Environews
FOCUS | Environmental Factors in Birth Defects: What We Need to Know
Birth defects are a major cause of infant deaths, but a lack of money, coordinated monitoring efforts, and high-quality data hampers efforts to fully understand why these conditions occur. Much remains unknown about the causes of birth defects, although experts believe environmental factors may contribute. Now researchers are looking at new ways to collect, analyze, and share data to help better understand environmental contributions to birth defects (p. A440).
SPHERES OF INFLUENCE | Learning Curve: Putting Healthy School Principles into Practice
With children spending about one-third of their day at school, so-called high-performance schools—designed to promote better attendance, achievement, and behavior by optimizing indoor air quality, ventilation, lighting, acoustics, and other elements of the built environment—could also provide children with the most pollution-free part of their day. But many children go to schools that are a far cry from this picture, and more than a few hurdles stand in the way to achieving high-performance schools. This article (p. A448) looks at ways some schools are overcoming those hurdles to create healthy learning environments.
Commentaries
POPULATION HEALTH | Infant Health and Malaria Control Insecticides
The World Health Organization Pesticide Evaluation Scheme (WHOPES) is used to assess the safety and efficacy of insecticides used for malaria control, including dichlorodiphenyltrichloroethane (DDT) and 12 other insecticides currently recommended for indoor residual spraying and bed nets. Bouwman and Kylin (p. 1477) note that WHOPES does not consider potential health effects of non-DDT pesticide exposures via breast milk, despite the potential for breast milk to be an important route of exposure for infants and children. The authors review exposures and health effects of these compounds in breast milk, and argue that this information should be incorporated in WHOPES evaluations. They suggest that additional research is needed to quantify exposures, including exposures through home and garden use, and clarify risks, including those risks associated with exposures to multiple insecticides. The authors conclude that risks are unlikely to outweigh benefits of malaria control, but note that this does not obviate the need to reduce risks whenever possible.
EXPOSURE SCIENCE | Sampling and Interpretation of Biomonitoring Data
Biomonitoring—the measurement of environmental chemicals or derivatives in biological specimens to estimate internal exposures—requires careful attention to factors that may affect accuracy and validity at any point from study design to laboratory analysis. Calafat and Needham (p. 1481) discuss the potential for sample contamination during collection and processing and emphasize the need to ensure that specimens analyzed in the laboratory represent both the biological sample of interest and the exposure of interest. The authors use examples based on prior research to support the routine use of field blanks (to identify contaminants introduced during collection, processing, or storage) and illustrate the potential influence of short-term exposures (e.g., to chemicals released from medical equipment), cross-matrix contamination (of one biological sample with another during collection), collection methods (e.g., extracting urine from diapers), and other factors that may affect biomonitoring for environmental exposures.
Reviews
EPIDEMIOLOGY | Chlorination and Congenital Anomalies
Water disinfection has substantial public health benefits, but also results in widespread exposure to chlorination disinfection by-products (DBPs) that have been inconsistently associated with congenital anomalies. Nieuwenhuijsen et al. (p. 1486) conducted a meta-analysis of existing studies to summarize current evidence and determine needs for future research. The authors included all published epidemiologic studies of DBP exposure (based on type of treatment, water source, or DBP measurements) and congenital anomalies, and derived summary estimates and dose–response estimates when appropriate. High versus low exposure to water chlorination or total trihalomethane concentrations was associated with all congenital anomalies combined [17%; 95% confidence interval (CI), 3–34] and with ventricular septal defects specifically (58%; 95% CI, 21–107), but the authors note that both estimates were based on small numbers of studies. They conclude that exposure classification and outcome ascertainment must be optimized and that additional DBPs and DBP mixtures should be evaluated in future research.
Also see Science Selections, p. A455
EXPOSURE SCIENCE | National Children’s Study: Environmental Exposures
The National Children’s Study (NCS), a longitudinal epidemiologic study to identify exposures and preventable causes of childhood disease, will enroll 100,000 children from 105 locations across the United States. Dust, water, and other samples will be collected to assess residential exposures, but additional data will be needed to fully characterize exposures over time. Lioy et al. (p. 1494) review information sources that will be used to determine exposures in Queens, New York (one of seven NCS pilot study locations), including existing national, regional, and local databases with information on air and water quality, waste sites, toxics emissions and spills, dietary intakes, pesticide use, geographic features, land cover and land use, meteorologic data, and other characteristics. These data will be used to develop and test a method for defining environmental exposure indices that may be applied across NCS sites. However, the authors note that data collection efforts at each NCS site will need to be customized to account for local characteristics and data sources, and that walk- or drive-through surveys will be necessary to fully characterize local variation within study sites.
