2

Informing Decision Making: Discussion of the Science and Critical Gaps

The session began with presentations from Tracey Woodruff from the University of California—San Francisco (UCSF), who discussed why children’s environmental health still matters. Aaron Bernstein, Harvard University, discussed the impact of climate change on children’s environmental health and his research efforts to improve civic engagement on the topic. Zhiwei Xu, University of Queensland, discussed the impact of heatwaves on children’s environmental health. The session closed with a panel discussion on critical gaps that hold back progress in environmental health policy. Brenda Eskenazi, the University of California—Berkeley, Mark Miller, California EPA, and Thomas Burke, Johns Hopkins University, participated. The entire session was moderated by Mona Hanna-Attisha, Michigan State University and Hurley Children’s Hospital Pediatric Public Health.

WHY ADDRESSING ENVIRONMENTAL THREATS TO CHILDREN’S ENVIRONMENTAL HEALTH STILL MATTERS¹

Despite progress on children’s environmental health issues, such as the improvements to child health and economic gains from policies that reduced lead exposure, a great deal of work remains to improve children’s health, asserted Woodruff. Children’s chronic disease risks are increasing. The number of children with leukemia has increased by 35 percent over

___________________

¹ This section is based on a presentation by Tracey Woodruff, professor, UCSF.

the last 40 years, 8.4 percent of children have asthma, and 1 in 42 boys have autism spectrum disorder (EPA, 2017). U.S. chemical production has increased 15-fold from 1945 to 2007, and 30,000 pounds of industrial chemicals are produced per capita; this includes PFAS, phthalates, phenols, and flame-retardants, which are known to be toxic. Health care professionals widely recognize that environmental exposures during development may increase health inequities (ACOG, 2021; Di Renzo et al., 2015; Zoeller et al., 2012).

Many information gaps remain regarding the impacts of chemical exposures on children’s health and development. “There are 9.5 trillion pounds of chemicals produced in the United States every year. There are 350,000 chemicals globally, and about 40,000 are registered in the U.S. under TSCA. Yet, a tiny fraction of these chemicals have any information about their health effects or understanding of their exposures, at least using a biomonitoring technique,” explained Woodruff.

Woodruff described an important opportunity that could fill in some of these knowledge gaps. The Environmental Influences on Child Health Outcomes (ECHO) program funded by the National Institutes of Health (NIH) is a synthetic cohort of about 50,000 children created by merging about 70 pediatric and perinatal cohorts across the country. ECHO provides an opportunity to identify and prioritize chemicals that have not been measured in biomonitoring studies (Pellizzari et al., 2019). Recently, an ECHO study, including a diverse group of pregnant women enrolled over 12 years, found rising exposure to chemicals from plastics and pesticides that may harm child development. Forty-five chemicals were detected in over 50 percent of pregnancy urine samples, with Hispanic women having increased exposure to many chemicals compared to other ethnicities. Many exposures are to replacements for chemicals of concern (Buckley et al., 2022).

The TSCA amendments include language requiring the EPA to consider developmental susceptibility. In risk evaluations, the EPA is now required to evaluate the potential for susceptibility to children and pregnant women from chemical exposures. In addition, the EPA can require industry to provide data to better document exposures and susceptibilities during development. Yet, the EPA has not fully met requirements in the law nor taken advantage of these new powers, said Woodruff. For example, it recently completed a risk evaluation for 1-bromopropane, a dry-cleaning solvent in widespread use. Although it identified that a single exposure during a critical window of fetal development may be sufficient to produce adverse developmental effects, it “did not calculate risk for children associated with acute exposure at dry cleaners” (EPA, 2020, p. 114), thus ignoring that children could be present in a dry cleaner, such as a family business. It also did not calculate risks for chronic exposure for children at dry cleaners because it believed “exposure to children at workplaces are unlikely to be chronic in nature” (EPA, 2020, p. 114). This assumption ignores situations where families may live on top of the facility or parents may bring their children to work.

