Children's Environmental Health Research Articles

Prenatal exposure to nitrate from household drinking water and the risk of preterm birth: A nationwide study from Denmark, 2011-2015

BACKGROUND AND AIM: Only two studies of prenatal exposure to nitrate from drinking water and the risk of preterm birth have been conducted, both with ecologic exposures and from the US. We aimed to examine this relationship utilizing individual-level exposure estimates in Denmark. METHODS: Nationwide register-based health data from singleton live births during 1991–2015 born to Danish-born parents were linked to national water monitoring and administrative data, including each maternal residence during pregnancy. The association between individual-level prenatal nitrate (NO₃⁻) exposure from household drinking water and the odds of preterm birth (37 weeks gestation) was evaluated using logistic regression with generalized estimating equations to account for sibling clusters. Models were adjusted for sex, birth year, birth order, urbanicity, and maternal age, smoking, education, income, and employment status at the individual level. RESULTS:Using a complete case analysis, 1,009,189 births were included in main models, including 3.6% of households with nitrate levels 25 mg/L NO₃⁻ (half the European Union standard) and 5.1% of births being preterm. The median exposure in the cohort was 1.9 (IQR: 1.0–3.4) mg/L NO₃⁻. An increased risk of preterm birth was seen across categories of exposure ({p}0.001) with an odds ratio in the uppermost category (25 mg/L NO₃⁻) of 1.05 (95% CI: 1.00, 1.10) compared to the referent (2 mg/L NO₃⁻). Continuous models also demonstrated an increased risk of preterm birth per 10 mg/L NO₃⁻ increase [OR: 1.01 (95% CI: 1.00, 1.03)]. CONCLUSIONS:We observed an increased risk of preterm birth as nitrate concentrations in maternal household drinking water increased. This estimated risk was of a similar magnitude as the two previous studies from the US. Our findings add to a body of evidence of the physical harm from nitrate in drinking water, even at concentrations below current regulatory levels. Policymakers may consider lowering maximum contaminant levels to protect public health. KEYWORDS: Big data, Birth outcomes, Children's environmental health, Environmental epidemiology, International collaboration, Water quality

Fine particulate matter relates to white matter microstructure alterations in U.S. children

BACKGROUND AND AIM: Recent epidemiological findings have increased concern that exposure to fine particulate matter air pollution (aerodynamic diameter ≤2.5 µm; PM2.5) may be neurotoxic, even at low levels. Earlier human MRI and animal studies suggest that PM2.5 may impact white matter, yet questions remain about potential localized effects on microstructure. The goal of the current cross-sectional study was to examine how PM2.5 exposure relates to white matter microstructure in a large, diverse sample of children. METHODS: We examined the associations between annual average outdoor PM2.5 exposure and white matter microstructure in 7,602 9-10 year-old children from the landmark Adolescent Brain Cognitive Development study. Diffusion Tensor Imaging was used to characterize white matter microstructure using fractional anisotropy (FA). Annual average PM2.5 concentrations were assigned to the primary residential address at the baseline visit in 2016-2018 using an ensemble-based model approach. RESULTS:After correcting for multiple comparisons and adjusting for study site (n=21), socio-demographics, and MRI-related covariates, residential PM2.5 exposure (mean 7.65; range 1.72–15.90 ug/m3) was found to have significant non-linear associations with FA in multiple tracts. Increasing exposure to PM2.5 was associated with bilateral increases in FA in the uncinate fasciculus, superior longitudinal fasciculus, and inferior longitudinal fasciculus and bilateral decreases in the cingulate cingulum. PM2.5 exposure from 1.72 to 7.5 ug/m3 was also positively associated with increases in the left hemisphere fornix and right hemisphere anterior thalamic radiation. CONCLUSIONS:White matter microstructure in today’s developing youth may be affected by ambient fine particulate matter. Different patterns were observed in the left and right hemispheres and along specific white matter tracts. Our findings suggest that annual average PM2.5 exposure may relate to greater restricted diffusion as indexed by FA in tracts important for motor, emotion, and attention, which may reflect a combination of alterations in myelination, microarchitecture, and/or packing density. KEYWORDS: Air pollution, Particulate matter, Neurodevelopmental outcomes, Children's environmental health

Associations of Prenatal First-Trimester Metal Mixtures with Adiposity during Childhood in the Project Viva Cohort

