Matthew Gillman

Matthew Gillman

Harvard Pilgrim Health Care Institute

Harvard epidemiologist to lead revamped NIH children’s study

A seasoned pediatric researcher and epidemiologist has been tapped to head the National Institutes of Health’s (NIH’s) revamped children’s study. Harvard Medical School in Boston’s Matthew Gillman had ties to the National Children’s Study (NCS), which NIH scrapped in late 2014. But Gillman says the new long-term study, which retains some of the original goals, is “set up in a way that will breed success.”

NIH spent $1.3 billion on planning and pilot studies before ending the NCS, which aimed to explore how everything from toxic chemicals to social factors shape the health of 100,000 U.S. children from birth to age 21. On orders from Congress, which called for the original NCS in 2000, NIH replaced it last year with a program called Environmental influences on Child Health Outcomes (ECHO).

One former NCS investigator who blamed its downfall in part on weak scientific leadership says the choice of Gillman, who will start work in July, is good news. “Dr. Gillman is an accomplished child health epidemiologist with extensive experience in pregnancy and birth cohort research. He is ideally positioned to make ECHO a success,” says pediatrician and epidemiologist Nigel Paneth of Michigan State University in East Lansing. In announcing the appointment, NIH Director Francis Collins in Bethesda, Maryland, said Gillman has “deep experience in the fields of epidemiology, pediatrics, and internal medicine.”

For 17 years, Gillman has led a study that looks at how factors such as maternal diet influence the health and development of 1700 children. He also chaired one of the NCS’s many advisory groups and co-led its coordinating center when the NCS was in pilot testing. He sees his new job as a chance to “create a public resource whose whole is greater than the sum of the parts and to do something larger than what I can do from the university setting.”

Unlike the sprawling NCS, ECHO will focus on just four disease areas: obesity; airway conditions such as asthma and allergies; neurodevelopment, including diseases such as autism; and prenatal and postnatal outcomes, such as birth defects. And instead of recruiting participants through doctors or by knocking on doors, ECHO will pull together existing studies of groups, or cohorts, of children.

ECHO is expected to run for 7 years, one-third as long as the original timeline for the NCS. Leaders of participating cohorts will spend 2 years on harmonizing existing data. Those that meet milestones will continue for another 5 years and gather new data. ECHO will also include data analysis and coordinating centers and a pediatric trials network at so-called IDeA states, or states with relatively little NIH funding. Although not part of the cohorts, they will focus on same diseases.

Congress has approved $165 million this year for ECHO, the same amount that had been planned for the NCS, and NIH has requested level funding in 2017. Proposals for the various components of ECHO were due on 15 April, and award decisions will be made by the end of September.

NIH has said the study will include about 50,000 children. Gillman says that number could grow, given that 160 ECHO proposals have been submitted, most from leaders of cohort studies. “I think 50,000 is a reasonable estimate for now, but perhaps it will be even more once we see what the applicants came in with,” he says. Gillman also anticipates that the program will allow for new enrollment, which some researchers say is essential to explore how exposures in the womb shape health.

One of Gillman’s first steps will be to convene a scientific advisory board to help ECHO’s steering committee develop a protocol. The original study used scores of advisory committees that generated an overly ambitious study plan. ECHO will have “fewer committees that have the remit to be part of the direction,” Gillman says. Instead, he says, decision making will be “streamlined” with the NIH director at the top of the chain of command.