EPIDEMIOLOGY | Occupational Pesticide Exposure and Childhood Leukemia
Childhood leukemias may develop from lymphoid or myeloid cells that acquire genetic abnormalities in utero. Known risk factors account for < 10% of cases, and exposure to ionizing radiation is the only established risk factor that is potentially modifiable. Parental pesticide exposure has been implicated based on experimental and epidemiologic research, and several common pesticides are considered probable or possible carcinogens. Wigle et al. (p. 1505) report on a systematic literature review and meta-analysis of childhood leukemia and parental occupational exposure to pesticides based on 26 case–control and 5 cohort studies. The analysis did not indicate an association with paternal exposures, but did support an association with maternal occupational exposures based on all studies combined and study subgroups classified according to quality scores, farm-related exposures, and other characteristics. The authors suggest that results should be interpreted cautiously, but conclude that efforts to limit maternal exposures may be warranted.
Research
EXPOSURE SCIENCE | Polycarbonate Bottle Use and Urinary BPA
Testicular germ cell tumors (TGCTs), the most common tumors diagnosed in men 15–34 years of age, are increasing in incidence. Etiologic mechanisms are uncertain, but evidence that sex hormones may influence pathogenesis has led to the hypothesis that environmental endocrine disruptors, including organochlorine (OC) compounds, may increase risk. Purdue et al. (p. 1514) conducted a nested case–control study within the Norwegian Janus serum bank cohort to estimate associations of 11 OC insecticides and 34 polychlorinated biphenyl (PCB) congeners (collected between 1972 and 1978) with incident TGCTs reported through 1999 to the Norwegian Cancer Registry (49 cases, 51 matched controls). Conditional logistic regression models were used to estimate associations between tertiles of lipid-adjusted serum levels and all TGCTs or cases histologically classified as seminomas (34 cases and matched controls). The authors report that p,p´-DDE, oxychlordane, trans-nonachlor and some PCB congeners were positively associated with the outcomes, but conclude that additional research is needed to corroborate their findings.
EXPOSURE SCIENCE | PBDEs in the Diet of the General U.S. Population
Polybrominated diphenyl ethers (PBDEs) are persistent environmental contaminants with evidence of endocrine, developmental, and other effects. North Americans are thought to be exposed primarily through contaminated dust and food sources, but information linking serum PBDE levels to diet is limited. Fraser et al. (p. 1520) analyzed data from a random sample of participants in the 2003–2004 National Health and Nutrition Examination Survey who had serum measures of PBDE congeners (including PBDEs 28, 47, 99, 100, and 153), [based on a 24-hr food recall (24FR) and a 1-year food frequency questionnaire (FFQ)], and information on potential confounders (n = 1,892 with complete data). The authors report that PBDE serum concentrations were 23–27% lower among vegetarians than omnivores. Poultry fat and red meat fat intakes were positively associated with serum PBDEs, but dairy or fish consumption was not. Results were similar for estimates based on 24FR and FFQ dietary instruments. The authors suggest that additional research is needed to determine mechanisms through which PBDEs enter the food chain.
Also see Science Selections, p. A455
EPIDEMIOLOGY | Lead and Gestational Hypertension
Blood lead level (BLL) is positively associated with hypertension, and previous epidemiologic studies have reported associations between late-pregnancy BLL and pregnancy-induced hypertension (PIH), an important cause of maternal and neonatal morbidity. Yazbeck et al. (p. 1526) studied PIH and maternal BLL measured at 24–28 weeks of gestation among 971 women in the EDEN (European Day Hospital Evaluation) cohort study. Blood pressure was measured during routine monthly visits, and women with elevated systolic or diastolic blood pressure at two or more visits were classified as having PIH (n = 106). The authors report that blood lead was positively associated with PIH, with a monotonic dose response from the lowest (< 1.2 µg/dL) to highest (> 2.3–8.5 µg/dL) quartile and an adjusted odds ratio of 3.3 (95% confidence interval, 1.1–9.7) for each 1-µg/dL increase in BLL. The authors conclude that lead levels well below the commonly accepted upper limit of 10 µg/dL may increase the risk of PIH.