BACKGROUND AND AIM: Prenatal metal exposure influences child health. We evaluated associations between prenatal metals and childhood adiposity. METHODS: Six nonessential (As, Ba, Cd, Cs, Hg, Pb) and four essential metals (Mg, Mn, Se, Zn) were measured in first trimester red blood cells from Project Viva participants. Sum of skinfold thicknesses (SST) (mm) (N=859), waist circumference (cm) (N=882), and body mass index (BMI) z-score (N=875) were measured at early childhood. In mid-childhood and late teen, SST (N=747-715), WC (N=748-717), BMI (N = 745-716), and dual-energy x-ray absorptiometry total- and trunk-fat-mass-index (kg/m2) (N=599-511) were measured. Linear regression and quantile g-computation models estimated associations between individual and mixtures of metals and adiposity, adjusting for confounders. RESULTS:Higher levels of prenatal cadmium were associated with increased SST at early childhood (β=0.30, 95% CI: 0.02, 0.58), and higher total (β=0.25, CI: 0.02, 0.48) and trunk fat (β=0.12, CI: 0.01, 0.23) at early teen. Cesium was associated with higher BMI at mid-childhood (β=0.18, CI: 0.02, 0.35) and trunk fat at early teen (β=0.24, CI: 0.003, 0.48). In mixture analyses, with concurrent adjustment for nonessential and essential metals, higher levels of prenatal nonessential metals were associated with higher total (β=0.07 per quartile, 95% CI: 0.01, 0.12), trunk fat (β=0.12, CI: 0.02, 0.22), WC (β=0.01, CI: 0.002, 0.01), and BMI (β=0.24, CI: 0.08, 0.41) at mid-childhood and higher total fat (β=0.06, CI: 0.002, 0.11) and BMI (β=0.19, CI: 0.02, 0.37) at early teen, while higher prenatal essential metals were associated with lower total (β=-0.07, CI: -0.12, -0.02), trunk fat (β=-0.13, CI: -0.21, -0.05), SST (β=-0.08, CI: -0.14, -0.01), WC (β=-0.03, CI: -0.05, -0.003), and BMI (β=-0.16, CI: -0.28, -0.04) at early teen. CONCLUSIONS:Prenatal nonessential metals were associated with increased adiposity, while essential metals were associated with decreased adiposity during childhood. KEYWORDS: Heavy metals, Obesity and metabolic disorders, Mixtures analysis, Children's environmental health

Short-Term Particulate Matter, Ultrafine Particles, and Blood Pressure among Children of the CorPuScula Study in Munich, Germany

BACKGROUND AND AIM: Short-term changes in particulate matter (PM) have been linked to blood pressure, but limited evidence exists on effects among children and for ultrafine particle (UFP) exposures. METHODS: We used data from 50 children (ages 6-10) in the German CorPuScula study, in which children attended up to 30 school health examinations in 2001 and 2002. Systolic and diastolic blood pressure (SBP and DBP, respectively; mmHg) were collected at each examination after children had rested for at least 5 minutes. PM10 (diameter ≤10 µm; µg/m3) and PM2.5 (diameter ≤2.5 µm; µg/m3) levels were measured at school using a low volume sampler for the entire study period, and UFPs (n/mL) were measured for six months using a particle counter. All exposures were evaluated for same-day working-hours (9:00-21:00) as well as one- and two-day lags, with UFPs also assessed for the 1-, 3-, 12-, and 24-hours prior to examination. Associations between air pollutants and blood pressure were evaluated using mixed linear regression models with a random participant intercept and adjusted for weekday as well as study day and physiological equivalent temperature index using natural cubic splines. Multipollutant models were conducted adjusting for centrally-monitored NO2. RESULTS:For the PM data (nobs=1,079; average 22 observations/child), median daily PM2.5 levels were 12.7 µg/m3. In the UFP data (nobs=420; average 8 observations/child), median daily UFP exposures were 21,544 particles/mL. We observed positive associations between PM and SBP that attenuated slightly upon adjustment for NO2, with strongest associations for 2-day lags (e.g., 1.45 mmHg [95% Confidence Interval: 0.78, 2.11] increase in SBP per 13.4 µg/m3 increase in PM2.5). PM10 was weakly inversely associated with DBP. No associations between UFP exposures and either outcome were apparent. CONCLUSIONS:While positive associations were observed between PM and SBP, we observed no evidence that UFP exposure influences blood pressure in children. KEYWORDS: Children's environmental health, Short-term exposure, Air pollution, Cardiovascular diseases, Multi-pollutant

Critical windows of perinatal particulate matter (PM2.5) exposure and preadolescent kidney function

BACKGROUND AND AIM: Air pollution exposure, especially particulate matter ≤ 2.5 μm in diameter (PM2.5), is associated with poorer kidney function in adults and children. Perinatal exposure may occur during susceptible periods of nephron development. We aimed to examine perinatal windows of susceptibility to PM2.5 exposure on preadolescent kidney function parameters and blood pressure (BP). METHODS: We used distributed lag nonlinear models (DLNMs) to examine time-varying associations between early life daily PM2.5 exposure (periconceptional through 1 year of life) and kidney parameters as well as BP in preadolescent children aged 8 to 9 years. Participants included 427 mother-child dyads enrolled in the PROGRESS birth cohort study based in Mexico City. Daily PM2.5 exposure was estimated at each participant’s residence using a validated satellite-based spatio-temporal model. Kidney function parameters included estimated glomerular filtration rate (eGFR), serum cystatin C, blood urea nitrogen (BUN), and BP. Models were adjusted for child’s age, sex and body mass index (BMI) z-score, as well as maternal education, indoor smoking report, seasonality and average first year of life PM2.5 exposure. RESULTS:Average perinatal PM2.5 was 22.7 µg/m3 and ranged 16.4-29.3 µg/m3. Early pregnancy PM2.5 exposures were associated with hyperfiltration in preadolescence. Specifically, we found that PM2.5 exposure between weeks 1 to 18 of gestation was associated with increased preadolescent eGFR. A constant 5 µg/m3 increase in PM2.5 sustained throughout this window would predict a cumulative increase in eGFR of 4.44 mL/min (95%CI: 1.37, 7.52). Additionally, late pregnancy levels of PM2.5 were associated with elevated BUN. In a stratified analysis, an association between PM2.5 and higher systolic BP was evident among boys but not girls. CONCLUSIONS:We identified perinatal windows of susceptibility to PM2.5 exposure with preadolescent kidney function parameters. Follow-up investigating PM2.5 exposure with peripubertal kidney function trajectories and risk of kidney disease in adulthood will be critical. KEYWORDS: particulate matter, kidney function, Cardiovascular diseases, Children's environmental health, environmental epidemiology,

