Environmental Epidemiology Research Articles

Gene-Environment Joint Linkage and Association Analysis of Arsenic Exposure and Diabetes-Related Traits in the Strong Heart Family Study

BACKGROUND AND AIM: Inorganic arsenic exposure is of increasing interest as a possible risk factor for diabetes, and although there is growing research on potential toxicological mechanisms, the biology remains incompletely understood. METHODS: This is a cross-sectional, gene-environment interaction study using data from n=1,915 participants from 40 pedigrees in the Strong Heart Family Study with measured arsenic exposure (sum of inorganic and methylated urine arsenic species, dichotomized for analysis), age, sex, educational attainment, body mass index, genotypes, family relationships, fasting glucose, HOMA-IR, and HOMA-B. We conducted a multipoint variance components linkage analysis testing for arsenic-locus interactions (GxE) predicting glucose, HOMA-IR, or HOMA-B in separate models, adjusted for potential confounders of the arsenic-outcome relationships. We then tested for SNP-trait associations and SNP GxE in linkage regions. Analyses were implemented using the software Sequential Oligogenic Linkage Analysis Routines (SOLAR). RESULTS:We localized three quantitative trait loci that showed suggestive evidence for differential genetic contributions to the outcome in the presence of higher vs. lower inorganic arsenic exposure. The first was an interaction locus for fasting glucose at chromosome 17q22 (LOD 2.6 in GxE model, LOD 1.5 in standard linkage model without interaction). The second was an interaction locus for HOMA-IR at chromosome 7p13 (LOD 2.0 in GxE model, 0.4 in the standard linkage model). The third was an interaction locus for HOMA-B at chromosome 18q22.2 (LOD 2.0 in GxE model, LOD 1.7 in standard linkage model). Several SNPs showed Bonferroni-significant associations with diabetes-related traits, with the strongest evidence for rs4793861 in the MSI2 gene for glucose; however, no tested SNPs showed Bonferroni-significant interactions with arsenic. CONCLUSIONS:We found suggestive evidence for different locus-specific contributions to diabetes trait variability in the presence of higher inorganic arsenic exposure, and SNPs in these regions that were associated with diabetes traits. Further research (e.g., linkage fine-mapping) might further elucidate this biology. KEYWORDS: Environmental Epidemiology, Toxicology, Heavy Metals

Associations between Long-term UFP Exposure and Certain Inflammatory Biomarkers and Blood Lipid Indicators: An Analysis Featuring Peak Exposure Metrics

BACKGROUND AND AIM: __ __Recent studies suggested that certain inflammatory biomarkers and blood lipid indicators may be associated with annual average of ultrafine particle (UFP) exposure. However, associations with repeated peaks of UFP exposure are rarely studied. __We previously proposed a novel metric for assigning UFP peak exposure. We hypothesize that, compared to annual average exposure, cumulative UFP peak exposure over a year may have stronger associations with the aforementioned biomarkers and indicators. METHODS: __ __Blood samples from 452 participants living in the Greater Boston Area were analyzed for three inflammatory biomarkers [high-sensitivity C-reactive protein, interleukin-6 and tumor necrosis factor receptor 2 (TNF-RII)] and four blood lipid indicators [total cholesterol, triglycerides, high-density lipoprotein (HDL) and low-density lipoprotein (LDL)]. __UFP exposure data was hourly averages of particle number concentration (PNC) estimated using a land use regression model based on mobile-monitored ambient UFP concentration, with time-activity adjustment applied. __Both blood sample and UFP exposure data were previously reported in the literature. __Multivariate regression models were used to analyze associations between our peak exposure metric and the outcome biomarkers and indicators. RESULTS:__ __An increase in cumulative UFP peak exposure was significantly associated with a decrease in blood total cholesterol and HDL level. It was also significantly associated with an increase in blood TNF-RII among Non-Hispanic Whites. __The association with total cholesterol level was not found using annual average UFP exposure. __The change of these three outcomes associated with one inter-quartile range’s (IQR) increase in peak exposure was greater than the change associated with one IQR’s increase in annual average exposure. CONCLUSIONS:__ __Associations between UFP exposure and inflammatory biomarkers and blood lipid indicators based on our peak exposure metric were generally similar to associations based on the annual average metric. However, for certain biomarkers and indicators, our peak exposure metric may show stronger associations than the annual average. KEYWORDS: Traffic-related, Particulate matter, Cardiovascular diseases, Exposure assessment, Modeling, Environmental epidemiology

