested for lead poisoning at older ages than current guidelines recommend (ages 1 and 2 years).

Our investigation has several limitations. First, our data set is limited to children with BLLs ≥15 mcg/dL because this is the level at which DOHMH conducted its investigations in 2005. It is possible that the potential association between SCD and lead poisoning only exists at these BLLs. Next, SCD information in the lead registry relies on reports from health care providers, rather than documentation of laboratory results, so some children with SCD may have been missed. Additionally, the small number of children with SCD restricted our ability to assess the relative contribution of different risk factors in a multivariable model, which might have provided insights into how various factors influence the relationship between SCD and lead poisoning. Furthermore, our analysis was limited to non-Hispanic Black children because of the demographic composition of the children with SCD in our data set; thus, we were unable explore the impact of lead exposure risk factors across diverse racial and ethnic groups. Finally, the cross-sectional nature of this study prevents us from establishing a causal association.

We found a potential association between SCD and childhood lead poisoning, given both the significantly higher peak BLLs in children with SCD and the increased prevalence of SCD among children with lead poisoning compared with the general population, with pica behaviors emerging as a potentially important risk factor. This investigation could have public health implications for childhood BLL testing and lead poisoning prevention guidelines. However, further research is needed to understand this potential association more fully. In NYC, we plan to conduct a future study that links BLL data with SCD results from the New York State Newborn Screening Program; this will allow for more accurate SCD classifications, a larger cohort of children with SCD, inclusion of all racial and ethnic groups in the analyses, and multivariable analyses.

BLL

blood lead level

CI

confidence interval

DOHMH

NYC Department of Health and Mental Hygiene

NYC

New York City

SCD

sickle cell disease

Sect.

Section

U.S.C.

US Code