Wildfire exposure {in utero} and use of respiratory medications in early childhood

BACKGROUND AND AIM: Wildfire smoke exposure is a growing concern as wildfires escalate in the western U.S. and abroad (bushfires in Australia). While gestational particulate matter exposure has been linked to increased incidence of childhood respiratory diseases, this association has not been examined with respect to wildfire smoke. Here, we use prescription claims data to investigate the relationship of childhood respiratory illness and in utero exposure to wildfire smoke. METHODS: The cohort consisted of infants (N182,000), born (2010-16) in one of the Metropolitan Statistical Areas (MSAs) of California, Oregon, Washington, Idaho, Nevada, or Montana and followed up to 8 years. We required had continuous prescription coverage and estimable birthdates via claims data in IBM’s MarketScan Commercial Claims Database. Our two outcomes were filled prescriptions of ‘upper respiratory’ and ‘lower respiratory’ medications. Using prescription claims, pediatric respiratory drugs (e.g., fluticasone, albuterol) were classified as ‘upper respiratory’ or ‘lower respiratory’ based on active ingredients, mode of administration, and drug class. Wildfire exposure, classified as number of smoke days during each trimester, for each MSA was determined from National Oceanic Atmospheric Association's Hazard Mapping System. First, second, and third trimesters were defined as 280-197, 196-99, and 1-98 days before estimated birthdate, respectively. We used proportional hazards models to assess the relationship of trimester-specific in utero wildfire smoke exposure and first use of respiratory medication. Models are adjusted for sex, birth season, birthyear, meteorological variables, and seasonality. RESULTS:Preliminary results suggest that gestational wildfire smoke is associated with both upper and lower respiratory medication usage during childhood. These associations are modified by season of birth and may differ by birthyear. Sensitivity analyses will account for preterm birth. CONCLUSIONS:Our results suggest that incidence of childhood respiratory illness, as indicated by respiratory medication usage, is impacted by in utero exposure to wildfire smoke. This abstract does not necessarily represent EPA policy. KEYWORDS: Air pollution, Wildfires, Children's environmental health, Respiratory outcomes, Natural disasters, Climate

Prenatal Maternal Phthalate Exposures and Trajectories of Childhood Adiposity from Four to Twelve Years

BACKGROUND AND AIM: Trajectories of child adiposity have lifelong consequences for future cardiometabolic risk. Prenatal phthalate exposures may affect processes that pre-program later childhood adiposity, but findings from studies examining these associations are conflicting. We examined associations between individual phthalate exposure biomarkers and total phthalate mixture during pregnancy with child adiposity trajectories. METHODS: Data were from 514 mother-child pairs in the PROGRESS cohort of Mexico City. We quantified 15 phthalate metabolites in 2nd and 3rd trimester maternal urine samples and created an average pregnancy measure using the geometric mean. We evaluated the 15 metabolites as nine biomarkers, including four metabolite molar sums. We measured child BMI z-score, bioimpedance-based fat mass index (FMI), and waist-to-height ratio (WHtR) at three visits between four and twelve years of age. We identified trajectories using multivariate latent class growth modeling, considering BMI z-score, FMI, and WHtR as joint indicators of latent adiposity. We then estimated associations of phthalate biomarkers with class membership using multinomial logistic regression. We used quantile g-computation to estimate the effect of the total phthalates mixture and assessed effect modification by sex. RESULTS:We detected three trajectories of child adiposity, “low-stable” remained stable at low adiposity (n=260), “low-high” started low then increased to high (n=147), and “high-high” started high and remained high (n=107). A doubling of the sum of di(2-ethylhexyl) phthalate metabolites (ΣDEHP), was associated with greater odds of being in the “high-high” trajectory (adjusted odds ratio [OR] = 1.53 [1.08, 2.19]; p = 0.02) in comparison to the “low-stable” group, whereas a doubling in mono(carboxy-isononyl) phthalate (MCNP) was associated with lower odds of being in the “low-high” trajectory (adjusted OR = 0.66 [0.45, 97]; p = 0.03). No sex-specific or total mixtures associations were detected. CONCLUSIONS:Pregnancy urinary concentrations of ΣDEHP metabolites and MCNP were associated with trajectories of child adiposity in opposing directions. KEYWORDS: Phthalates, Children's environmental health, Environmental epidemiology, Mixtures analysis