EPIDEMIOLOGY | Bone Lead and Tooth Loss
Lead exposure has been associated with dental caries and periodontitis, but has not been evaluated as a risk factor for tooth loss. Arora et al. (p. 1531) analyzed data collected between 1992 and 1994 from 333 men enrolled in the U.S. Veterans Affairs Normative Aging Study to estimate associations between prevalent tooth loss and blood and bone lead levels. Tooth loss was increased among men with bone lead levels in the highest versus lowest tertile, even after adjusting for smoking, diabetes, age, and other potential confounders. Estimated associations were comparable in magnitude to those for current smoking versus never- or ex-smoking, and were similar between men with higher and lower levels of education. The authors note that the absence of an association between tooth loss and blood lead levels is consistent with an effect of long-term versus recent exposure that could continue to affect older individuals despite recent reductions in lead exposure.
ENVIRONMENTAL MEDICINE | Mass Lead Intoxication in Senegal
Haefliger et al. (p. 1535) describe an investigation prompted by the unexplained deaths of 18 children in the NGagne Diaw suburb of Dakar, Senegal. From 1995 to 2008 many NGagne Diaw residents were involved in lead-acid battery recycling, including manually recovering lead particles from contaminated soil. Clinical and laboratory assessments were performed on 32 siblings and 23 mothers of deceased children and on 18 unrelated children and 8 adults from the area, and lead contamination was measured in soil and homes. Mean blood lead levels were 129.5 µg/dL (range, 39.8–613.9 µg/dL) in children and 55.5 µg/dL (range, 32.5–98.8 µg/dL) in adults. Seventeen children had clinical signs of severe toxicity, and homes and soil were heavily contaminated with lead. The authors conclude that most if not all of the deaths resulted from severe lead intoxication, which may be widespread in the community despite recent decontamination attempts. These findings highlight the need to strengthen national and international efforts to address this global public health problem.
Also see Science Selections, p. A454
GENE POLYMORPHISM | Positive Selection at the Human CYP3A Locus
Cytochrome (CYP) P450 enzymes metabolize a wide variety of drugs, environmental contaminants, dietary constituents, and endogenous compounds. Substantial interindividual differences in CYP3A activity may partly reflect natural selection of functional polymorphic variants relevant to new environmental exposures encountered by early humans as they dispersed from Africa to other parts of the world. Previous studies have supported this hypothesis but have not completely profiled allelic structure across the CYP3A locus. Chen et al. (p. 1541) resequenced 54 genomic DNA fragments (43,675 bases) spanning four genes (CYP3A4, CYP3A5, CYP3A7, and CYP3A43), two pseudogenes (CYP3AP1 and CYP3AP2), and randomly selected intergenic regions in samples from 24 Africans, 24 Caucasians, and 29 Chinese. According to the authors, the inferred hyplotype structure indicates a recent or ongoing selective sweep (natural selection for specific haplotypes) in CYP3A4 and CYP3A7 in all three populations, and a recent selective sweep in CYP3A43 and CYP3A5 among non-Africans and Caucasians, respectively. Results suggest that evolutionary adaptation has contributed to phenotypic variation in CYP3A activity, and may partly explain differences in susceptibility to adverse effects of environmental exposures.
TOXICOLOGY | Perinatal Bisphenol A and Early Adipogenesis
Humans are exposed to bisphenol A (BPA) in utero, and experimental studies suggest effects of BPA on reproductive development, glucose homeostasis, and adipogenesis. Somm et al. (p. 1549) studied pups born to Sprague-Dawley rats given 1 mg/L BPA in drinking water (approximately 70 µg/kg/day) from gestation day 6 until postnatal day (PND) 21. The authors report that dams were not overtly affected by BPA exposure, and that litter sizes and sex ratios were comparable between exposed and unexposed litters. However, BPA-exposed male and female pups were heavier at birth than unexposed pups. At PND21, body weight remained elevated in exposed females but was comparable in exposed and unexposed males. White adipose tissue was increased with exposure in both sexes, along with adipocyte hypertrophy and increased expression of proadipogenic transcription factors and lipogenic enzyme genes; however brown adipose tissue was increased with BPA exposure in females only. At 14 weeks BPA-exposed females were heavier than unexposed females regardless of diet, but in males BPA exposure was associated with increased body weight only in rats on a high-fat diet. Body weight differences were not explained by increased food intake, suggesting effects on energy expenditure.