Temperature related health burden in the Attica region, Greece, under two different climatic scenarios for the near and distant future

BACKGROUND AND AIM: The gradual increase in ambient temperature and in the frequency and intensity of heat waves, related to climate change, may lead to premature deaths. Our study aimed to evaluate temperature-related deaths in the Attica region, Greece, under two climate change scenarios for the periods 2031-2050 and 2081-2100. METHODS: We estimated the number of deaths from natural, cardiovascular and respiratory causes attributed to the change of temperature between a reference period (1991-2005), and two future periods (2031-2050 and 2081-2100) under two climatic scenarios (RCP4.5 -“medium” and RCP8.5 - “high”) using exposure response functions based on the literature. We also estimated the number of excess deaths due to heatwaves. RESULTS:Under the RCP4.5 scenario, the mean annual number of deaths from natural causes attributed to the increased temperature during the warm season (April-September) was 2 times higher in the 2031-2050 period and about 2.5 times higher in the 2081-2100 period compared to the reference period (230, 95% CI: 207-253 deaths).Under the RCP8.5 scenario, the number of attributable deaths was 511 (95% CI: 461-560) and 1316 (95% CI: 1191-1439) for the periods 2031-2050 and2081-2100, respectively. The excess number of deaths during heatwaves from 138 (95% CI: 118-159) in the reference period increased to 1145 (95% CI: 980-1314) under the RCP8.5 in the 2081-2100 period without taking into account population adaptation to climate change. The impact of increasing temperature was more severe on vulnerable populations, such as the elderly. The number of avoidable deaths due to increasing temperature in the cold season (October- March) did not compensate for the excess mortality in the warm season. CONCLUSIONS:The increase in temperature as a result of climate change is a major public health issue and urgent policies to mitigate its effects are crucial. KEYWORDS: climate, environmental epidemiology, mortality, temperature, temperature extremes

Effect of heatwaves and greenness on mortality among Chinese elderly people

BACKGROUND AND AIM: Both heatwaves and greenness have been shown to affect health, but the evidence on their joint effects is needed to be better elucidated. We aim to assess the associations of the combined exposure to greenness and heatwaves and whether the effect varies between different subgroups. METHODS: We utilized the Chinese Longitudinal Healthy Longevity Survey (CLHLS), a prospective cohort of older adults aged 65. We used the number of heatwave days in one year before death to estimate individual long-term heatwaves exposure and cumulative Normalized Difference Vegetation Index (NDVI) during follow-up to assess individual greenness exposure. Cox proportional hazards models were used to assess the effects of greenness, heatwaves, and their interaction on mortality, adjusted for a range of covariates. We conducted subgroup analyses by residence, gender, and age. RESULTS:There were 20,758 participants in our study, totaling 67,312 person-years of follow-up. The mean NDVI was 0·41 (SD 0.13), and the mean number of heatwave days was 8.92 (2.04). In the adjusted model, the mortality hazard ratio (HR) for each 3-day increase in heatwave days was 1.04 (95% CI 1.04,1.05), each 0.1-unit decrease in cumulative NDVI was 1.06 (1.05,1.07). In the adjusted model with an interaction term, the HR for the interaction term was 1.01 (1.01,1.02) with a p-value less than 0.001. In our subgroup analyses, the HR for each 3-day increase in heatwave days was higher in urban areas than in rural areas (1.06 vs. 1.03), and the HR for 0.1-unit decrease in NDVI was higher in urban areas than in rural areas (1.08 vs. 1.04). CONCLUSIONS:Greenness can protect against the effect of heatwaves on mortality, and the number of heatwave days will affect the health effects of greenness. Urban dwellers have a higher response to the detrimental effect of heatwaves and have a higher marginal benefit from greenness exposure. KEYWORDS: Green space, Temperature, Environmental epidemiology