Street-View Greenspace Exposure and Objective Sleep Characteristics

BACKGROUND AND AIM: Epidemiologic studies on the association of residential greenness with children’s sleep quality are limited. Furthermore, most studies rely on satellite-derived measures of greenspace, which do not capture ground-level visible exposures. Google Street View (GSV) images offer the opportunity to measure greenspace from a street-based view and capture greenspace that may be most relevant for health. We aim to examine associations of GSV-based greenspace with objective sleep characteristics in Project Viva, a highly-phenotyped cohort in eastern Massachusetts participating in the Environmental influences on Child Health Outcomes (ECHO) consortium. METHODS: We applied the PSPNET deep learning segmentation algorithm to GSV images from 2007-2010 (mid-childhood visit) and 2012-2016 (early adolescence visit) to derive novel metrics of greenspace (e.g. % trees, %grass within each 360 degree image) at a 250m radius around participant’s residential addresses (N=717). In early adolescence (mean[SD] age 13.0[0.7] years), participants completed 5 days of wrist actigraphy recording for 10 hours/day. Sleep duration and efficiency were derived from actigraph data. We used linear regression to examine associations of GSV-based greenness metrics at mid-childhood and early adolescence with sleep at early adolescence, adjusting for child’s sex, race/ethnicity, and age, mother’s education and marital status, father’s education, household income, neighborhood median income, and population density. RESULTS:Among the 500 adolescents, 249 were girls. In unadjusted models, mid-childhood GSV-based total greenness (per interquartile range) was associated with longer sleep duration at early adolescence (9.3 min/day; 95%CI:3.2,15.5). This association was mostly driven by trees (8.2; 95%CI:1.7,14.6). In early adolescence, unadjusted associations were similar. However, in all fully adjusted models, the association was no longer present (e.g. 1.1; 95%CI:-5.5,7.7 for mid-childhood greenness). No associations were observed between GSV-based greenness and sleep efficiency. CONCLUSIONS:GSV-based greenness was not associated with sleep duration and efficiency in early adolescence among Project Viva participants. KEYWORDS: Built environment, green space, Children's environmental health, Environmental epidemiology

Widespread Exposure to Emerging and Previously Unmeasured Chemicals in Commerce in Pregnant women Across the US

BACKGROUND AND AIM: Prenatal chemical exposures are ubiquitous, can influence maternal, fetal and child health, and lead to adverse pregnancy and child health outcomes. The NIH Environmental influences on Child Health Outcomes (ECHO) Program, –comprised of geographically, racially, and ethnically diverse prospective pregnancy and birth cohorts, provides an unparalleled opportunity for understanding environmental exposures among US pregnant women. Our study applies a new and efficient biomonitoring method for quantifying 100 priority chemicals in commerce among pregnant women and characterizes predictors of prenatal exposures. METHODS: We developed methods to prioritize and measure 148 chemicals and their metabolites in eight priority classes in urine: pesticides (30), environmental phenols (17), parabens (6), organophosphate flame retardants/esters (OPFRs/OPEs) (9), phthalates/alternative plasticizers (31), polycyclic aromatic hydrocarbons (11), aromatic amines (42), and tobacco biomarkers (2). Using HPLC-MS/MS methodology, we analyzed 172 urine samples from pregnant women in nine of 72 ECHO pregnancy cohorts from five US states (CA, GA, IL, NH, NY) and Puerto Rico, representing a broad range of demographic characteristics as well as recruitment and sample collection strategies. We calculated detection frequencies and descriptive statistics for concentrations and assessed predictors by estimating associations of sociodemographic and sample collection characteristics with creatinine-standardized analyte concentrations. RESULTS:We detected 56 analytes in greater than 50% of samples (6 pesticides, 6 environmental phenols, 3 parabens, 2 OPFRs/OPEs, 18 phthalates/alternative plasticizers, 10 polycyclic aromatic hydrocarbons, and 11 aromatic amines). Forty-two analytes were not detected in any sample. Among widely detected analytes were three pesticides (two neonicotinoids and a metabolite of chlorpyrifos), BPS (a BPA replacement), several previously unmeasured phthalate metabolites, melamine and cyanuric acid. For many analytes, concentrations varied by cohort/location. CONCLUSIONS:Our study is the largest to date to measure these mostly understudied chemicals for which we found widespread exposure in a diverse sample of U.S. pregnant women. KEYWORDS: Chemical exposures, Phthalates, Phenols, Pesticides, Children's environmental health, Pregnancy outcomes

Early-life exposure to lead and antisocial behaviors during childhood and adolescence

BACKGROUND AND AIM: An association between lead (Pb) exposure and antisocial behaviors has been documented over the last decades, but the findings have been inconsistent. We aimed to estimate the association between prenatal, early-childhood, and preadolescent/adolescent (periadolescent) Pb exposure, and aggression and conduct problems in periadolescents of Mexico City. METHODS: We studied participants from the Early Life Exposures in Mexico to Environmental Toxicants (ELEMENT) cohorts. We used 1) maternal patella Pb at puerperium as a proxy for prenatal Pb exposure, 2) cumulative blood lead index (a summary of repeated blood Pb (BPb) measurements) between birth and 60 months of age as a measure of early-childhood Pb exposure, and 3) a single BPb measurement concurrent with behavioral evaluation as a measure of periadolescent Pb exposure. The parent-reported scales of aggression and conduct problems of the Behavioral Assessment System for Children-2nd version (BASC-2) were evaluated during periadolescence (at a mean age of 10.38 years). We estimated the association between Pb exposure at each stage and having obtained an elevated T-score (≥60) in each scale, while adjusting for household socioeconomic status, maternal education and age at delivery, child sex, and Pb exposure at previous stages. RESULTS:743, 704, and 595 participants were included in the models for prenatal, early-childhood, and periadolescent Pb exposure, respectively. On average, mothers had 10.6 years of education, and were 26.3 years old at delivery. 49% of children were female. A one-unit increase in Pb exposure during the prenatal period (1 µg/g of maternal patella Pb), but not during childhood or periadolescence, was associated with increased odds of aggression (Odds Ratio (OR)=1.021; 95% Confidence Interval (CI): 1, 1.041) and conduct problems (OR=1.019; 95% CI: 1.002, 1.037) CONCLUSIONS:Exposure to Pb during the prenatal stage, but not during childhood or adolescence, was associated with aggression and conduct problems during periadolescence in this longitudinal cohort study. KEYWORDS: Heavy metals, Neurodevelopmental outcomes, Children's environmental health

