How many children in the United States suffer from lead poisoning? Tracking California discovered that the numbers of lead-poisoned children may be much higher than previously reported.
Using new analytical methods, Tracking California estimated that during 1999-2010, most states may have missed more than half of their lead-poisoned children.Nationally, only 64% of lead-poisoned children were identified. In California, only 37% of lead-poisoned children were identified.
The findings suggest that recent lead poisoning estimates that are based solely on clinical data are too low and many lead-poisoned children remain undiagnosed and untreated. Explore results through our interactive maps of predicted lead exposure and testing rates.
Lead poisoning and its public health impacts are completely preventable. However, lead-based paint and lead dust remain in millions of older homes in the United States and continue to harm the developing brains of children. Other sources of lead include lead pipes, contaminated soil, and consumer products. While we should not wait until children are poisoned to find and remove hazardous lead from the environment, identifying all lead-exposed children (children with lead in their bodies) is essential to addressing this issue.
An incomplete picture of lead poisoning
Commonly cited estimates of lead poisoning in children are based on results of blood tests conducted by medical providers. These estimates are incomplete because lead testing is not required for all children in the U.S., testing guidelines vary by state, and not all states report lead testing data. To assess the scope of the problem, Tracking California developed a statistical model using data from the National Health and Nutrition Examination Survey (NHANES) to estimate the actual number of lead-exposed children across the U.S. by state.
How many children are falling through the cracks?
To better understand how many lead-poisoned children may have been missed, we compared our estimates of the true number of lead-poisoned children with the numbers of children diagnosed with lead poisoning and reported to the Centers for Disease Control and Prevention (CDC) from 1999-2010. For this analysis, lead poisoning was defined as having a blood lead level of 10 ug/dL or higher among children 1-5 years old.
State-level estimates of the number of children with lead exposure have been based on results of lead tests conducted by medical providers. Therefore, the actual number of lead-exposed children in the each state is unknown. Using advanced statistical methods, we generated estimates of blood lead prevalence in order to better understand the scope of the problem. The methods described below were published in the journals Statistics in Medicine and Pediatrics.
We developed a statistical model that used 1999-2010 data from the National Health and Nutrition Examination Survey (NHANES) to predict the prevalence of elevated blood levels (EBLL) in children. This model considered factors such as race/ethnicity, household poverty, age of housing, geographic region, and year. However, some data are missing from NHANES in ways that would introduce bias to the model. Specifically, data on age of housing is missing for over a third of NHANES participants, and those participants are more likely to live in an older home. Therefore, we developed statistical methods to overcome this bias. Our methods were published in Statistics in Medicine.
Using our statistical model and data from the U.S. Census Bureau, we estimated the numbers of children with EBLL (measured as micrograms of lead per deciliter of blood) for 1-5 year olds at three different blood lead levels: >= 2.5 ug/dL and above, >= 5 ug/dL, and >= 10 ug/dL. Our results were published in Statistics in Medicine and are displayed as state and local area maps.
To assess the extent to which medical providers in each state are identifying children with EBLL, we compared our prevalence estimates with the numbers of lead-poisoned children reported by states to the Centers for Disease Control and Prevention (CDC) from 1999-2010. If a state did not report to CDC throughout this entire time period, we only compared the years for which data were reported. For this analysis, we looked at children age 1-5 years with an EBLL of 10 ug/dL and higher. This was CDC's reference level during 1999-2010 and was the level most consistently reported by states. Our results were published in Pediatrics and can be explored on our interactive maps.
We estimated that 1.2 million children had lead poisoning from 1999-2010. In states that reported data to the CDC during this time, there were 944,000 lead-poisoned children, of whom only 607,000 were identified and reported. This suggests that:
Our results indicate that:
This research was conducted by Tracking California, a program of the Public Health Institute. This work was supported by Cooperative Agreement Number 5U38EH000953 from the CDC. Its contents are solely the responsibility of the program and do not necessarily represent the official views of the CDC.