15-F2t-Isoprostane during the lifespan: from children to middle age

BACKGROUND AND AIM: 15-F2t-Isoprostane is a product of arachidonic acid non-enzymatic oxidation by free radicals. Validated in animal and clinical models, it is a golden standard tool for the oxidative status assessment. The aim of this study is to assess how this biomarker levels could vary during the lifespan. METHODS: The 15-F2t-Isoprostane urinary concentration was measured in 889 subjects aged between 7 and 69. The analyses were performed by ELISA kit. The levels of 15-F2t-Isoprostane were normalised by creatinine (Kinetic Jaffé method). The analysis was performed considering 10-year age classes: 10 years, 10-19 years, 20-29 years, 30-39 years, 40-49 years, 50-59 years, 60-69 years. RESULTS:The average 15-F2t-Isoprostane levels were 4.41 ± 4.69 ng/mg creatinine. The descriptive analysis reveals an apparent reduction in the age-class 11-19 years (2.78 ± 2.54 ng/mg creatinine). The average predicted values for the age classes resulting from a GLM analysis are: 10 years: 3.72 ng/mg creatinine (p0.001, 95%C.I. 2.87-4.58); 10-19 years: 2.78 ng/mg creatinine (p0.001, 95%C.I. 2.25-3.31); 20-29 years: 4.08 ng/mg creatinine (p0.001, 95%C.I. 3.04-5.12); 30-39 years 5.09 ng/mg creatinine (p0.001, 95%C.I. 4.20-5.98); 40-49 years 4.98 ng/mg creatinine (p0.001, 95%C.I. 4.35-5.60); 50-59 years: 4.28 ng/mg creatinine (p0.001, 95%C.I. 3.68-4.88) and 60-69 years: 5.23 ng/mg creatinine (p0.001, 95%C.I. 3.83-6.62). CONCLUSIONS:Oxidative stress biomarkers levels are currently studied in various populations due to their increase in pre-clinical stages of diseases that could be related to lifestyle, environmental or occupational exposures. The main evidence consists in the reduction in the concentration of 15-F2t-isoprostane in adolescence, which leads to an interruption of the constant increase in oxidative stress over the years. The knowledge of how the 15-F2t-Isoprostane levels change during the lifespan and the assessment of the trend-shape of this phenomenon are crucial to better understand the meaning of this biomarker in molecular epidemiology research. KEYWORDS: Long-term exposure, Environmental epidemiology, Molecular epidemiology

Knot placement in the Distributed non-linear lag models (DNLM) framework

BACKGROUND AND AIM: Distributed non-linear lag models (DNLM) provide flexible methods to model non-linear exposure-response relationships over time, with common applications in environmental epidemiology. When the lag functions of the cross-basis include a spline function, the number and placement of knots for the function are specified in the model although the preferred number and placement of knots may not be known a priori. Our goal is to develop a flexible algorithm to explore the impact of knot selection on the final model. METHODS: We use data describing weekly maternal exposure to nitrogen dioxide (NO2) and fetal growth trajectories during the first 12, 20 and 12-20 weeks of pregnancy for 2,021 women from the INfancia y Medio Ambiente (INMA) project (2003-2008) for the exploration of two knot selection methods. We generated multiple search grids containing potential knot placement for the lag dimension of the cross-basis; each search grid had a distinct level of resolution for knot placement. We compared the number and placement of knots from each grid as well as the inference from the resulting DLNM models utilizing knots selected from each search grid. RESULTS:Our results showed that models with 12 lags resulted in the same number and placement of knots regardless of search grid used. Similarly, the models with 20 lags (i.e., both the 20 week and 12–20-weekfetal growth trajectories), produced the same number of knots regardless of search grid used; however, the placement of the knots changed depending on the search grid utilized. CONCLUSIONS:Choosing an appropriate search grid for knot selection in the model may impact the knot vector selected in some contexts. A simulation study to investigate the model impact of different knot selection processes may reveal more detailed information about the overall importance of this step in the DLNM framework and model building. KEYWORDS: DLNM, knot selection, splines, DLM