TOXICOLOGY | Selenium and Diabetes in NHANES 2003–2004
Previous studies have shown abnormal reproductive and developmental outcomes in sheep grazing on pastures treated with sewage sludge, which includes a complex mixture of endocrine-disrupting compounds (EDCs). Bellingham et al. (p. 1556) hypothesized that effects of EDCs on reproductive outcomes may result from effects on hypothalamo–pituitary–gonadal axis regulation by the kisspeptin (Kisspeptin-1) neuropeptide and its receptor, GPR54 (G-protein–coupled receptor 54). The authors quantified KiSS-1, GPR54, and estrogen receptor- (ERα) expression in maternal and fetal hypothalami and pituitary glands; assessed colocalization of Kisspeptin-1 with luteinizing hormone (LHβ) and ERα in fetal pituitary glands; and compared results in control versus treated animals. Exposure did not appear to influence measured outcomes in ewes, but fetal exposure was associated with reduced hypothalamic KiSS-1 mRNA expression and reduced colocalization of Kisspeptin-1 cells with LHβ and ERα in the pituitary gland. The authors conclude that in utero disruption of the Kisspeptin-1/GPR54 system may be an additional pathway through which EDCs cause long-term effects on reproduction.
TOXICOLOGY | Pyrethroids and Mixtures
Toxic effects of widely used pyrethroid insecticides are related to effects on neuronal excitability, but symptoms following acute high-dose exposures differ according to the chemical structure of specific compounds. Consequently, it is uncertain whether dose additivity is an appropriate default assumption for estimating risks associated with mixed pyrethroid exposures. Wolansky et al. (p. 1563) exposed adult male rats to a mixture of 11 pyrethroids that included equipotent ratios of each compound at individual doses below compound-specific threshold concentrations for behavioral effects. Compounds were simultaneously administered by oral gavage, or were administered in three separate mixtures according to the time to peak effect of individual compounds. The authors observed significant dose-related decreases in motor activity following exposure, despite the presence of each individual compound at subthreshold doses, and concluded that observed effects were consistent with predictions based on dose additivity, regardless of the dosing protocol used.
EXPOSURE SCIENCE | Model and Validation of Personal Exposure to VOCs
Direct measurement of personal exposure to volatile organic compounds (VOCs) provides accurate exposure estimates but is impractical for large-scale epidemiologic research. Delgado-Saborit et al. (p. 1571) describe models developed to estimate personal exposures based on microenvironment VOC concentrations (measured at participant homes and workplaces or in representative microenvironments at other common locations), and information on individual characteristics and activities. Models incorporating information on stratified microenvironment VOC concentrations, lifestyle factors, and individual-level activities explained the greatest amount of exposure variance use of solvents and exposure to road traffic and environmental tobacco smoke were important sources of exposure in this United Kingdom study population of 100 healthy adult nonsmokers. The authors note that models developed to estimate VOC exposures may require modification if applied to other populations.
Children's Health
EPIDEMIOLOGY | Postnatal Cadmium Exposure and Child Neurodevelopment
Epidemiologic studies conducted in the 1970s and 1980s reported evidence of adverse neurologic and developmental effects of cadmium, but results may have been confounded by other exposures (e.g., lead) or biased due to exposure misclassification. Cadmium has also been associated with hypertension in adults, but has not been evaluated as a risk factor for hypertension in children. Cao et al. (p. 1580) analyzed data from the Treatment of Lead-Exposed Children (TLC) trial, a randomized clinical trial of succimer treatment in 2-year-old children with elevated blood lead levels (20–44 µg/dL) who were subsequently followed to 7 years of age. Blood cadmium levels were measured at 2 years, and blood pressure and neuropsychological and behavioral test scores were measured at 2, 5, and 7 years of age. The average cadmium concentration in study children was 0.21 µg/L, comparable to expectations based on a nationally representative sample. The authors report that blood cadmium levels at 2 years of age were not associated with neurodevelopmental end points or blood pressure at any of the time points assessed, either before or after accounting for blood lead levels, treatment group, and other potential confounders.