According to Woodruff, the EPA has also not fully accounted for the population variability in TSCA risk evaluations. Typically, an adjustment factor of 10 is used, although a National Academies report has said this may not address the full range of variability in response to chemical exposures (NRC, 2009). Moreover, the FQPA requires an additional adjustment factor for age-related susceptibility, but TSCA does not. Thus FQPA provides an opportunity to upgrade adjustment factors, which other agencies have done. For example, the California Office of Environmental Health Hazard Assessment has increased adjustment factors from 10 to 30 depending on age-specific differences in chemical metabolism between children and adults.

Woodruff concluded that newer exposure data find multiple chemical exposures, some of which appear to be increasing and can influence children’s health. Public policies can address children’s health, but the onus must be on those who financially benefit from chemical sales to pay for chemical safety testing, says Woodruff. Systematic changes will upgrade existing approaches to integrate the best available science in hazard and risk assessment, including making decisions on early indicators of harm.

Children’s Environmental Health, Research, and Civic Engagement²

According to Bernstein, environmental health concerns are not at the top of parents’ minds. This lack of attention among parents to the environment as a driver of health demonstrates that parents often do not understand the connection, and environmentally attributed diseases are common among children. Traffic-related air pollution, such as NO₂, significantly contributes to asthma development (Achakulwisut et al., 2019). In some geographic locations, such as downtown Oakland, where more than 70 percent of the population is people of color, the impact of traffic on asthma is even more pronounced, as about 50 percent of asthma cases are associated with air pollution. However, in Oakland Hills, where less than 30 percent of the population is of color, air pollution accounts for only about 20 percent of asthma cases, explained Bernstein.

Another example of children’s health concerns is that climate change has very different effects based on when someone is born. A child born today will experience a greater lifetime exposure to extreme events than previous generations have. Using a cohort approach based on birth year, scientists estimate that children born in 2020 will experience a two- to sevenfold increase in extreme events, particularly heatwaves, compared with people born in 1960, under current climate policy pledges (Thiery et al., 2021). Health equity requires climate action, said Bernstein. Unfortunately, the scholars who study health equity are substantially disconnected from those who focus on environmental health.

Another area that has solid ties to the environment is mental health. An international survey of 10,000 people aged 16 to 25 (1000 people in each of 10 countries) showed astonishingly high levels of worry about climate change. Many young people described that “their concerns about climate change impacted their daily functioning.” The countries where the highest proportion of respondents (74 percent) described climate anxiety as affecting daily functioning were India and the Philippines. Although the U.S. proportion was much lower, it was still shockingly

___________________

² This section is based on a presentation by Aaron Bernstein, professor, Harvard T.H. Chan School of Public Health.

high, with 26 percent of young people reporting climate anxiety’s impact on functioning (Hickman et al., 2021).

Bernstein’s fundamental question is, “How do we capture these environmental health concerns of children in our policies?” Tools that can help include the social cost of carbon analysis, cost-benefit analysis, and assessment of state- and city-level policies. The first is a value placed upon CO₂ emissions to encapsulate the externalized cost of greenhouse gas emissions on society; calculations typically do not account for health impacts due to data gaps. The second is complicated; it is easy and quick to describe the costs, but it is rare and much more challenging to consider and account for the benefits.

Lastly, the lack of consideration of children in policy decisions is not just an environmental health problem. Children receive a tiny slice of support from government agencies. Investments in children have been trending downward for about a decade. A spike in spending followed the Great Recession, mainly due to federal relief efforts during the pandemic, but spending on children is expected to go back to pre-pandemic levels (Daly et al., 2021). Bernstein added while there is a disconnect in the United States between health equity researchers and children’s environmental health, there is also this “under investment of our children more broadly.” Children’s environmental health cannot be improved if only pollution and climate change are addressed; the social environmental factors, such as poverty and adverse childhood experiences, also can be hazardous to children.

The Center for Climate, Health, and the Global Environment (C-CHANGE) at Harvard has begun to think about ways of changing thinking and mindset around children’s environmental health. Since the 1990s, Gallup surveys have asked U.S. adults about their understanding of climate. Public understanding of climate issues has improved over the last 20 years, with most of those improvements among those who grew up learning about climate change, a reflection of the benefit of education on children’s environmental health, said Bernstein. Another important point is that most Americans support climate change solutions to protect public health and trust health professionals for climate change information (Speiser, 2021).