Relationship between blood cadmium and iron status among Korean population

BACKGROUND AND AIM: In recent studies for the association between blood Cd and iron deficiency (ID), it was found that ID elevates blood cadmium levels in children, adolescents or adults. But, the previous studies almost only used either serum ferritin or Hemoglobin (Hb) for evaluating ID or iron deficiency anemia (IDA). Blood Cd level according to actual iron status have not been studied. The objective of this study was to investigate the association between Cd and Iron status(normal, negative iron balance, iron deficiency erythropoiesis, IDA) and the risk for high cadmium level depending each iron status using a large-scale nationwide representative sample of the Korean. METHODS: A nationwide cross-sectional study was conducted to examine the relationship between iron status and blood Cd level in 3,366 men and 3,443 women over 10 years of age, using data from the 2010–2012 Korean National Health and Nutrition Examination Survey (KNHANES). We used analysis of variance (ANOVA) or chi-square test to estimate differences in the demographic and biochemical characteristics of participants. Multiple logistic regression models were also used to examine the associations between iron status and high blood Cd level (75th percentile of Cd level in the current sample; 1.37μg/L) RESULTS:Mean values of blood Cd gradually increased in accordance with poor iron status (P-trend0.001). We found that increased risk for high Cd level was associated with poor iron status after adjusting for age, sex, BMI, education, income, cigarette smoking and alcohol consumption. In a sex and menopause stratified analysis, a significant association was stronger only in premenopause women, not in men, menopause women CONCLUSIONS:We found a relationship between poor iron status and blood Cd in a representative population sample of Korean adult. Our results suggested considering the reproductive toxicity of Cd, management for blood Cd level would be necessary in women of childbearing age with poor iron status KEYWORDS: Heavy metals, Environmental epidemiology

Longitudinal analysis of DNA methylation in relation to gestational perfluoroalkyl substance exposure: An epigenome-wide association study

BACKGROUND AND AIM: Alterations to DNA methylation may underlie the association between gestational exposure to perfluoroalkyl substances (PFAS) and adverse health outcomes in children. However, few studies have examined the association between gestational PFAS exposure and DNA methylation, and no studies have repeated DNA methylation data across childhood. We examined associations between gestational PFAS exposure and repeated measures of offspring peripheral leukocyte DNA methylation among 266 mother-child pairs enrolled in the HOME Study (Cincinnati, OH). METHODS: We quantified serum concentrations of perfluorooctanoate (PFOA), perfluorooctane sulfonate (PFOS), perfluorononanoate (PFNA), and perfluorohexane sulfonate (PFHxS) in mothers at ~16 weeks gestation. We measured DNA methylation at delivery (cord blood) and age 12 years using the Illumina HumanMethylation EPIC BeadChip. We analyzed the associations between log2-transformed PFAS concentrations and repeated DNA methylation measures using generalized estimating equations. Visit by PFAS interaction terms were used to test the stability of these differences across time. We performed Gene Ontology (GO) enrichment analysis to identify significant biological pathways. RESULTS:A total of 35 loci were significantly associated with PFAS (false discovery rate, q 0.05). Among the 5 loci for PFOS, 10 for PFOA, 7 for PFHxS, and 13 for PFNA (q 0.01), none overlapped. These loci mapped to genes (e.g., AGAP1, HPSE2, HABP2, RNF13, RADIL, and TMEM56) that are associated with cancers, cardiovascular disease, and cognitive function. We found little evidence that associations between PFAS and DNA methylation changed over time. The most significant GO pathways included homophilic cell adhesion, cell-cell adhesion and integral component of plasma membrane. CONCLUSIONS:Using longitudinal data, we identified loci associated with PFAS that have not been reported. Several loci were in genes linked to PFAS-associated health outcomes. Future studies are needed to confirm our findings and examine whether DNA methylation mediates associations between gestational PFAS exposure and offspring health. KEYWORDS: PFAS, Epigenomics, Biomarkers of exposure, Environmental epidemiology