EPIDEMIOLOGY | Phthalates and Preterm Birth in Mexico City
Previous studies have reported inconsistent associations between gestational age and phthalates, which are commonly used in personal care products, pharmaceuticals, and plastics manufacturing. Meeker et al. (p. 1587) conducted a nested case–control study within a Mexican birth cohort comparing third-trimester urinary phthalate metabolite concentrations between 30 women who delivered preterm (< 37 weeks gestation) and 30 controls. The authors report that geometric mean urinary concentrations of several phthalate metabolites were higher among cases than controls, and that differences were attenuated but still apparent after correcting for urine dilution. Cases were also more likely than controls to have urinary phthalate metabolite concentrations above median levels based on logistic regression models adjusted for potential confounders. The authors note that concentrations of some phthalate metabolites were elevated in the population compared with previous estimates for women in the United States, and suggest that additional research is needed to identify exposure sources and confirm findings in a larger study population.
EXPOSURE SCIENCE | Predictors of Serum Dioxins and PCBs among Russian Boys
Extensive chemical manufacturing in Chapaevsk, Russia, has resulted in high environmental contamination with dioxins, furans, and polychlorinated biphenyl compounds (PCBs). Burns et al. (p. 1593) analyzed baseline data from 482 Chapaevsk boys 8–9 years of age (participants in the Russian Children’s Study cohort) to identify predictors of serum dioxin and PCB levels at enrollment in 2003–2005. Boys who were older, consumed local foods, and were breast-fed longer, and boys whose mothers were employed at the Khimprom chemical manufacturing plant or had a local garden plot, had significantly higher serum dioxins and PCBs than other boys. Distance of the boys’ homes from the Khimprom plant was strongly associated with total toxic equivalents (TEQs) for all measured compounds combined, with an estimated adjusted mean TEQ of 30.6 pg/g lipid [95% confidence interval (CI), 26.8–35.0] for boys that lived < 2 km from the plant compared with 18.8 pg/g lipid (95% CI, 17.2–20.6) for boys living > 5 km away. The authors recommend soil remediation and garden plot relocations to reduce exposures in this community.
EPIDEMIOLOGY | Hydroxylated PCBs and Child Neurodevelopment
Hydroxylated polychlorinated biphenyls (OH-PCBs) are water-soluble metabolites that may be actively transported across the placenta or formed by fetal metabolism. Experimental studies suggest effects on hormonal homeostasis and neurodevelopment, but health effects of OH-PCBs have not been studied in humans. Park et al. (p. 1600) estimated associations of six OH-PCBs measured in maternal (n = 147) and cord (n = 80) serum samples with 16-month Bayley Scales of Infant Development Mental Development Index (MDI) and Psychomotor Development Index (PDI) scores. Study participants were a stratified random sample of cohort study participants from two eastern Slovakia sites, including an area with high environmental PCB contamination from a manufacturing facility. Cord and maternal serum levels of 4-OH-CB-107 were significantly associated with lower MDI scores, and cord 4-OH-CB-107 was also associated with lower PDI scores, but other OH-PCB metabolites were not associated with decreased test scores. The authors note that inverse associations were previously observed between Bayley Scores and levels of the parent PCBs from which 4-OH-CB-107 is derived, but conclude that results require confirmation in a larger study population.
EPIDEMIOLOGY | Lead Exposure and Behavior in India
Blood lead levels have decreased in Indian children since the phaseout of leaded gasoline in 2001, but levels continue to exceed acceptable limits in many areas. Effects of lead on IQ (intelligence quotient) scores are well established, but few studies have assessed effects on specific domains of behavior and cognition. Roy et al. (p. 1607) conducted a cross-sectional study of 756 children 3–7 years of age in Chennai, India, to evaluate relations between blood lead levels and anxiety, social problems, inattention, hyperactivity, attention-deficit/hyperactivity disorder (ADHD), and executive function, which were assessed based on the Conners’ Teacher Rating Scales-39, the Conners’ ADHD/Diagnostic and Statistical Manual for Mental Disorders, 4th Edition Scales (CADS), and the Behavior Rating Inventory of Executive Function. The mean blood lead level in the study population was 11.4 µg/dL (range, 2.6–40.5 µg/dL), with more than half of the children having levels > 10 µg/dL. The authors report that blood lead level was associated with higher anxiety, social problems, ADHD, and deficits in executive function, and conclude that executive function and attention are particularly vulnerable to adverse neurobehavioral effects of lead.
