Reducing Elevated Blood Lead Levels in Children in North Carolina & Vermont
Reducing Elevated Blood Lead Levels in Children in North Carolina & Vermont
Background: Few studies have examined factors related to the time required for children's blood lead levels (BLLs) ≥ 10 μg/dL to decline to < 10 μg/dL.
Objectives: We used routinely collected surveillance data to determine the length of time and risk factors associated with reducing elevated BLLs in children below the level of concern of 10 μg/dL.
Methods: From the North Carolina and Vermont state surveillance databases, we identified a retrospective cohort of 996 children < 6 years of age whose first two blood lead tests produced levels ≥ 10 μg/dL during 1996-1999. Data were stratified into five categories of qualifying BLLs and analyzed using Cox regression. Survival curves were used to describe the time until BLLs declined below the level of concern. We compared three different analytic methods to account for children lost to follow-up.
Results: On average, it required slightly more than 1 year (382 days) for a child's BLL to decline to < 10 μg/dL, with the highest BLLs taking even longer. The BLLs of black children [hazard ratio (HR) = 0.84; 95% confidence interval (CI), 0.71-0.99], males (HRmale = 0.83; 95% CI, 0.71-0.98), and children from rural areas (HRrural = 0.83; 95% CI, 0.70-0.97) took longer to fall below 10 μg/dL than those of other children, after controlling for qualifying BLL and other covariates. Sensitivity analysis demonstrated that including censored children estimated a longer time for BLL reduction than when using linear interpolation or when excluding censored children.
Conclusion: Children with high confirmatory BLLs, black children, males, and children from rural areas may need additional attention during case management to expedite their BLL reduction time to < 10 μg/dL. Analytic methods that do not account for loss to follow-up may underestimate the time it takes for BLLs to fall below the recommended target level.
Studies reporting an inverse relationship between children's blood lead levels (BLLs) of 10-24 μg/dL and cognitive function/performance (Bellinger et al. 1991; Ruff et al. 1993) motivated the Centers for Disease Control and Prevention (CDC) to lower the level of concern from a BLL ≥ 25 μg/dL to ≥ 10 μg/dL in 1991 (CDC 1991). Although prevention efforts have resulted in an 86% decline in BLLs ≥ 10 μg/dL in children in the United States from the 1970s to the present, an estimated 310,000 children < 6 years of age still have BLLs ≥ 10 μg/dL (Schwemberger et al. 2005). A key prevention strategy has been to identify children with BLLs ≥ 10 μg/dL and to enroll them in case management to reduce their lead burdens.
Collecting blood lead is considered to be the most useful tool for screening and diagnostic testing (Agency for Toxic Substances and Disease Registry 2005). Blood lead levels correlate most closely with recent environmental exposure. The excretory half-life of lead in adult blood is approximately 36 days (Todd et al. 1996), but estimates in children are longer (Manton et al. 2000). Prolonged exposure, which can come from both internal and external sources, prevents BLLs from declining to lower levels. The most common external source of exposure in children is deteriorated lead-based paint dust (CDC 1997). The most common internal exposure source is from lead reservoirs in bone, which are the largest component of body burden, accounting for 70% of all lead in children, and from where replenishment of lead in the blood occurs. Half-life of lead in bone is usually measured in years or even decades (Graziano 1994).
Although identification of the factors influencing the time to reduce a child's BLL is essential to improving case management, few studies have investigated this issue. New York City investigators estimate that, on average, it takes 6-12 months for children's BLLs to decline from ≥ 20 μg/dL to < 10 μg/dL, when medical and environmental management is consistent with CDC recommendations for managing children with elevated BLLs (Matte T, personal communication).
An analysis of Wisconsin surveillance data from 1995 to 1997 showed that BLLs decreased 2.6 μg/dL per year, and the mean time for decline among children with confirmed BLLs ≥ 20 μg/dL to drop below 10 μg/dL was just over 4 years (Wisconsin Childhood Lead Poisoning Prevention Program, unpublished data). A study published in 2001 found that those with higher peak BLLs take longer to decline to < 10 μg/dL (Roberts et al. 2001). Specifically, children with BLLs of 25-29 μg/dL required on average 24 months to fall below 10 μg/dL compared with children with BLLs of 10-14 μg/dL, who required 9.2 months. However, those findings may have been biased because the analysis did not take into account the contribution of children whose BLLs did not fall below 10 μg/dL. A more recent study reported times of decline of 11.6 months in the 10-14 μg/dL category and 12.7 months in the 15-19 μg/dL category among 2,109 children ≤ 3 years of age from six states enrolled in case management; however, the authors also excluded children who were censored in their analysis (Whitehead and Leiker 2007).
To assess factors related to the time required for children's BLLs ≥ 10 μg/dL to decline, we analyzed routinely collected blood lead surveillance data from lead poisoning prevention programs in Vermont and North Carolina. We compared three different analytic approaches: a) a central analysis that included censoring children whose BLLs did not fall below 10 μg/dL at the time of their last test; b) an analysis that simply excluded children who were censored; and c) a sensitivity analysis that inferred the time for children's BLLs to fall below 10 μg/dL on the basis of linear interpolations between the times of the last measurement above and the first measurement below the level of concern.
