3

Environmental Influences on Children’s Health Across the Life-Span and Generations

The session included a series of presentations from experts on the influences of the environment on children’s health and development. Paul Juarez, Meharry Medical College, discussed the public health exposome and health equity. Frederica Perera, Columbia Center for Children’s Environmental Health, discussed biomarkers to air pollution and children’s health and development. Niels Erik Skakkebæk, Rigshospitalet and the University of Copenhagen, described testicular cancer as a late symptom of testicular dysgenesis syndrome. Kelly Ferguson, National Institute of Environmental Health Sciences, described the importance of disentangling heterogeneous health outcomes to improve research evaluating chemical exposures during pregnancy. Manish Arora, Icahn School of Medicine at Mount Sinai, presented on prenatal exposures as markers for autism risk. The session was moderated by Germaine M. Buck Louis, former dean of the George Mason College of Public Health, who chaired the planning committee.

THE EXPOSOME AND HEALTH EQUITY¹

Juarez described what he and his colleagues have learned by developing a conceptual model, the public health exposome, and an accompanying database. A major focus of their effort with the exposome is to characterize causes of health disparities that can be addressed to achieve health equity

___________________

¹ This section is based on a presentation by Paul Juarez, professor, Meharry Medical College.

(Juarez et al., 2014). A motivation behind this project was that “zip code is a greater predictor of children’s health outcomes or personal health outcomes than genetic code” and that some estimate that 70–90 percent of disease risks are due to environmental and behavioral factors (NASEM, 2016; Rappaport and Smith, 2010). The database has metadata (a data dictionary) that describes the data in depth and standardizes how they are captured, reported, and useful for analyses. The database is now shareable through data use agreements.

Juarez also described their ontologies. The exposome measures all a person’s exposures throughout their lifetime and how those exposures relate to health. Exposures begin before birth and include environmental and occupational sources. It is important to understand how environmental exposures interact with each person’s unique characteristics, such as genetics, physiology, and epigenetics, to affect health (Juarez et al., 2014).

The environment includes both internal and external environments. The latter is divided into five domains: natural, built, social, policy, and health care. The natural environment is air, water, and land; the built environment is the places people live, work, play, pray, and learn; the social environment is the social and demographic characteristics and

political and economic factors; the policy environment is composed of federal, state, and local policies and regulations that directly or indirectly impact health; and the health care environment includes providers and population-level measures, such as medically underserved areas and areas of health care profession shortages.

From a population perspective, health disparities are considered a measure of the inequities between population groups. Social demographic characteristics, such as age, race, and gender, and social measures, such as poverty status, socioeconomic status, and place, measure these. Measures of population health are measured against a comparison group, typically White people. Relative measures of risk, such as odds ratios, are then calculated to compare morbidity and mortality across racial and ethnic groups. Several different outcomes can be used, such as incidence, prevalence, mortality, life expectancy, years of personal life loss, or disability-adjusted life years.

Health equity is achieved when every person can “attain his or her full health potential,” and no one is “disadvantaged from achieving this potential because of social position or other socially determined circumstances,” described Juarez. Public health aims to achieve health equity by eliminating health disparities and achieving optimal health for all Americans. According to Juarez, doing so will require that public health professionals work together to eliminate unjust differences in health and health care using an antiracist approach.

The public health exposome is vital to improving children’s health. The leading causes of child and adolescent mortality are motor vehicle crashes, firearm-related injury, malignant neoplasms, suicide, drug overdose, drowning, congenital anomalies, heart disease, fire or burns, and chronic lower respiratory disease. External environmental and behavioral factors are drivers of these causes of child and adolescent mortality.

Exposures are important factors in understanding where public health can intervene and improve health in populations. Measuring exposures includes identifying the source (such as chemicals), the natural and built environments, and nonchemical exposures (such as the social and policy environments and socioeconomic conditions or psychosocial stressors). Exposure routes (inhalation, ingestion, and dermal) are also important to consider, as well as the characteristics of exposures, duration, concentration, and frequency. Exposure assessments require aligning the external environmental exposures with internal biomarkers of exposure effect and disease, both spatially and temporally.

Longitudinally aligning exposures to biological effects is challenging. Some exposures are instant but may have cumulative effects or interactions. Those are the challenges that Juarez and colleagues are trying to address with the public health exposome data.

