Children’s Study Faces Tough Year

The next year will be critical for the National Children’s Study (NCS). Still in its pilot phase after a decade of planning, the ambitious project is facing low enrollment numbers and questions about its environmental testing methods.

Mandated by Congress under the Children’s Health Act of 2000, the study plans to follow 100,000 children from before birth to age 21. It will examine relationships between the environment, genetics, development, and health, with the goal of shedding light on associations between environmental exposures and disorders such as asthma, autism, diabetes, and obesity.

The study is funded on an annual basis and thus needs continued support to survive. New leadership has energized the project, but some observers say it is in trouble.

Congress directed the National Institutes of Health and other federal agencies, including the Centers for Disease Control & Prevention and the Environmental Protection Agency, to plan, develop, and implement the study, but it provided no funding for the study until fiscal 2007.

Funding for the study has picked up significantly under the Obama Administration. Study officials say that the current annual budget of $194 million is sufficient, but they are concerned about future budgets. “We are going to need broad support to continue this study for more than two decades,” says Carol J. Henry, acting chair of the NCS federal advisory committee, a panel of outside experts who provide advice to the federal government about the study.

Over the years, NCS has had mixed support. Several researchers are vested in the study through various oversight committees, and they still see great value in it. The study has the potential to explore links between asthma and air pollutants such as particulate matter, nitrogen oxides, allergens, and endotoxins. It is likely to examine neurodevelopmental problems associated with exposure to various pesticides, as well as endocrine effects associated with long-term exposure to bisphenol A, phthalates, polybrominated diphenyl ether flame retardants, perchlorate, triclosan, perfluorinated compounds, and phytoestrogens.

But there are huge challenges in making associations between exposure to an environmental contaminant and disease. NCS officials realize that the links are weak.

“That’s the conundrum of biomonitoring,” Henry says. “We only understand the relationship between exposure and disease for a few chemicals: lead, mercury, carbon monoxide, and cotinine,” she asserts. “We can measure other chemicals, but we don’t know what it means for health.”

Another problem facing the study is the lack of standard methods for analyzing environmental contaminants in matrixes such as urine, blood, and house dust. “We found a great challenge in the variability and in the validation of many of the environmental methods,” says NCS Acting Director Steven Hirschfeld, a captain in the U.S. Public Health Service and associate director for clinical research at NIH’s National Institute of Child Health & Human Development (NICHD). “We are finding inconsistencies” in how to interpret the results, even with some of the preferred methods, he notes.

To help overcome some of the challenges associated with environmental methods, dozens of NCS study centers located across the country will be engaging in formative research projects over the coming months. Some researchers will be focusing on real-time analysis of biospecimens; others will be examining methods for analyzing house dust.

Each center is under contract with NICHD and has its “support and encouragement to engage in subcontracts with the best available talent and resources,” Hirschfeld tells C&EN. Study officials are particularly interested in learning about environmental methods currently under development and what researchers think they could bring to the study.

Researchers have already collected about 120,000 biospecimen and environmental samples from seven vanguard study centers as part of the study’s pilot phase. They are currently examining the stability of those samples and looking for effects related to collection methods and thawing conditions. They are also examining analytical methods and instrumentation.

“We are in the midst of evaluating the feasibility, acceptability, and cost, as well as the informative value of each measurement method, to ascertain our ability to choose the method that will provide the optimum amount of information at the least cost,” says Michael Dellarco, a senior scientist at NICHD and NCS project officer.

If real-time analysis becomes part of the main study, Hirschfeld estimates that in the short term only a few dozen analytes will be measured. “But the number that could potentially be analyzed based on stored samples could be in the thousands,” he notes.

The prospect of real-time analysis has study officials questioning whether the study’s return-of-results policy needs to be updated. Real-time analysis wasn’t readily available 10 years ago when the policy was developed.

“To date, the assumption has been that the vast majority of samples and data would be analyzed in the future,” Hirschfeld says. At an NCS federal advisory committee meeting in July, committee members debated which results should be returned to study participants and when.