The Role of Immigrant Enclaves and Ambient Air Pollution Exposure in the Immigrant Birthweight Paradox

BACKGROUND AND AIM: Foreign-born Black and Latina women have higher birthweight infants that their US-born counterparts, despite generally worse socioeconomic status and prenatal care access, i.e. immigrant birthweight paradox (IBP). Residence in immigrant enclaves and associated socio-cultural and economic benefits may be drivers of IBP. Yet, enclaves are known to have higher air pollution, a risk factor for lower birthweight. We investigated IBP accounting for both residential immigrant composition and air pollution exposure. METHODS: In the Boston-based Children’s HealthWatch cohort of non-smoking women-child dyads (n=2,116), we obtained birthweight-for-gestational age z-scores (BWGAZ) for US-born births from 2006-2015. Using maternal address, we estimated trimester-specific PM2.5 concentrations, proximity to major roadways, and immigrant orientation level based on census-tract percentages of foreign-born, non-citizen, and linguistically-isolated households (proxy for immigrant enclave). We fit multivariable linear regressions of BWGAZ by maternal nativity and sequentially adjusted for immigrant orientation and ambient air pollution sources. We also examined effect modification of IBP by immigrant orientation. RESULTS:Fifty-one percent of women were foreign-born. Immigrant orientation was negatively correlated with PM2.5 concentrations and proximity to major roads. Overall, children of foreign-born women had 0.19 (95% CI: 0.10, 0.28) higher mean BWGAZ than those of US-born mothers, demonstrating IBP in our cohort. BWGAZ did not significantly change after adjustment for immigrant orientation and air pollution sources. However, we found an additive effect of immigrant orientation on maternal nativity (p-interaction: ≤0.01), with IBP observed only in moderate (0.27, 95% CI: 0.09, 0.46) and high (0.29, 95% CI: 0.15, 0.44) immigrant-oriented tracts. CONCLUSIONS:Foreign-born women in higher immigrant-oriented areas had higher BWGAZ, supporting the role of immigrant enclaves on IBP. Prenatal air pollution exposure did not impact associations observed in this study. Future research should consider the role of immigrant enclaves and their spatial correlation with potential protective resources and environmental risk factors. KEYWORDS: Air pollution, Particulate matter, Birth outcomes, Environmental disparities, Exposures, Environmental epidemiology

Investigating the role of clinical measures before dialysis treatment as mediators in the association between ambient temperature and mortality/hospital admissions

BACKGROUND AND AIM: Typical thermoregulatory responses among healthy individuals include reduced blood pressure and perspiration. Individuals living with end-stage renal disease (ESRD) are susceptible to physiological fluctuations caused by ambient temperature changes that may increase health complications. We investigated whether systolic blood pressure (preSBP) and interdialytic weight gain change (IDWG) can independently mediate the association between ambient temperature and hospital admissions and between mortality outcomes among ESRD patients. METHODS: The study population consisted of ESRD patients in Philadelphia County, PA, from 2011 to 2019 (n=1981). Within a time-to-event framework, we estimated the association between daily maximum dry-bulb temperature (TMAX) and, in separate models, all-cause hospital admissions (ACHA) and all-cause mortality (ACM) outcomes during warmer calendar months. The study employed the difference (c-c') method to decompose total effect models for ACHA and ACM using preSBP and IDWG as time-dependent mediators and then applied proportion mediated (PM) using indirect effects. Covariate adjustments for total and direct effect models include age, race/ethnicity, blood pressure medication use, treatment location, preSBP, and IDWG. RESULTS:Based on Lag 2- Lag 1 temporal ordering, 1°C TMAX increase in was associated with an increased hazard of ACHA by 7.6% (adjusted hazard ratio (HR), 1.076; 95% confidence interval, 1.070-1.082) and ACM 7.5% (adjusted HR, 1.075, 1.050-1.100). Mediation analysis for hospital admission events identified significant indirect effects for all three studied pathways and significant indirect effects for IDWG and conjoined preSBP + IDWG pathways for mortality. The mechanistic path for IDWG, independent of preSBP, demonstrated inconsistent mediation in ACHA (-2.67%) and ACM (-6.26%). PM estimates from the conjoined pathway suggest that it is likely not a candidate path in combined pathway models. CONCLUSIONS:This work has provided a step toward understanding physiological linkages that may explain or suppress temperature-driven hospital admissions and mortality risks. These findings could help discern candidate heat-adaptation interventions. KEYWORDS: Survival, temperature, Short-term exposure, Epidemiology, Mortality, Environmental epidemiology

