Reduced teen pregnancy and violence: is lower lead exposure part of the explanation?

Since I started my career in 1975, the teen birth rate has seen substantial (and welcome) declines, from 55.6 births per 1,000 females ages 15–19 in 1975 to 24.2 in 2014—less than half the 1975 level.

Similarly, youth violence has also declined significantly over this time period. For example, the rate of serious violent crime among juveniles ages 12–17 declined from almost 40 per 100,000 in the early 1980s to 7.6 in 2015, and the homicide rate among youth ages 18–24 has fallen substantially since the early 1990s.[1] While levels of violence vary considerably across locations and populations, the overall declines are historic and have occurred across regions and population groups.

Many organizations, agencies, local programs, parents, teachers, and health care providers have made invaluable contributions to these declines. And while we have never been able to fully parse out the reasons for these changes, some fascinating research suggests a potentially powerful explanation: lower exposure to lead among young children may contribute to reduced risk factors later in life.

I recently worked on a project on lead exposure with The Health Impact Project, a collaboration of the Robert Wood Johnson Foundation and The Pew Charitable Trusts. I knew at the project’s onset that lead is toxic to the nervous system and affects cognitive development and behavior. I also knew that, with some exceptions, lead had been removed from various products (e.g., most paint, fuel, and children’s products) beginning in the late 1970s. But I was unaware of other facts—for example, that the phase-out of leaded gas took place over two decades (primarily from 1975–1985).

Ample research documents the harm done to young children who are exposed to lead; greater exposure to lead causes neurological damage, impulsivity, conduct disorders, behavior problems, and lower inhibition. However, because I study youth, I was interested in whether these early difficulties continue to have negative implications for teenagers. Sadly, there is little research about the effects of early exposure to lead on teen and young adult outcomes. One study by Rick Nevin followed cohorts over time and reported that rates of violence fell about 20 years after declines in blood lead levels. The alignment between lower levels of lead in the environment and lower rates of teen childbearing and violence is remarkable, but only correlational.

Of course, it is unethical to conduct an experimental study in which some kids are assigned to be exposed to lead and others are not. But analyses of longitudinal data are possible. (Note to reader: this is about to get a little wonky.) A very sophisticated peer-reviewed study by Jessica Reyes used the National Longitudinal Survey of Youth (NLSY) (both the 1979 and 1997 cohorts) to examine teen behaviors. Reyes used data on average levels of gasoline lead in states (by year and month) in the late 1970s and early 1980s and linked it with children’s data in the National Health and Nutrition Examination Survey by age, sex, income, and race. She then applied these estimates of lead exposure to the NLSY samples when the kids were ages 0–3, matching each individual by age, year, and state to get an estimated level of lead. Finally, she studied their behaviors a decade or more later, when they were teens.

Reyes found a strong association between estimated lead exposure in early childhood and the incidence of teen parenthood and violence more than a decade later. Recognizing that exposure to lead and other adverse experiences tend to coincide, she statistically accounted for economic and social disadvantages often associated with violence and pregnancy.

In short, Reyes found that exposure to higher levels of lead from gasoline in early childhood is associated with a greater risk of teen sex and pregnancy, higher alcohol and marijuana use, and higher levels of various forms of violence. These effects are substantial: “A change that approximates the population-wide reduction that resulted from the phaseout of lead from gasoline … yields a predicted 12 percentage point decrease in the likelihood of pregnancy by age 17, and a 24 percentage point decrease in the likelihood of pregnancy by age 19 (from a 40 percent chance to a 16 percent chance).”

We know that the factors underlying negative behaviors are complex, but Reyes’ peer-reviewed study suggests that we have overlooked an important environmental contributor to problem behaviors. In addition, the scientific literature describes multiple mechanisms through which lead exposure can increase criminal behavior, consistent with work linking biological, psychological, and social factors to increased crime.

Dramatic declines in children’s exposure to lead may contribute to less risk-taking among many of today’s youth, relative to previous generations. If so, it behooves us to recognize that lead exposure remains a risk factor in many disadvantaged communities. We should now strive to eradicate exposure entirely.

[1] An upward spike was documented in the early 1990s for both teen childbearing and violence, but the long-term trends have been downward.