Health care contributes to climate change, Bernstein raised. An equitable, patient-centered approach to climate action could enable health care to meet these obligations and its responsibility to deliver high-quality care for the betterment of all (Bernstein et al., 2022). In Bernstein’s opinion, clinics do not need to take on decarbonization. Still, they can address climate by reducing waste in the medical system by avoiding unnecessary tests and treatments or solarizing federally qualified health clinics, as in Puerto Rico. Many of those solarized health centers were able to remain

operational during a blackout in April 2022, improving the consistency of health care operations on the island and reducing emissions of greenhouse gases and other air pollutants.

Health care clinics are the glue that holds communities together after disasters, which is why Bernstein said that C-CHANGE will release a series of reports on patient-centered climate action. The first set will be focused on heat, followed by wildfires. The series is focused on frontline providers, not population health systems. These providers and all the non-profits that work with them have trusted relationships in their communities and often provide medication, access to water, and even housing. Bernstein and his team are working on these issues because he believes we radically need to change how we address emergency response. The National Weather Service sets heat warnings at temperatures that are too high. The media then does not reach the people at risk, and then cooling centers are opened that vulnerable people (which includes pregnant persons and children) do not use. Clinicians must discuss climate-related health issues with their patients because clinicians are essential allies in addressing foundational concerns around climate change.

Bernstein then discussed the need to transform the ways that knowledge is translated into action. In most academic settings, research is conducted, an organization publishes a press release, a media outlet may or may not pick it up, and an audience reads it. Recently, Bernstein and colleagues have partnered with Peak Action, an organization that focuses on social media campaigns, to identify a group of 15 “climate creators” to watch.³ They range from people with Ph.D.s in climate science to inter-sectional race ethnicity activists. They reach many more people than a press release from an academic institution usually does and engage with younger and more diverse people. “The power of this alliance is that we can define how knowledge gets into the hands of people it matters to,” said Bernstein. Social media influencers get directly to audiences. Because of their different personas, some reach very targeted audiences that are politically active but would never read a press release or a newspaper or watch TV. Bernstein cautioned that his center’s influencers likely have more scientific training than the average media outlet.

The last shift that Bernstein discussed came from a conversation between Gina McCarthy, the nation’s climate czar, and former governor of Vermont, Peter Shumlin. They created the Harvard Chan C-CHANGE Youth Summit on Climate, Equity, and Health. The summit, which just finished its second year, allows 125 high school students from around the country to engage with climate leaders, such as Howard Koh, the former assistant secretary of health and human services and commissioner for

___________________

³https://climatecreatorstowatch.com/ (accessed December 19, 2022).

public health in Massachusetts. The youth who have attended these summits ask informed questions and are on top of the subject matter.

C-CHANGE has been working on a large but growing effort to communicate how climate actions benefit health and equity. Bernstein added, “it is critically important to not just focus on gloom and doom but also to show the path forward, to encourage public buy-in.”

Heat and Children’s Health⁴

Climate change is arguably the most significant public health threat, explained Xu. It leads to extreme weather events, rainfall, rising temperature, and other issues. Climate both directly and indirectly affects children’s health. Direct impacts include temperature change, heatwaves, precipitation and floods, droughts, and bush fires. Indirect impacts include changes to the ecosystems and changes in vector patterns. Two ways to cope with climate impacts exist; mitigation and adaptation. Mitigation involves steps to prevent or reduce greenhouse gas emissions. Adaptation means adjusting the system or behaviors to adapt to the changing climate in a way that will reduce the health impacts of climate (Helldén et al., 2021).

Children are more sensitive to heat for many reasons, described Xu. First, they have a greater body surface area–to–mass ratio than adults. Second, they spend more time outdoors and participate in more vigorous activities. Third, children, especially the very young, cannot take care of themselves and are highly dependent on their caregivers. Fourth, children have far more expected future years of life and will be facing far more years of exposure to heat. Heat affects children across their different life stages, from before they are born, to infancy, to early childhood, through adolescence. If heat exposure causes a health condition in childhood, some may suffer it for life (Perera and Nadeau, 2022).