Background: Few studies have examined factors related to the time required for children's blood lead levels (BLLs) ≥ 10 μg/dL to decline to < 10 μg/dL.
Objectives: We used routinely collected surveillance data to determine the length of time and risk factors associated with reducing elevated BLLs in children below the level of concern of 10 μg/dL.
Methods: From the North Carolina and Vermont state surveillance databases, we identified a retrospective cohort of 996 children < 6 years of age whose first two blood lead tests produced levels ≥ 10 μg/dL during 1996-1999. Data were stratified into five categories of qualifying BLLs and analyzed using Cox regression. Survival curves were used to describe the time until BLLs declined below the level of concern. We compared three different analytic methods to account for children lost to follow-up.
Results: On average, it required slightly more than 1 year (382 days) for a child's BLL to decline to < 10 μg/dL, with the highest BLLs taking even longer. The BLLs of black children [hazard ratio (HR) = 0.84; 95% confidence interval (CI), 0.71-0.99], males (HRmale = 0.83; 95% CI, 0.71-0.98), and children from rural areas (HRrural = 0.83; 95% CI, 0.70-0.97) took longer to fall below 10 μg/dL than those of other children, after controlling for qualifying BLL and other covariates. Sensitivity analysis demonstrated that including censored children estimated a longer time for BLL reduction than when using linear interpolation or when excluding censored children.
Conclusion: Children with high confirmatory BLLs, black children, males, and children from rural areas may need additional attention during case management to expedite their BLL reduction time to < 10 μg/dL. Analytic methods that do not account for loss to follow-up may underestimate the time it takes for BLLs to fall below the recommended target level.
Studies reporting an inverse relationship between children's blood lead levels (BLLs) of 10-24 μg/dL and cognitive function/performance (Bellinger et al. 1991; Ruff et al. 1993) motivated the Centers for Disease Control and Prevention (CDC) to lower the level of concern from a BLL ≥ 25 μg/dL to ≥ 10 μg/dL in 1991 (CDC 1991). Although prevention efforts have resulted in an 86% decline in BLLs ≥ 10 μg/dL in children in the United States from the 1970s to the present, an estimated 310,000 children < 6 years of age still have BLLs ≥ 10 μg/dL (Schwemberger et al. 2005). A key prevention strategy has been to identify children with BLLs ≥ 10 μg/dL and to enroll them in case management to reduce their lead burdens.
Collecting blood lead is considered to be the most useful tool for screening and diagnostic testing (Agency for Toxic Substances and Disease Registry 2005). Blood lead levels correlate most closely with recent environmental exposure. The excretory half-life of lead in adult blood is approximately 36 days (Todd et al. 1996), but estimates in children are longer (Manton et al. 2000). Prolonged exposure, which can come from both internal and external sources, prevents BLLs from declining to lower levels. The most common external source of exposure in children is deteriorated lead-based paint dust (CDC 1997). The most common internal exposure source is from lead reservoirs in bone, which are the largest component of body burden, accounting for 70% of all lead in children, and from where replenishment of lead in the blood occurs. Half-life of lead in bone is usually measured in years or even decades (Graziano 1994).
Although identification of the factors influencing the time to reduce a child's BLL is essential to improving case management, few studies have investigated this issue. New York City investigators estimate that, on average, it takes 6-12 months for children's BLLs to decline from ≥ 20 μg/dL to < 10 μg/dL, when medical and environmental management is consistent with CDC recommendations for managing children with elevated BLLs (Matte T, personal communication).
An analysis of Wisconsin surveillance data from 1995 to 1997 showed that BLLs decreased 2.6 μg/dL per year, and the mean time for decline among children with confirmed BLLs ≥ 20 μg/dL to drop below 10 μg/dL was just over 4 years (Wisconsin Childhood Lead Poisoning Prevention Program, unpublished data). A study published in 2001 found that those with higher peak BLLs take longer to decline to < 10 μg/dL (Roberts et al. 2001). Specifically, children with BLLs of 25-29 μg/dL required on average 24 months to fall below 10 μg/dL compared with children with BLLs of 10-14 μg/dL, who required 9.2 months. However, those findings may have been biased because the analysis did not take into account the contribution of children whose BLLs did not fall below 10 μg/dL. A more recent study reported times of decline of 11.6 months in the 10-14 μg/dL category and 12.7 months in the 15-19 μg/dL category among 2,109 children ≤ 3 years of age from six states enrolled in case management; however, the authors also excluded children who were censored in their analysis (Whitehead and Leiker 2007).
To assess factors related to the time required for children's BLLs ≥ 10 μg/dL to decline, we analyzed routinely collected blood lead surveillance data from lead poisoning prevention programs in Vermont and North Carolina. We compared three different analytic approaches: a) a central analysis that included censoring children whose BLLs did not fall below 10 μg/dL at the time of their last test; b) an analysis that simply excluded children who were censored; and c) a sensitivity analysis that inferred the time for children's BLLs to fall below 10 μg/dL on the basis of linear interpolations between the times of the last measurement above and the first measurement below the level of concern.