EPIDEMIOLOGY | Traffic-Related Air Pollution and Neurobehavioral Function
Animal studies suggest that traffic-related air pollution may have adverse neurologic effects, but studies of neurobehavioral effects in children are limited. Wang et al. (p. 1612) conducted a cross-sectional study of children from two primary schools in Quanzhou, China to estimate associations between chronic exposure to ambient air pollution and neurobehavioral test scores. Participants were 8–10 years of age, including 431 children from a school in an area with low traffic density and 430 from a school in a high traffic area (median campus nitrogen dioxide concentrations of 7 µg/m3 and 36 µg/m3, respectively). The authors report that neurobehavioral test scores for children from the school with higher traffic-related air pollution exposure indicated poorer performance than scores for children with lower exposure, and conclude that additional studies are needed to clarify effects of traffic-related air pollution on neurobehavioral function and development.
EXPOSURE SCIENCE | Herbicides and Adverse Pregnancy Outcomes
Atrazine and other corn herbicides are routinely detected in drinking water. Two studies on atrazine’s potential association with small-for-gestational-age (SGA) and preterm birth prevalence have found inconsistent results. Moreover, these studies did not control for individual-level potential confounders. Ochoa-Acuña et al. (p. 1619) used a retrospective cohort study to evaluate whether atrazine in drinking water is associated with increased prevalence of SGA and preterm birth. They found that atrazine in drinking water during the third trimester, as well as during the entire pregnancy, was associated with a significant increase in the prevalence of SGA. Levels of atrazine > 0.1 µL in drinking water during the third trimester were associated with a 17–19% increase in the prevalence of SGA. Mean atrazine concentrations over the entire pregnancy > 0.644 µg/L also were associated with higher SGA prevalence. No association was found for preterm delivery.
EXPOSURE SCIENCE | Chemical Concentrations during Lactation
It has generally been assumed that concentrations of persistent organic chemicals in breast milk decline over the course of lactation and that initial levels reflect the mother’s cumulative lifetime exposure. LaKind et al. (p. 1625) studied changes in persistent organic chemicals in breast milk samples collected 1, 2, and 3 months postpartum and at the end of lactation from a convenience sample of 10 women from Pennsylvania. The same chemicals also were measured in serum samples collected at the first and last milk donation to assess milk:serum partition coefficients. Lipid-adjusted concentrations of polybrominated diphenyl ethers, polychlorinated biphenyls, polychlorinated dibenzo-p-dioxins and furans, and organochlorine pesticides in serum and milk did not consistently decrease during lactation; milk: serum partitioning was unpredictable and may have been dependent on chemical characteristics. The authors conclude that decreasing concentrations of lipophilic chemicals during lactation should no longer be assumed.
EPIDEMIOLOGY | Variation in PON1 Enzyme Activity in Young Children
Activity of the paraoxonase 1 (PON1) enzyme, which detoxifies activated organophosphorus pesticides (OPs) and reduces oxidative stress, has previously been shown to vary according to genotype and age, with lower levels of PON1 activity in children than in adults. However, information on age-related changes in PON1 activity is limited. Huen et al. (p. 1632) measured PON1 genotypes (polymorphisms at positions 192 and –108) and PON1 activity from birth through 7 years of age in 458 children (108 with data from four or more time points) in the CHAMACOS study, a cohort of predominantly Mexican-American women and children with relatively high occupational and environmental OP exposures. The authors report that relatively low levels of PON1 activity persisted in young children through at least 7 years of age, suggesting a larger window of vulnerability to adverse effects of OPs than expected based on previous research. In addition, they noted that children with low-activity genotypes may be particularly vulnerable to OPs compared with other children because of lower PON1 activity overall and a slower increase in activity with age.
Also see Science Selections, p. A454