Recently, Juarez and colleagues linked the public health exposome data to the Southern Community Cohort Study. The Southern Community Cohort Study is a cohort of 85,000 adults recruited through community health centers in 12 southeastern states beginning in 2005. It aims to evaluate risk factors for cancer and other diseases.² Daily measures of particulate matter (at a 3 km grid cell level over 12 years) were connected to the addresses of the patients in the study. They also have included heat metrics, such as minimum and maximum temperature, heat index, and precipitation measures, for the 12 states, and environmental exposures to stress. Chemicals, socioeconomic factors, and psychosocial stress (from racism, food insecurity, violence, poverty, and housing insecurity) affect health through stress and inflammation mechanisms.

Juarez offered thoughts for future directions. First, he suggested consensus on a standardized ontology for assessing the external environment and spatial and temporal measures of external and internal exposures. The Systematized Nomenclature of Medicine (SNOMED) has helped standardize health care ontology, but work is needed on ontologies for the public health exposome. The ontology he and his colleagues have developed may work for others, but more input is needed. Next, Juarez suggested aligning external exposures and internal biomarkers of exposure, effect, and disease. It is important to look at multiple or cumulative exposures with interactive effects over the life course, which needs development on a conceptual or theoretical level, to test those theories. Finally, Juarez called for a human “exposome” project, like the Human Genome Project, that allows investigators from different fields to come together and standardize these types of measures.

BIOMARKERS OF PRENATAL EXPOSURE TO AIR POLLUTION AND CHILDREN’S HEALTH AND DEVELOPMENT³

Over 80 percent of toxic air pollutants and CO₂, a greenhouse gas, are emitted by the combustion of fossil fuels. Globally, the climate is getting worse, and in many places, air quality is deteriorating as well, said Dr. Frederica Perera, founder of the Columbia Center for Children’s Environmental Health. Some estimate that up to 90 percent of the global population is exposed to air pollution above the World Health Organization (WHO) guidelines for health protection.⁴ Additionally, at the current pace

___________________

²https://www.southerncommunitystudy.org/ (accessed December 20, 2022).

³ This section is based on a presentation by Frederica Perera, director, Columbia Center for Children’s Environmental Health.

⁴https://www.who.int/news/item/04-04-2022-billions-of-people-still-breathe-unhealthy-air-new-who-data (accessed December 21, 2022).

of climate change, within the next 2 decades, the Earth is likely to breach the threshold set by the Intergovernmental Panel on Climate Change.⁵

Air pollution and climate change have combined effects and synergies with respect to exposure (Perera, 2017; Perera and Nadeau, 2022). Air pollution includes the criteria pollutants that the EPA can regulate under the National Ambient Air Quality Standards, particulate matter (PM_2.5, PM₁₀), ground-level ozone (O₃), carbon monoxide, lead and nitrogen dioxide (NO₂), and air toxics and hazardous air pollutants like benzene, formaldehyde, and ethylene oxide that are regulated from industrial facilities. Fossil fuel use leads to polycyclic aromatic hydrocarbons (PAHs; a component of the fine particulate fraction [PM_2.5]), NO₂, black carbon, mercury, sulfur dioxide, and ozone precursors (NO_X, VOCs) in the air. Climate change contributes to temperature changes, precipitation extremes (rain bombs, droughts), weather events, forest fires, and plant allergen formation. Ozone is an example of the synergistic effects between air toxics and climate change; the reactions that generate ozone occur faster at higher temperatures. Forest fires, made more frequent and severe by climate change, emit massive quantities of air pollution that add to those from fossil fuels (Perera, 2017; Perera and Nadeau, 2022).

Air toxics and climate change can lead to cumulative health effects on children (Perera, 2017; Perera and Nadeau, 2022). Air toxics are associated with infant mortality, preterm birth, low birth weight, new and exacerbated cases of asthma, decreased lung function, immune disorders, neurodevelopmental effects (intelligence loss, autism spectrum disorder, and attention deficit and hyperactivity disorder), mental health disorders, and alterations in the brain structure. Climate change is associated with preterm birth, low birthweight, allergy and asthma exacerbation, neurodevelopmental effects, heat-related illness, malnutrition, stunted growth, infectious diseases, forced migration, physical trauma, post-traumatic stress disorder, chronic stress, and mental health disorders. Evidence is accumulating that these two fossil fuel–derived threats can interact synergistically to heighten disease risks among children.