Much evidence supports the impact of heat on adverse perinatal outcomes. A recent systematic review found consistent evidence of an association between heat exposure and an increased likelihood of adverse pregnancy outcomes (Chersich et al., 2020). For a one-degree increase in temperature, the likelihood of preterm birth increased by 5 percent. Among people exposed to heat waves, the risk of preterm birth increased by 16 percent; with a one-degree temperature increase, the risk of still birth also increased by 5 percent. Despite a greater amount of heterogeneity in effect estimates, positive associations were observed between heat exposure and low birthweight (Chersich et al., 2020).

Heat also impacts infant health, said Xu. Using hospital admissions data from 2005 to 2015 from Queensland Health, Xu and his colleagues

___________________

⁴ This section is based on a presentation by Zhiwei Xu, research fellow, Griffith University.

examined how heatwaves affect hospital admissions in infants (Xu et al., 2017). They found that when the daily temperature was above 32° C (97th percentile in Brisbane) for more than 2 days, the risk of admissions increased by about 5 percent. When the high temperature lasted 4 days, it increased by 18 percent (Xu et al., 2017). Another study in Spain found that heatwaves were associated with a 25 percent increase in infant mortality, most deaths were associated with conditions that originated prenatally, and the mortality risk was slightly increased among girls (Basagaña et al., 2011). Studies in Canada and the United States have found that the risk of sudden infant deaths dramatically increases when temperature increases (Auger et al., 2015; Jhun et al., 2017).

Research has also shown connections between pediatric morbidity and extreme ambient temperatures. A recent scoping review of the literature on heat and pediatric outcomes in a U.S. urban population identified associations between heat and heat stroke, electrolyte imbalance, diarrhea, infectious diseases, and asthma (Uibel et al., 2022). Xu suspects that the underlying mechanisms vary. Heat may increase diarrhea and infectious diseases due to increasing exposure to pathogens, but the mechanism between heat and asthma may differ. Xu says his colleagues in Brisbane have also found an association between heat exposure and the risk of childhood pneumonia, diarrhea, and asthma.

Population vulnerability is also an important consideration, said Xu. Three factors determine this: exposure level, sensitivity to heat, and adaptive capacity. Although children are sensitive to heat, good adaptive practices can help protect them, such as heat health action plans or heatwave response plans. For example, Queensland has a Heatwave Management Plan: children, especially young children, are listed as a vulnerable subgroup, and a series of actions will be taken in the face of heatwaves to protect them. Community strategies, such as school-based adaptation programs, can also help. For example, schools could have guidelines for outdoor activities during heat waves, such as flexible scheduling. Individual-level adaptation strategies can also be helpful; these include parents choosing to limit their children’s exposure to heat. However, even the best heat adaptation strategies or policies will be less effective if the vulnerable populations are not motivated to act.

Xu recommended several areas for future research. A top priority is adaptive strategies co-designed with stakeholders—children, their caregivers, pediatricians, child care professionals, and decision makers. This collaboration will improve understanding of the barriers to action during heatwaves. Another critical research gap is evaluating the efficacy of the existing strategies to confirm that they are having the expected impacts on child morbidity and mortality. Xu also noted a need for more large epidemiologic studies to improve the generalizability of results, a

more consistent definition of heat, and robust statistical methods. Most importantly, Xu echoed other presenters by noting the need for improved translation of scientific findings. Animated videos, advertisements, and other consumable media may spread knowledge about the impacts of heat and heatwaves on child health more quickly than traditional media.