Association between prenatal metal exposure and respiratory symptoms in childhood

BACKGROUND AND AIM: Arsenic, manganese, and lead have been cross-sectionally associated with adverse respiratory outcomes in childhood but there is limited data on their combined effects starting in utero. Our aim was to examine associations between prenatal lead, arsenic, and manganese exposure and respiratory symptoms in childhood. METHODS: We assessed 637 mother-child dyads enrolled in the Programming Research in Obesity, Growth, Environment, and Social Stressors (PROGRESS) birth cohort in Mexico City. Blood manganese (BMn), arsenic (BAs), and lead (BPb) were measured in mothers at 2nd trimester. Ever wheeze, current wheeze and asthma diagnosis was based on the caregiver’s answers (yes/no) from the International Study of Asthma and Allergies in Childhood (ISAAC) survey applied at 4-5 and 6-7 years old. Covariates included maternal education, maternal asthma status, maternal age at enrollment, child’s sex, and prenatal environmental tobacco smoke exposure. A logistic mixed model regression was generated to simultaneously assess the association between prenatal metals and ever wheeze, current wheeze, and asthma diagnosis longitudinally across both the 4-5 and 6-7 year study visits. RESULTS:The mean ± standard deviation for log-transformed BMn, BPb, and BAs at 2nd trimester were 0.14 µg/dL ± 0.15, 0.48 µg/dL ± 0.27, and -1.11 µg/dL ± 0.19, respectively. In the adjusted model, BPb (Odds Ratio (OR): 1.95, 95% CI: 1.04-3.66) and BAs (OR: 3.44, 95% CI: 1.05-11.34) at 2nd trimester were associated with increased odds of ever wheeze and current wheeze, respectively. BMn at 2nd trimester (OR: 0.05, 95% CI: 0.01-0.27) was negatively associated with current wheeze. CONCLUSIONS:Prenatal exposure to arsenic and lead was associated with higher odds of ever and current wheeze while manganese was negatively associated with odds of current wheeze in childhood. These findings underscore the need to consider prenatal metal exposure, including low levels of exposure, in the study of respiratory diseases. KEYWORDS: Respiratory outcomes, Heavy metals, Allergies, Asthma, Children's environmental health

Residential greenspace and internalizing behaviors in early childhood

BACKGROUND AND AIM: Prior studies suggest that greenspace promotes better mental health among adolescents and adults. We investigated associations between greenspace and internalizing behaviors in earlier childhood. METHODS: This study included children in the Conditions Affecting Neurocognitive Development and Learning in Early Childhood (CANDLE) pregnancy cohort in Memphis, TN, within the ECHO-Pathways Consortium. Child behaviors were parent-reported using the Child Behavior Checklist. We examined the internalizing behavior broadband score and six related subscales. Greenspace was characterized at participant residences as the Normalized Difference Vegetation Index (NDVI) within 300m, percent tree canopy coverage of the census block group, and distance to the nearest park. Poisson regression models of raw scores were adjusted for a suite of child, maternal and neighborhood factors. RESULTS:In this sample (n=955), 65% of mothers identified as Black/African-American, 29% as White; 53% had at least a college/technical school degree. Children were on average 4.3 (SD 0.4) years old. Mean NDVI weighted across the year prior to outcome ascertainment was 0.59 (SD 0.08). The mean internalizing raw score was 6.2 (SD 2.1); 5.2% of the sample scored above the clinical threshold (63 t-score). A 0.1 unit higher NDVI was associated with a 9.4% lower internalizing score (95%CI: -17.5, -0.5), a 10.3% lower anxious/depressed score (95%CI: -19.3, -0.2), an 11.3% lower somatic complaints score (95%CI: -21.1, -0.3), and a 12.7% lower anxiety problems score (95%CI: -20.8, -3.8). Tree canopy was associated with the anxious/depressed and anxiety scores (-5.7% [95%CI: -11.0, -0.2] and -7.7% [95%CI: -12.5, -2.6], respectively, per 10% higher tree canopy coverage), but for the internalizing score confidence intervals included the null (-4.9% [95%CI: -9.7, 0.2]). No associations were observed with park proximity. CONCLUSIONS:More residential greenness and tree canopy were associated with fewer internalizing problems, suggesting that nature-based interventions may be relevant even early in childhood. KEYWORDS: Green space, Neurodevelopmental outcomes, Children's environmental health

Variability of urinary organophosphate esters (OPEs) during childhood: The HOME Study