Fetuses are highly susceptible to exposure from air toxics and climate change; this is a period of complex and rapid development of the brain, and other systems are undergoing precise programming that can be vulnerable to disruption by a toxic chemical or stressor, says Perera. Further, the biological defense mechanisms to detoxify chemicals, repair damage to DNA, and mount an immune response are immature. The same factors

___________________

⁵ This threshold is a goal of 1.5 °C increase over preindustrial levels over by the end of the century within the next 2 decades (see: https://www.ipcc.ch/sr15/ (accessed November 10, 2022).

make the fetus vulnerable to maternal stress. As a result, chronic diseases can start from biological changes that begin in utero. There is also evidence that exposures before birth can impact later generations through epigenetic changes.

Physical and psychosocial toxic exposures during the fetal period impact brain development. Figure 3-1 illustrates that most of the brain’s architecture is established during the first 9 months. However, the brain is still developing throughout childhood and adolescence; thus, these are also vulnerable periods (Giedd, 1999).

Environmental and climate injustice has been increasing and takes many forms. “First, the young bear the brunt, yet they are not responsible for the problem,” said Perera. Children in marginalized communities face disproportionately higher exposure to toxic air pollution and climate change, such as children in developing countries and communities of color and low-income communities in developed countries, such as the United States. They are more exposed to air pollution because of where stationary sources of pollution and major highways are cited. Moreover, the risks of heatwaves, severe storms, and floods are higher in marginalized communities due to discriminatory land use and housing policies. It is also well known that stress from poverty and racism takes a biological toll and can interact with environmental toxicants (Payne-Sturges et al., 2019).

Biomonitoring is often conducted in epidemiologic studies to understand the connection between environmental exposures and clinical disease. Biological markers in samples of blood, placental tissue, and others can measure or estimate internal dose, biologically effective dose, preclinical effect, and susceptibility. The concept underlying molecular

epidemiology is that biological markers can be used to provide an early warning of risk and to estimate inter-individual variability (see Figure 3-2).

Many biomarkers have been evaluated for prenatal exposure to air pollutants, such as PAHs, PM, and traffic-related air pollutants. In some cases, biomarkers have been associated with an increased risk for an adverse outcome. For example, shortened telomere length in cord blood or placental tissue has been measured following exposures to PAHs and traffic-related air pollution. Telomeres play a role in fetal developmental programming, and shorter telomeres have been associated with worse health outcomes (Bijnens et al., 2015). However, the linkages from exposure to preclinical effects to health outcomes have not been fully established. Several indicators of oxidative damage have been associated with air pollution exposure but have not yet been associated with a health outcome: 3-nitrotyrosine, 15-F2t-isoprostane (cord blood), and mitochondrial 8-hydroxy-2’-deoxyguanosine (8-OHdG) levels (maternal blood). These are indicators of damage (Ambroz et al., 2016; Grevendonk et al., 2016; Saenen et al., 2016). Confirmed links between exposure, a biomarker of preclinical effect, and a health outcome include:

• Decreased mitochondrial DNA, a potential marker of oxidative stress production, has been associated with exposure to NO₂, PM_2.5, and PM₁₀; and also to lower birth weight (Clemente et al., 2015; Janssen et al., 2012).
• PAH-DNA adducts, a marker of potential DNA damage from PAH exposure, have been associated with adverse developmental outcomes, such as decreased birth weight, length, and head circumference, reduced IQ scores, attention problems, delayed maturation

of emotional self-regulatory capacity and autistic traits (Margolis et al., 2016; Perera et al., 2005; Tang et al., 2006; Vishnevetsky et al., 2015).

The PAH-DNA adducts are formed because PAHs can be metabolized to reactive intermediates that form covalent bonds with guanine nucleotides in DNA. Perera has used the biomarkers as an indicator of exposure and biologically effective dose of PAHs that incorporates both exposure and biologic response. The Columbia Center for Children’s Environmental Health cohort studies have been based in New York City, specifically northern Manhattan and the South Bronx; Krakow, Poland; and Chongqing, China.

The New York City cohort enrolled 727 Latina and Black pregnant women and is following their children through adolescence. The oldest children are 23 years old. The observed effects of prenatal exposure to PAHs (measured by personal air samples or PAH-DNA adducts) include behavioral problems, anxiety and depression symptoms, ADHD, decreased emotional regulation, autistic traits, reduced birth weight, and reduced head circumference and IQ. Interaction of PAH and maternal hardship due to poverty on IQ and ADHD, as well as between PAH and early-life stress on attention problems and ADHD (Margolis et al., 2016; Perera et al., 1998, 2012, 2011; Vishnevetsky et al., 2015). These associations are all significant after adjusting for potential confounders, Perera clarified.