Xu shared a famous Chinese fable. A 90-year-old man living near the mountains was rankled by the obstruction caused by the mountains and decided to dig through them. Later, when his neighbors questioned whether he would finish, he replied, “If I can’t finish that, my children will be able to finish that. And if they can’t, my grandchildren, and my great grandchildren, they will continue until it is finished.” Of the fable, Xu said,

PANEL DISCUSSION: CRITICAL GAPS THAT HOLD BACK PROGRESS IN ENVIRONMENTAL HEALTH POLICY

Hanna-Attisha allowed each panelist to reflect on the day’s presentations. Brenda Eskenazi, of the University of California—Berkeley, described that the major theme was that young children and fetuses are not well protected from environmental hazards. Eskenazi said that Bernstein’s presentation about engaging youth with the most energy and the most to lose resonated with her.

Mark Miller of the California EPA said what resonated most for him was that “there is this intersection between climate change, plastics, air pollution, and chemical pollution and that the intersection magnifies the health impacts of each of these. There is a history in environmental health of identifying a problem and dealing with it within a particular industry, only to have it reemerge later in a different context. And we see that over and repeatedly.” Miller added that the underinvestment in children in general especially struck him. The social and economic underinvestment in children is a crucial backdrop given the interactions between chemical exposures and climate change factors and that societal factors, equity factors, and social environment all can exacerbate the impact of exposures. Finally, Miller mentioned that we have not translated science into policy; one example is mixtures. Mixtures are more than the sum of their parts; whether they are carbon impacts, chemical exposures, or externalities related to climate change, we need to do a better job looking at them, said Miller.

Thomas Burke of Johns Hopkins University said he hopes this workshop will be a turning point where the discussions can help guide toward a new lens for children’s health. As a risk assessor, Burke said he was struck that risk assessment can cause long delays in actions to protect our children. Another message was that it is time to change the question of research and risk assessments from the quest for identifying permissible exposures to how to prevent exposure and protect health; it is time to move from single pollutants to cumulative impacts, including the social impacts. According to Burke, focusing on the whole system when addressing a problem would advance children’s environmental health. Burke said to Hanna-Attisha, “When we first met at Flint, that was a systems failure that let down those kids.”

Responding to a question about the main barriers to understanding the relationships between pesticide exposures and health, Eskenazi described that exposure assessment has been limiting our ability to look at the potential health effects. To assess exposure to modern pesticides, such as pyrethroids or organophosphates, a metabolite of that pesticide’s parent compound is measured in urine. That may cause measurement error, as that urinary metabolite may not reflect differences in exposure to the parent compound. The difference in metabolite levels might reflect differences in people’s metabolism or direct exposure to the metabolite; this is true for other chemicals. In addition, said Eskenazi, modern pesticides have half-lives of a few hours or days, not years. A short half-life does not preclude chronic effects from repeated exposure or even a single exposure. Benzene is a highly toxic compound associated with leukemia; a single exposure to a high dose or repeated exposure to lower doses can have long-term effects.

Eskenazi pointed out that assessing exposure in pregnancy is challenging. Pesticide exposure represents a snapshot in time, so fully capturing exposures and allowing an inference during critical windows would require 24-hour urine samples every day during pregnancy. This detailed level of exposure assessment is infeasible and cost prohibitive, as pesticide samples cost $100–150 per sample for just one chemical class.

Another point made by Eskenazi is that many pesticides are barely studied, partly because the methods to measure them as human biomarkers are new. Very little is known about human exposure to neonicotinoids, especially in children, because it was not possible to measure them in humans until recently. Glyphosate is the same; although it is the most commonly used herbicide worldwide, we know virtually nothing about its health effects on children and fetuses. Developing a reliable and valid assay takes time, and few laboratories are capable.

When asked what scientific actions she had for the EPA, Eskenazi replied, “We need to invest in the exposure assessment.… NIH funds

research looking at health effects, but actual exposure assessment, understanding the validity and the reliability of measures, and developing new measures have all been insufficient.”

Hanna-Attisha asked Miller how children’s unique vulnerabilities make it challenging to assess the nature of the risks. Miller listed several differences that cause children’s vulnerabilities: toxicokinetics, behavior, intake rates for fluids and food, and breathing rates. For intake rates, a 1-month-old is not the same as a 3-year-old, and an adolescent differs from both. As children grow and develop, their metabolism and excretion change.