BACKGROUND AND AIM: Organophosphate esters (OPEs) are widely used as flame retardants and plasticizers. Previous research has shown that gestational exposure to OPEs is variable, but no studies have examined the patterns and variability of postnatal exposure to our best knowledge. METHODS: In a prospective birth cohort (HOME Study, Cincinnati, OH, USA), we measured urinary OPE metabolite concentrations three times (16 and 26 weeks of gestation, at delivery) during pregnancy and five times (ages 1, 2, 3, 5, 8 years) during childhood. We log10-transformed creatinine-standardized OPE metabolite concentrations. We calculated Spearman correlation coefficients between postnatal OPE metabolites at each time point. We then used linear mixed models to calculate intraclass correlation coefficients (ICCs) for metabolite concentrations across the five postnatal time points. We further examined the association between three prenatal measures and early childhood (ages 1, 2, 3 years) exposure and subsequent metabolite concentrations at ages 5 and 8 years, including the main effect and the interaction term with the subsequent time points of metabolite measurement using linear mixed models. RESULTS:We detected (2-chloroethyl) phosphate (BCEP), bis(1,3-dichloro-2-propyl) phosphate (BDCIPP), and diphenyl phosphate (DPHP) in 79 to 100% of children. The median concentrations (µg/g creatinine) were 0.78-4.76 for BCEP, 4.02-7.42 for BDCIPP, and 2.57-11.48 for DPHP. We found moderate correlations between BCEP and BDCIPP (rs: 0.28-0.49, p 0.001) at all time points. ICCs for metabolites across five postnatal time points, however, indicated poor within-person correlation (0.09–0.25). Prenatal urinary DPHP concentrations were positively associated with postnatal DPHP concentrations at some time points. Positive associations between early (1-3 years) and late childhood concentrations (age 8 years) were found for BCEP and BDCIPP. CONCLUSIONS:We found high within-person variability of OPE exposure within children across the first 8 years of life, suggesting that multiple urine samples are needed to characterize children’s exposures to OPEs. KEYWORDS: Chemical exposures, Endocrine disrupting chemicals, Children's environmental health, Environmental epidemiology

Urban Environment and Growth and Obesity in Preschool Children from Six European Birth Cohorts

BACKGROUND AND AIM: Urban environmental exposures have been investigated separately in relation to growth and childhood obesity. In this study we aim to determine which prenatal and postnatal outdoor exposures are associated with childhood BMI, weight and height at age 3-4 years. METHODS: Height and weight were measured in 11,325 children aged 3-4 years from a multi-centre European cohort (France, Greece, Lithuania, Norway, Spain, and UK). BMI was calculated using the formula weight/height². Age-and-sex–standardized {z}-scores were calculated for height, weight and BMI. Urban environment exposures were estimated during pregnancy and childhood, including air pollution, built environment, green spaces, traffic, noise, meteorology, and socioeconomic deprivation index. Exposure-wide association study was performed using multiple single exposure linear regression models and accounting for multiple testing. RESULTS:Non linear associations were observed between exposure to Normalized Difference Vegetation Index (NDVI), a measure of greeness, and the outcomes leading to its categorization into tertiles. Pregnancy and postnatal NDVI in the first tertile were associated with greater height compared to the reference (second tertile), and remained statistically significant after correction for multiple testing (prenatal corrected {p}-value threshold for significance = 0.002; postnatal corrected {p}-value = 0.003). Other prenatal and postnatal markers of the urban environment such as walkability, building density and land use, nitrogen dioxide and fine particles were also associated with greater height, but were no longer statistically significant after correction for multiple testing. Fine particles during pregnancy were associated with higher BMI, and prenatal NDVI in the first tertile and air pollution were also associated with higher weight, but the {p}-values for these associations were above the corrected threshold for multiple testing. CONCLUSIONS:Living in a less green and more densely built and walkable area may affect preschool height in a positve manner. KEYWORDS: External exposome, Green space, Built environment, Children's environmental health, Obesity and metabolic disorders

Health Risk Assessment, Pollution Indices and Ecological Risk of Chemical Elements in Dust and Soils of Warri Metropolis, Southern Nigeria

BACKGROUND AND AIM: Adverse health effects have been linked with human exposure to chemical contaminants via diverse environmental pathways. This study assessed pollution levels, ecological risk and health risk of chemical elements in dust and soils of Warri Metropolis, Southern Nigeria. METHODS: Forty topsoil and settled dust samples were collected from ten sampling sites and analysed for chemical elements using Inductively Coupled Plasma Mass Spectrophotometry. The results were subjected to descriptive and inferential statistical analysis. Pollution Index (PI), Ecological Risk Index (ERI), Nemerov Integrated Pollution Index (NIPI) and Health Risk [Hazard Quotient (HQ) and Cancer Risk (CR)] were also calculated. RESULTS:Results revealed high concentrations (mg/kg) of Cu (2.1-38.4), Zn (21.4-268.0), Ag (0.015-2.720), Hg (0.09-0.51), P (120-1230) and B (3.0-7.0) in dusts; and Cu (1.0-42.7), Zn (6.2-492.0), Ag (0.016-100), Hg (0.05-0.33), P (30-1960) and B (3.0-5.0) in soils. Cu, Zn, Hg and P concentrations were highest and exceeded permissible limits in poultry farm dusts and Pessu soils. PI indicated very strong pollution (P15) for Ag and low pollution (P11) for Zn and Pb. Heavy pollution (NIPI3) in Effurun-Sapele road and Ekpan dusts was attributed to heavy metals. ERI revealed moderate risk for heavy metals in Igbudu market dusts and Pessu soils while other locations had low risks. Non-carcinogenic health risk evaluation showed HQs greater than 1.0 for V, Cr, Mn, Fe, Co, Cu, Zn, Pb, As, Al and Ca, signifying adverse health effects. The CRs of As, Cr, Cd, Ni and Co for children and; Cr and Co for adults were higher than the acceptable limit of 1.0×10⁻⁴, suggesting probable development of cancer. CONCLUSIONS:This study established the occurrence of dust and soil pollution by chemical contaminants from industrial emissions, vehicular sources and open-dumping. The chemical element concentrations pose carcinogenic and non-carcinogenic risks to children and adult residents of the study area. KEYWORDS: Chemical exposures, Heavy metals, Risk assessment, Cancer and cancer precursors, Children's environmental health