Associations observed with air pollution are concerning because research on child development shows that outcomes often persist into or set the stage for sequelae in adulthood. Impairment of cognitive functioning, learning, and mental health are often lifelong conditions, and they can be triggered by climate change impacts. Adversity and stress due to poverty and racism compound the harm in terms of cognitive and behavioral problems, according to many studies, and all of this adds up to decreased resilience and less ability to contribute to society. “We certainly need all the brainpower and energy we have in the population to help us solve these problems that we’re facing,” said Perera.

Studies on biomarkers can support policy interventions. Perera cited examples from her center’s work in New York City, Krakow, Poland, and Chongqing, China. In New York City, Columbia Children’s Environmental Health Center data supported citywide policies to reduce air pollution. Those regulations included clean taxis, cleanup of diesel buses and trucks, and regulations to promote clean heat in buildings. In 2009, Michael Bloomberg wrote to Perera to tell her that her center’s findings had encouraged actions and commitment to reducing air pollution in the city.

Perera then explained the center’s research had demonstrated a reduction in exposures due to policy changes. Across three successive

NYC cohorts (1998–2022), in which they monitored prenatal air pollution exposure using small backpack samplers, they observed a significant downward trend in perinatal exposure to airborne PAHs. In addition, the detectable levels of PAH-DNA adducts in cord blood followed the same downward trajectory.

In Krakow, a collaboration between colleagues at the Jagellonian University and the Columbia Children’s Environmental Health Center has shown the benefits of environmental policies. Following a ban on burning solid fuels in coal stoves, citywide PM_2.5 declined by 50 percent during the heating season and about 40 percent overall in annual average, stated Perera.

In Chongqing, after closing a coal-burning plant in 2002, ambient air levels of PAHs significantly reduced in 2005; the levels of PAH-DNA adducts in newborn cord blood were also significantly lower in the cohort born in 2005. Closing the coal power plants also led to observable cognitive benefits in children. The mean level of brain-derived neurotrophic factor (BDNF), a key molecule involved in learning and memory, was significantly higher after the closure (1,267 pg/ml versus 753 pg/ml, p < 0.05). In addition, an association between PAH-DNA adducts and reduced scores in developmental tests at age 2 was no longer observed (Perera et al., 2008; Tang et al., 2006, 2008, 2014).

Perera believes more studies using biomarkers in assessing effects are needed to explore life course and transgenerational effects of air pollution, and to examine the cumulative effects of multiple exposures, interactions between air pollution and climate change, and interactions of air pollution and psychosocial stress. Echoing comments by Landrigan, Burke, Woodruff, and others, Perera commented that environmental health professionals need more research that shows the benefits of regulation and other public health interventions and is more focused on pregnant women and children in environmental justice communities.

TESTICULAR CANCER AS A LATE SYMPTOM OF TESTICULAR DYSGENESIS SYNDROME⁶

Industrialized countries have had birth rates below sustainability levels for several decades and are now facing declining populations, said Skakkebæk. Many couples are experiencing infertility, with about 10 percent of all children in Denmark born after assisted reproduction, including donor and partner inseminations (Skakkebæk et al., 2022). Semen quality is generally poor. Today, more than 90 percent of abnormal sperm cells are paradoxically considered “within normal range” by WHO standards.

___________________

⁶ This section is based on a presentation by Niels Erik Skakkebæk, senior researcher, Rigshospitalet and University of Copenhagen.

Furthermore, serum testosterone levels have been falling for three or more decades. Frequencies of genital abnormalities, including cryptorchidism (undescended testicles), are increasing (Skakkebæk et al., 2016). Finally, it is well documented that there has been a worldwide increase in testicular germ cell cancer for decades. Testicular germ cell cancer is linked to developmental disorders of the testis, including hypospadias, cryptorchidism, and poor sperm production in testicular dysgenesis syndrome (Skakkebæk et al., 2022).