Miller gave an example of exposure to nitrates. If nitrates are in drinking water, they can be absorbed into the blood, where a portion is metabolized into nitrite. That reacts with hemoglobin to create methemoglobin, reducing the amount of oxygen the blood can carry, causing methemoglobinemia. Methemoglobinemia is primarily a disease of infants because their water intake can be much higher; if a formula-fed infant is ingesting nitrate-contaminated water, that may be their single source of food. That is unusual because across different life stages, adults and even older children spend the day out of the home. Absorption of nitrates also differs for young infants because for the 1st month or 2, they have a relative overgrowth of bacteria in their gut that converts nitrates to nitrites, which are absorbed. Metabolism is also different because infants do not have methemoglobin reductase, a normal enzyme that converts methemoglobin back to hemoglobin. Toxicodynamic issues are different, as body systems are programmed during infancy in ways that can cause lifelong impacts.

Miller also pointed out the critical windows of sensitivity. Sometimes, those windows are short, or multiple windows could have different outcomes. One example would be fetal alcohol syndrome. Toxicity occurs throughout pregnancy, but outcomes are phenotypically different depending on the timing.

Adjustment factors are a common way to account for life stage vulnerabilities or other uncertainties in toxicity data. For example, a point of departure from a dose-response curve will be divided by an adjustment factor to determine a reference concentration or dose. Miller described that California uses different adjustment factors for different life stages: an adjustment of factor of 10 is used between the 3rd trimester of pregnancy and 3 years of age, and an adjustment factor 3 is used between the ages of 2 and 16. However, the California Office of Environmental Health and Hazard Assessment evaluated animal studies that specifically looked at the early life and adult periods within those studies. It observed a tenfold factor: in that early-life period, it varied from sometimes being less sensitive to a chemical to being a hundredfold or more sensitive.

Therefore, a factor of 10 is not necessarily fully accounting for childhood vulnerabilities.

Miller discussed the topic of vulnerability, particularly adolescence, which is a unique life stage. Adolescents have unique exposures and vulnerabilities. Their sexual organs are developing, which makes them more susceptible. California developed a report that evaluated the effects of secondhand smoke on health and found that for most children, the risk factor for developing breast cancer begins after menarche and goes through the first completed pregnancy (Miller et al., 2007).

Hanna-Attisha asked a follow-up: “Recognizing these vulnerabilities, what can EPA do better to protect children, especially during these critical windows?” Miller discussed the need for continued funding that is focused and long term. Much of the best research on children’s environmental health came from the children’s environmental health centers; children are unique and need that special attention, but there have not been as many new cohorts funded.

Hanna-Attisha asked Burke whether, in his role as a science advisor, he felt that science gaps held the agency back from protecting children. Burke noted that during his time at the EPA, it was working on lowering the ozone standard, addressing the lead and copper rule, and dealing with the national problem of lead in drinking water, especially in Flint. It also tried to use TSCA to address hazards posed by trichloroethylene exposure during fetal development and overcome the controversies from using data from animal toxicity studies to determine the critical endpoints. The EPA also fought a battle to push forward bans on chlorpyrifos, based on data from the children’s center cohort studies mentioned by Miller.

According to Burke, one of the biggest science gaps is exposure data. Exposures to children are still not well characterized. Collecting the right kind of samples is expensive, but the tools now exist. The air and water quality data are not granular enough to identify all areas of increased exposure. “We know we are missing vulnerable groups. We missed Flint, right?” According to Burke, the current exposure data does not allow the EPA to address environmental justice. No current exposure surveillance system can recognize other risk factors and high-risk communities. It would be helpful to have more basic epidemiology and exposure surveillance, such as the Centers for Disease Control and Prevention (CDC) environmental public health tracking network.

Burke added that the EPA is vulnerable to continual delay to policy making from a well-organized assault on science from consulting firms that are working on behalf of the regulated communities. The attacks on science lead to delays in regulatory decision making. As a result, so much of what is driving current environmental policies are studies from the 1970s and 1980s. The lack of new science and its policy translation is due to not a lack of a research community but a continual decline in funding.