If a result is clinically significant and medically actionable, there’s an ethical obligation to return the results to the participant, NIH officials and committee members agreed. But for chemicals of unknown clinical significance, such as environmental analytes with no agreed-upon safe values, it isn’t clear whether the results should be given to participants.

Many of the challenges facing NCS today were foreshadowed in a 2008 report by the National Research Council and the Institute of Medicine, Henry says. “The issues that we are now intensely discussing come from recommendations in that report,” she tells C&EN.

The report assessed the strengths and weaknesses of the study’s research plan and provided numerous recommendations. Some of those recommendations, such as instituting a pilot phase before the main study and centralizing the data, have already been implemented. Many others are being discussed by various oversight and advisory committees.

The study has also undergone significant changes in leadership since the 2008 report. In July 2009, the study’s first director, Peter Scheidt, stepped down and Hirschfeld took the helm as acting director. A few months later, Duane Alexander, then director of NICHD, announced he was stepping down to serve as an adviser to NIH’s Fogarty International Center. Alan Guttmacher became acting director of NICHD in December 2009 and was named director in July.

The change in leadership came in the wake of a dispute between senators on the appropriations committee and NIH over the ballooning costs of the study. The angry senators accused NIH of committing “a serious breach of trust” by withholding costs related to the study.

Despite the changes and new leadership, some researchers are not convinced that the study is on the right path. In this month’s issue of the journal Epidemiology, two former investigators in the study—David A. Savitz of Mount Sinai School of Medicine and Roberta B. Ness of the University of Texas School of Public Health—call for “fundamental change in the study methods and the decision-making process” (2010, 21, 598).

In their commentary, the two researchers point out that there are numerous advisory groups making recommendations for the study, but it is unclear how major decisions are being made. “Putting the study on track requires an organizational structure with clear lines of responsibility and authority that can promptly finalize and implement a plan,” they write.

NCS does have a lot of advisory committees and multiple layers of oversight, Hirschfeld acknowledges. But he welcomes all of the input. “We’ve been able to, we hope, integrate the input from all of the different perspectives,” he notes.

Ultimate responsibility for the study falls in the hands of the NICHD director. Oversight is also provided by the NIH director, who controls the NCS budget; a federal interagency coordinating committee made up of government officials from participating agencies; a federal advisory committee of external experts; an independent study monitoring and oversight committee; a steering committee; and an executive committee of the steering committee.

With so many layers of oversight, it is difficult to make changes in a timely manner, study officials admit. For example, it became clear in August 2009 that knocking on doors to identify pregnant women eligible for the study may not be the best recruitment strategy. But it took nearly a year to implement alternative strategies.

As of June, 1,569 women had enrolled in the study via the door-to-door approach. Of these, 832 were pregnant, 325 were actively trying to become pregnant, and 412 women had given birth.

“We were expecting about 3.5 times that,” Hirschfeld tells C&EN. Assumptions regarding pregnancy rates “weren’t as robust as we’d hoped,” he says. Although the study could be conducted using the knock-on-the-door recruitment method, “it would take much longer than anticipated, and it would be more resource intensive,” he says.

To boost enrollment numbers, 30 new study centers are now actively recruiting new participants using alternative strategies. Some recruiters are continuing to knock on doors, but some are contacting pregnant women through their prenatal care providers, and some are mailing out questionnaires. They will continue their efforts until at least next summer.

“We hope that by increasing the number of study centers from seven to 30 and intensifying the effort, we will have at a minimum four times the amount of data,” Hirschfeld notes. And data are what drive the design of the main study, which will have about 100 study centers, he emphasizes. “We have shifted our model to one that is completely data-driven and evidence-based.”

Under Hirschfeld’s leadership, the study’s business model has also been overhauled. “We recognized that the process was so resource intensive that it was going to be unaffordable and unsustainable to scale it up from seven to 100 sites,” Hirschfeld says.

Despite numerous challenges facing the study, its leadership is confident that the study is moving in the right direction. Empirical data will be gathered at 30 centers over the next year or so, and that data will be used to inform the main study. “We frame what has occurred in the past as successful because we learned from it,” Hirschfeld tells C&EN. “It should not be viewed as a failed attempt. It should be viewed as an informative attempt that led us to take next steps.”