Skakkebæk said that his original observation of precursor cells for testicular cancer occurred in 1972 when he detected carcinoma-in-situ of the testis in two infertile men and published it in the Lancet (Skakkebæk, 1972). In 2016, WHO’s International Agency for Research on Cancer accepted the term germ cell neoplasia in-situ for this precursor lesion (Moch et al., 2016). Interestingly, the neoplasia in-situ germ cells, which are present in young adult men before the development of invasive cancer, express fetal markers for normal embryonic primordial cells. Thus, laboratory research aligns with epidemiological studies pointing to the fetal origin of adult germ cell tumors. The good news, according to Skakkebæk, is that methods have been developed to detect germ cell neoplasia in-situ in biopsies of men at risk of developing testicular cancer (e.g., in the contralateral testis of men with unilateral tumors). Germ cell neoplasia in-situ cells could be detected before tumor development. In that case, the patient can be treated with localized radiation to the affected testis, orchiectomy can be avoided, and testosterone production in that testis can be preserved (Rajpert-De Meyts et al., 2022).

Normally, the gonocytes in the fetal testes differentiate into prespermatogonia late in fetal life and become spermatogonia at birth. During and after puberty, the germ cells differentiate into spermatocytes, spermatids, and sperms. According to the current model for the pathogenesis of testicular germ cell tumors, an arrest of cell differentiation at the fetal gonocyte stage due to poor Leydig or Sertoli cell functions will lead to the proliferation of gonocytes into germ cell neoplasia in-situ (see Figure 3-3).

Skakkebæk then segued to the epidemiologic and histologic evidence that testicular cancer is associated with decreasing male fertility. A case-control study showed that men with testicular cancer had fewer children than men without testicular cancer 2 years before the tumor was detected (Møller and Skakkebæk, 1999). There has also been histological evidence of poor spermatogenesis among men with testicular cancer. A pathology study of 218 unilateral testicular cancer cases found 25.2 percent of patients had a cumulative presence of one or more signs of testicular dygenesis syndrome, and more than 70 percent of patients had Leydig cell hyperplasia (Hoei-Hansen et al., 2003).

Testicular germ cell cancer may be just one symptom of testicular dysgenesis syndrome due to dysfunction of fetal Sertoli and Leydig cells, said Skakkebækk. The dysgenesis of the Sertoli and Leydig cells and impaired germ cell differentiation may also lead to reduced semen quality and androgen insufficiency (see Figure 3-4). Lower testosterone

production can cause hypospadias and cryptorchidism. In severe cases, even disorders of sex differentiation may occur (Skakkebæk et al., 2021, 2022; Toppari et al., 1996; van den Driesche et al., 2017).

Although there are genetic causes of testicular cancer, they cannot explain the increasing rates across the globe (see Figure 3-5), which are most likely due to environmental factors causing damage to the fetal testis (Skakkebæk et al., 2022).

Skakkebækk concluded that the testicular germ cell cancer rate, which is linked to the maldevelopment of fetal gonads and male infertility, is a

warning sign that male reproductive health is at risk. “The role of industrial exposures in the current crisis in male reproduction, including solid data on worldwide increasing rates in testicular germ cell cancer, urgently needs to be explored,” said Skakkebækk.

DISENTANGLING HETEROGENEOUS OUTCOMES TO IMPROVE UNDERSTANDING OF ENVIRONMENTAL IMPACTS ON PREGNANCY AND CHILD HEALTH⁷

Ferguson discussed two major priorities: disaggregating heterogeneous health outcomes and conducting new research to evaluate new exposures. Development is a critical period of susceptibility to environmental impacts. Birth outcomes associated with later life disease include preterm birth and low birth weight, which both capture heterogeneous disease etiologies. Environmental toxicities are typically very specific, which dilutes the association with these heterogeneous outcomes. Therefore, disaggregating outcomes could potentially improve the ability to identify associations with environmental toxicant exposures and understand the mechanisms of action. In addition, new research is needed to address new exposures because chemical exposure profiles are changing dramatically. As chemicals were discontinued because of identified toxicities, new chemicals replaced them, so historic cohorts will not elucidate anything about the chemical exposures children are experiencing today.