Sexually dimorphic associations between prenatal blood lead exposure and temporal processing in 6- to 7-year-old children in Mexico City

BACKGROUND AND AIM: Temporal-Response Differentiation (TRD) is an operant testing procedure to assess time perception. Few human studies have assessed whether lead (Pb) affects temporal processing. METHODS: As part of a prospective cohort study in Mexico, 547 six- to seven-year-old children completed a TRD task to assess temporal processing. Pb was measured in maternal blood collected during the second trimester (2T), third trimester (3T), and on the day of delivery; and, in infant’s umbilical cord (CB) at delivery. During the TRD task, children repeatedly held a lever down for 10-14 seconds to receive a reward. We examined the frequency, average duration, and standard deviation of timing holds. General linear models were used to examine associations between each Pb measure and each outcome adjusted for child’s age at testing, maternal education, and socioeconomic status. Sexually dimorphic effects were evaluated using 2-way interactions between child sex and exposure. RESULTS:In the number of holds models, there was a significant interaction between CB lead levels and sex (β=1.5; 95%CI: 0.12, 2.9) with boys having a reduced number of holds with higher Pb than girls. In the average duration of holds models, there were significant interactions of 3T and delivery Pb with sex (β=-0.72; 95%CI: -1.33, -0.12 and β=-0.59; 95%CI: -1.1, -0.1, respectively) with males having longer durations with higher Pb than girls. In the standard deviation of hold durations models, there were significant interactions of 3T, delivery, and CB Pb with sex (β=-0.79; 95%CI: -1.3, -0.23, β=-0.78; 95%CI: -1.2, -0.3, and β=-1.1; 95%CI: -1.7, -0.4, respectively) with boys having greater variability in their hold durations with higher exposure than girls. CONCLUSIONS:Our results offer new insights into the cognitive mechanisms involved in prenatal Pb-related neurodevelopmental deficits, implicating temporal processing mechanisms, and provide a new approach to evaluate sex-specific effects of lead exposure. KEYWORDS: Heavy metals, Neurodevelopmental outcomes, Children's environmental health

Contributions of nearby agricultural insecticide applications to indoor residential exposures

BACKGROUND AND AIM: Pesticide exposure has been associated with adverse health effects; individuals may be exposed to pesticide drift from nearby agricultural fields. We evaluated relationships between agricultural insecticide applications and insecticide levels in carpet dust. METHODS: We measured concentrations (ng/g) of allethrin, azinphos-methyl, carbaryl, chlorpyrifos, cyfluthrin, cypermethrin, deltamethrin, diazinon, malathion, permethrin, phosmet, propoxur, tetramethrin in carpet-dust samples collected from 558 California homes. After using the California Pesticide Use Reporting database (reporting unit: ~2.59km2) to estimate agricultural insecticide use within buffers with radii of 0.5-, 1-, 2-, 3-, and 4km around each home, we calculated the density of use (kg/km2) during 30-, 60-, 180-, and 365-day periods prior to dust collection and evaluated relationships between insecticide density and concentrations in dust. We modeled natural-log transformed dust concentrations using Tobit regression for insecticides with 40% detection. For insecticides detected ≤40%, odds of detection were modeled with logistic regression. Models were adjusted for season, year, occupation, and home/garden uses. RESULTS:Detection frequencies ranged from 10% for allethrin, azinphos-methyl, and malathion to 80% for diazinon, chlorpyrifos, and permethrin. Chlorpyrifos applications within 1-4km of the residence during all time periods were associated with higher dust concentrations compared to homes without nearby applications. Carbaryl applications within 2-4km of homes 30- and 60-days prior to dust collection were associated with higher concentrations (30-days/2km, β any applications vs. none=8.2, 95%CI 1.8-37.6; 60-days/4km, β median density vs. none=4.2, 95%CI 1.1-15.7). Cyfluthrin and phosmet applications 180- and 365-days prior to dust collection were associated with higher odds of detection (365-days/4km, highest density vs. none: OR=2.2, 95%CI 1.2-4.0 and 6.2, 95%CI 3.3-12.0, respectively) compared to homes with no nearby use. CONCLUSIONS:Agricultural insecticide use within 4km of a home could be an important determinant of indoor contamination. Our findings provide valuable information for the development of exposure metrics for epidemiologic health studies. KEYWORDS: Pesticides, Exposures, Exposure assessment, Environmental epidemiology, Children's environmental health

Identifying periods of susceptibility to perfluoroalkyl substances and bone mineral density in early adolescence: the HOME Study