Water as a relevant source of inorganic arsenic exposure in U.S. cities

BACKGROUND AND AIM: Water arsenic represents a health concern for US residents served by community water systems (CWS) exceeding the Environmental Protection Agency (EPA) Maximum Contaminant Level (MCL) of 10 µg/L, and for residents reliant on unregulated private wells. Little is known on the contribution of CWS arsenic exposures below the MCL, particularly in urban areas. We evaluated whether water arsenic is a relevant source of urinary arsenic in urban settings in the Multi-Ethnic Study of Atherosclerosis (MESA). METHODS: We examined the association of CWS arsenic with urinary arsenic concentrations (µg/L) for 910 MESA participants from Baltimore, Chicago, Los Angeles, New York City, St. Paul, MN, and Winston-Salem, NC. We assigned three-year population weighted averages of CWS arsenic concentrations using 2006-2008 EPA data (before MCL implementation) to participants by zip code. Dietary information and urinary samples were collected in 2000-2002. Urinary arsenic species were measured using HPLC-ICPMS. We accounted for seafood intake using a residual based method. Linear models adjusted for race/ethnicity, income, smoking status, diet (rice) and urinary creatinine to account for urine dilution. RESULTS:CWS arsenic levels ranged from 0.5 µg/L (LOD) in several cities to 3.05 µg/L in Chicago (median 0.44 µg/L). CWS arsenic was positively correlated (p 0.05) with urinary inorganic arsenic (Spearman’s rho 0.11), dimethylarsinate (DMA) (0.09), monomethylarsonate (MMA) (0.16) and their sum (0.10). In adjusted models, a 1 µg/L change in CWS arsenic was associated with 1.29 (95%CI 1.01, 1.64) higher inorganic arsenic and 1.19 (1.03, 1.38) higher MMA. The association with DMA, sum of inorganic and methylated species was not statistically significant. CONCLUSIONS:Drinking water arsenic is a relevant source of exposure and internal dose in US urban areas, determined by urinary arsenic biomarkers. Research is needed to determine health implications of chronic exposure to water arsenic below the MCL, including vulnerable subpopulations. KEYWORDS: arsenic, drinking water, exposure assessment, environmental epidemiology

Nitrate in Danish household tap water and the risk of small-for-gestational-age, 1991-2015

BACKGROUND AND AIM: Findings from studies examining the effects of prenatal exposure to nitrate from drinking water on fetal growth have been inconsistent, small, and mostly ecologic in nature. The aim of this study was to assess maternal exposure to nitrate (NO₃⁻) in drinking water and small-for-gestational-age in a large population-based cohort. METHODS: Exposure estimates from each maternal address during pregnancy were obtained from a national public and private water monitoring database. We linked these individual household nitrate concentrations with individual-level birth registry data to create a cohort of 1,078,892 singleton live births in Denmark to Danish-born parents during 1991–2015. Exposures were modeled as log-transformed continuous estimates and categorically. Small-for-gestational-age was defined as the lightest 10% of infants at birth per sex and gestational week. Multivariable logistic regression with generalized estimating equations was used to assess associations and account for the non-independence of siblings. National register-based data were used to adjust for sex, birth year, birth order, and maternal age, smoking, education, income, and employment status at the individual level. RESULTS:In this cohort, the median nitrate concentration was 1.9 mg/L NO₃⁻. An increased risk of small-for-gestational-age was observed when comparing half the European Union standard (25 mg/L NO₃⁻) to 1 mg/L NO₃⁻ [OR: 1.03 (95% CI: 1.01, 1.05)] and in categorical analyses, except for the highest category (25 mg/L NO₃⁻) compared to the referent (2 mg/L NO₃⁻), possibly due to small numbers. CONCLUSIONS:In this nationwide study spanning 25 years, we found that nitrate in maternal household tap water during pregnancy was associated with an increased risk of her child being born small-for-gestational-age. While the increased risk may be small at an individual level, given the widespread presence of nitrate in drinking water it is important on a global scale. When evaluating nitrate regulations, policymakers should focus on health impacts at levels below current standards. KEYWORDS: Big data, Birth outcomes, Children's environmental health, Environmental epidemiology, International collaboration, Water quality