Associations between phthalate exposure and preterm birth are mixed across the literature, said Ferguson. Phthalate exposure was associated with spontaneous preterm births in a pooled study of 16 cohorts (Welch et al., 2022) and linked with changes in fetal to early childhood growth (Ferguson et al., 2022). However, within each study, there is a small number of preterm births, which limits the power to observe an effect. In a case-control study, Ferguson and colleagues observed associations between prenatal phthalate exposures [di(2-ethylhexyl) phthalate (DEHP) metabolites and monobutyl phthalate (MBP)] and preterm birth (DEHP OR: 1.33, 95 percent CI: 1.04–1.70; MBP OR: 1.27, 95 percent CI: 0.99–1.63) (Ferguson et al., 2014). They recognized that preterm birth is heterogeneous and can be spontaneous or placental, so the study collected specific phenotypes on the type (McElrath et al., 2008). They observed larger effect estimates when they disaggregated the impact of phthalate exposure by subtype of preterm birth. For example, the odds of spontaneous birth were five times higher among the highest compared to the lowest quartile of exposure for mono(2-ethyl-5-carboxypentyl) (MECPP;

___________________

⁷ This section is based on a presentation by Kelly Ferguson, investigator, National Institute of Environmental Health Sciences.

a metabolite of DEHP). On the other hand, associations with placental preterm birth were null (Ferguson et al., 2014).

Having a large number of cases can improve statistical power and allow for discovery of effects despite heterogenety of outcomes, described Ferguson. Recently, data from all U.S. studies with prenatal measures of one or more urinary phthalate metabolites (16 cohorts) were pooled for analysis (n = 6,043 cases and n = 538 preterm births) (Welch et al., 2022). The study found that phthalate metabolites mono-isobutyl phthalate, MBP, and mono (3-carboxypropyl) phthalate were associated with increased odds of preterm birth in adjusted models. Ferguson summarized that exposure to several phthalates is associated with preterm birth, but the spontaneous preterm birth associations were greater in magnitude. Combining data across multiple studies with this level of clinical detail is rare; studies require sufficient sample size to overcome this heterogeneity (Welch et al., 2022).

Ferguson is also working to tease out the heterogeneous outcome of child growth and determine if evaluating specific child growth phenotypes allows for elucidating phthalate–child growth mechanisms. People with the lowest index of birthweight-to-length ratio and highest body mass index at age 11 are at the highest risk of death from coronary heart disease in adulthood (Eriksson et al., 1999). Few environmental epidemiology studies examine associations with a longitudinal trajectory of catchup growth (a period of rapid growth that follows a period of reduced growth), Ferguson said. In one study of 780 participants with measures of weight from in utero to age 6 and adiposity beginning at birth, MEP and MBP metabolites were associated with decreased adiposity at birth but increased adiposity in childhood (Ferguson et al., 2022). Exposure to several phthalates may be associated with decreased adiposity at birth and higher adiposity in childhood. It is challenging to capture complete longitudinal data from pregnancy through childhood on a large sample size to identify the impacts of phthalate exposure in these various growth trajectories.

Ferguson’s final points reflected the need to evaluate chemicals being used to replace those phased out for toxic effects. As phthalates were discontinued, she saw decreasing exposures to them but increasing exposures to terephthalate and Di(isononyl) cyclohexane-1,2-dicarboxylate (DINCH) metabolites from 2007 to 2018. The National Health and Nutrition Examination Survey had similar findings. The challenge, according to Ferguson, is that we need to study pregnant people with these exposures who do not exist in the older cohorts.

Ferguson suggested two research priorities. The first would be capturing detailed, standardized information on pregnancy and child health outcomes to facilitate the subtyping of disease for improved detection

of toxicity and understanding of mechanisms. Detailed child health outcomes could be helpful in studies of executive function. Exposure to replacement chemicals will also be important, as exposure patterns change over time, such as for pesticides, flame retardants, and PFAS; new recruitment into cohorts is needed to understand the consequences of these new exposures.

PRENATAL EXPOSURES AND BIOMARKERS AS MARKERS FOR AUTISM RISK⁸

“What role does the environment play in precision medicine?” asked Dr. Manish Arora of Mount Sinai School of Medicine. Precision medicine is often focused on genetics, but studies of twins have shown that for a whole range of diseases, heritable factors are less important than nonheritable factors (Rappaport, 2016). For precision environmental medicine, both genetics and environment are needed because it is clear they both play a role in many different outcomes.

Moreover, said Arora, the technologies for measuring environmental chemicals and responses in the human tissues (e.g., exposomics and metabolomics) have been improving vastly every 5 years or so. Measuring tens of thousands of chemicals in the tiniest drop of blood is now possible. Genetic profiling has also improved, but its predictions for human health have not. Precision medicine has a missing piece.