BACKGROUND AND AIM: Perfluoroalkyl substance (PFAS) exposures may affect childhood bone mineral density (BMD), but no studies have assessed periods of heightened susceptibility. We estimated associations of individual PFAS and their mixture during gestation and three times during childhood with BMD in early adolescence. METHODS: We examined 222 mother-child pairs enrolled in a prospective pregnancy and birth cohort in Cincinnati, OH from 2003-2006. We measured concentrations of perfluorooctanoic acid (PFOA), perfluorononanoic acid, perfluorohexanesulfonic acid, and perfluorooctanesulfonic acid in maternal serum collected at 16 weeks gestation and child serum collected at age 3, 8, and 12 years. At age 12 years, we measured areal BMD at six skeletal sites with dual x-ray absorptiometry and calculated height-, age-, sex-, and population ancestry-specific BMD Z-scores. Using linear regression, we estimated covariate-adjusted differences in BMD Z-scores per doubling of PFAS concentrations at each period. Using hierarchical Bayesian kernel machine regression (hBKMR), we estimated period-specific associations and posterior inclusion probabilities (PIPs) to determine periods of heightened susceptibility to PFAS mixtures. RESULTS:Associations were strongest for PFOA and forearm (1/3 distal radius) BMD, with differing periods of susceptibility for males and females. Among males, forearm BMD Z-score differences (95% confidence interval) per doubling of PFOA were -0.26 (-0.50, -0.02), -0.33 (-0.68, 0.02), -0.24 (-0.61, 0.13), and -0.00 (-0.30, 0.29) for gestation and ages 3, 8, and 12, respectively. Among females, the corresponding estimates were -0.12 (-0.38, 0.15), -0.07 (-0.44, 0.30), -0.30 (-0.76, 0.17), and -0.44 (-0.81, -0.07). Patterns were generally similar but weaker for other PFAS and skeletal sites. Period-specific PIPs from hBKMR models were highest for the PFAS mixture at age 8 for males (0.76) and age 12 for females (0.62). CONCLUSIONS:PFOA and PFAS mixtures were associated with lower BMD in early adolescence. Susceptibility to PFAS may occur earlier in life for males compared with females. KEYWORDS: Chemical exposures, Children's environmental health, Mixtures analysis, PFAS, Endocrine disrupting chemicals, Environmental epidemiology

Proximity to petrochemical industry and risk of childhood asthma occurrence

BACKGROUND AND AIM: Petrochemical industrial complexes would emission s several pollutants that were associated with adverse effect in respiratory system. Aim of present study is using different surrogate indices of petrochemical exposure to evaluate the effect on childhood asthma. METHODS: We conducted a singleton birth cohort by using the 2000-2014 Taiwan Birth Certificate Database (TBCD), linked with Nation Health Insurance Research Database (NHIRD) for follow-up. The distance to the nearest petrochemical industrial parks (PIPs) and benzene exposure level were uses as surrogating petrochemical exposure indices. Exposure windows were divided into prenatal and postnatal periods. The Cox proportional hazard models were applied to evaluate the effects. RESULTS:In total, 227601 asthma cases were newly diagnosed until to 2016/12/31. Younger maternal age and greater proportions of male, preterm birth, low birth weight, delivery by Caesarean section, maternal gestational hypertension were observed for asthma children than non-asthma children. The results showed that compared with those living in the farthest, living close to PIPs (HR = 1.67, 95% CI: 1.65–1.70, for 10 km vs. ≥ 40 km) were observed to have an increased risk of asthma occurrence. Higher benzene exposure levels during prenatal or postnatal period having higher risks of asthma were also found. CONCLUSIONS:In present study, we found living near PIPs and exposure to higher level of benzene during prenatal or postnatal period would increase the risk of childhood asthma occurrence. KEYWORDS: Children's environmental health, Asthma, Chemical exposures

Influence of Dose Metrics on Associations between Exposure to Black Carbon and Exhaled Nitric Oxide in Children with Asthma

BACKGROUND AND AIM: Few studies examined the associations between fractional exhaled nitric oxide (FeNO) and traffic-related air pollution (TRAP) from individual dosimetry models. Our study aimed to assess the associations between FeNO and black carbon (BC, as a surrogate of TRAP) from four dose metrics. Besides, we aimed to examine if more refined dose metrics would impact the health risk estimates. METHODS: We recruited 35 children with asthma living in environmental justice communities. Daily FeNO measurements were followed for up to 30 days for each child. Four dose metrics were calculated based on measurements from a central-site and personal monitors, with and without estimates of inhalation rates based on a self-reported time-activity questionnaire. Linear mixed effect models were used to test the associations of BC at different lag periods and FeNO. RESULTS:Exposure to BC was positively associated with FeNO within the first 24 hours of exposure. A log-transformed interquartile range increase in inhaled dose from metrics that used central-site estimates (Model A, C) increased 7% (95% CI: 1% - 14%) and 6% (95% CI: 0% - 13%) in FeNO at lag 0-6 hours in the adjusted model 4. The strength of association persisted at lag 7-12 hours and then decreased and became insignificant at longer lag periods. Similarly, the inhaled dose from personal monitored estimates (Model B, D) was positively associated with FeNO within the first 24 hours. Using personal BC measurements or dynamic inhalation rate did not improve prediction of changes in FeNO. CONCLUSIONS:Our findings revealed that inhaled BC dose significantly increases FeNO level within the first 24 hours of exposure. Our results provide evidence that personal exposure to BC was associated with airway inflammation among children with asthma. Our results also add to the limited number of studies that compared the effect estimates among alternative dose metrics within a single study. KEYWORDS: traffic-related, air pollution, Short-term exposure, asthma, Children's environmental health