One aspect that differentiates the environment from the genome is that the environment is highly dynamic, at not just one scale but many. It is more dynamic than other biological processes, such as those connected to circadian rhythm. Some environmental factors operate at millisecond scales; others have seasonal, annual, or even longer cycles. Thus, the environment is also dynamic at different scales of time.

Arora said he and his colleagues felt that fundamentally, the gene-by-environment paradigm (health = genes × environment) is incomplete and that “complex systems cannot interact directly or exist in isolation.” The fundamental aspect of the biodynamic interface is time (Arora et al., 2021). Without considering the temporal nature of physiology and environment, how genes and environment together affect health will never be fully understood. For example, consider a study where the exposure intensity is higher in the controls than in the cases, such as with an essential nutrient or a dietary supplement. The next study could show an inverse relationship. The differences between these examples could be the timing of the exposure, Arora explained.

___________________

⁸ This section is based on a presentation by Manish Arora, professor, Icahn School of Medicine at Mount Sinai.

Throughout the workshop, presenters discussed the importance of the prenatal period. Arora has used exposures measured in teeth and hair to reconstruct exposures during the prenatal period. Teeth and hair can be used to reconstruct prenatal exposures because both matrices have incremental markings that mark time. Using baby teeth, which have rings like a tree to mark different developmental periods, weekly measures of lead exposure can be estimated going all the way back to the prenatal period (Arora et al., 2021). If a blood sample was collected, only one snapshot of exposure could be measured, which may miss an exposure at an important time. Arora and colleagues have recently studied not just exposure to environmental chemicals but also a biological response to those stressors. They have observed that even twins have variations in response to the environment, as reflected by different patterns of C-reactive protein.

ASD is a neurodevelopmental condition characterized by many behavioral traits and repeated behavioral problems with socialization, such as language delays. The 1970s prevalence was 1 in 10,000; now some estimate it at 1 in 44, which is not wholly explained by an increase in diagnosis, explained Arora. The average age of diagnosis is about 4–5. Early therapies can help improve outcomes, but no test can detect when therapies would be more beneficial.

Arora and colleagues have developed a biomarker that can detect risk for autism at birth; it is under review by the Food and Drug Administration and has been given breakthrough status.⁹ They use laser ablation and mass spectrometer analysis of hair combined with machine learning. The biomarker has a sensitivity of 0.964 and a specificity of 0.754. The negative predictive value of this test is very good (if it is negative, the person almost certainly does not have autism). It is also important to differentiate ASD from similar conditions, such as ADHD, and compare that to control conditions that are dissimilar to autism.

The environment is highly dynamic. “Measuring more and more exposures is not enough to understand its impact on health. How exposures and physiology change over time also needs to be considered. Precision environmental medicine is an urgent need,” says Arora.

PANEL DISCUSSION: RETHINKING EXPOSURES TO INFORM DECISION MAKING

Buck Louis asked the panelists about suggestions the EPA might consider. Juarez responded that being able to disseminate results back to

___________________

⁹https://www.prnewswire.com/news-releases/linus-biotechnology-inc-receives-fda-breakthrough-device-designation-for-stranddx-asd-exposome-sequencing-diagnostic-301438415.html (accessed December 21, 2022).

the communities through trusted messengers is important and reiterated that a consensus ontology in environmental exposures and associated biomarkers is needed. The ontology can help advance computational machine learning techniques to look for those patterns in relationships of exposures and biomarkers.

Ferguson had a different perspective on potential EPA priorities. She believes that when the EPA conducts hazard assessments for chemicals, it does so one at a time due to policy requirements. It would be helpful to move toward evaluating classes of chemicals, or mixtures, which would shift away from the replacement problem. There may not always be information now to be able to conduct hazard reviews that account for mixtures. Still, the EPA may be able to standardize approaches for chemical mixture analyses to achieve more harmonization across that literature and be able to synthesize some of the results.

Perera said that she would like to see the treatment of chemicals as classes based on their underlying biological pathways of action. That would go a long way toward ending the chemical-by-chemical approach. Perera also said that support is needed for long-term cohort studies capable of assessing the effects of early-life exposures over an extended period, through adolescence and beyond.

Skakkebæk agreed with Perera. Puberty is important, but so is mini puberty, which occurs between birth and 6 months. In one of his studies, Skakkebæk found that the best predictor of male semen quality was testosterone levels at 3 months.

Arora said that an important goal over the next 25 years would be to make progress on climate change. If this crisis is not addressed, many other problems will become secondary.

This page intentionally left blank.