Assessing the Effect of Ultraviolet Radiation, Residential Greenness and Air Pollution on Vitamin D Levels: A Cohort Study in China

BACKGROUND AND AIM: Vitamin D is obtained in the diet or from exposure to ultraviolet rays, and thus can be modified by environment factors. We aim to assess individual and combined effects of ultraviolet (UV) radiation, residential greenness, and air fine particulate matter (PM2.5) with vitamin D levels in a community-based cohort study of elderly aged over 65 years old in China. METHODS: We used the 2012-2014 Chinese Longitudinal Healthy Longevity Survey data for analysis, obtained daily UV radiation from the closest monitoring station and measured residential greenness through satellite-derived Normalized Difference Vegetation Index (NDVI) in a 500 m radius around address. Annual PM2.5 was calculated using satellite-based data (0.1 degrees × 0.1 degrees). Vitamin D levels were defined according to serum 25-hydroxyvitamin D: non-deficiency (≥50 nmol/L) and deficiency (50 nmol/L). We used the generalized estimating equation for longitudinal analysis, adjusting sociodemographic information, eating habits, lifestyle, and season of blood draw. RESULTS:We included 1336 participants, with a mean age of 83 at baseline. In single exposure models, UV radiation and greenness were positively associated with vitamin D non-deficiency, while PM2.5 was inversely associated with it. In the saturated model with three-exposures, each 0.1-MJ/(m2·d) increase in annual UV radiation was associated with a 86% higher odds of vitamin D non-deficiency [odds ratio (OR) 95% CI: 1.50, 2.30], each 0.1-unit increase in annual NDVI was 22% higher odds of vitamin D non-deficiency (OR 95% CI: 1.10, 1.36), while each 10-μg/m3 increase annual PM2.5 was associated a 20% lower odds of vitamin D non-deficiency (OR 95% CI: 0.74,0.85). In our stratified analysis assessing effect modification, the association of UV and vitamin D non-deficiency was strongest in the lowest PM2.5 area and highest greenness area. CONCLUSIONS:UV radiation, residential greenness, and PM2.5 are associated with vitamin D deficiency, and can influence aging pathways and elderly health. KEYWORDS: Air pollution, Green Space, Environmental epidemiology

Intervention efficacy in reducing childhood lower respiratory infections and indoor fine particulate matter among US households with wood heating stoves

BACKGROUND AND AIM: Millions of rural United States (US) households are heated with wood stoves that can lead to high indoor concentrations of fine particulate matter (PM₂.₅). Wood stove use is associated with lower respiratory tract infection (LRTI) in children, the leading cause of childhood hospitalization in the US. In the Kids Air Quality Interventions for Reducing Respiratory Infections (KidsAIR) study, we assessed the impact of low-cost interventions on childhood LRTI and indoor PM₂.₅ in rural homes with wood stoves. METHODS: KidsAIR was a parallel three-arm (education, portable air filtration unit, control), post-only randomized trial in households from Alaska, Montana, and Navajo Nation with a wood stove and one or more children under five years of age. We tracked LRTI cases for two consecutive winter seasons and measured indoor PM₂.₅ over a six-day period during the first winter. We assessed results using two frameworks: intervention efficacy on LRTI and PM₂.₅ (intent-to-treat) and association between PM₂.₅ and LRTI (exposure-response). RESULTS:There were 61 LRTI cases from 14,636 child-weeks of follow-up among 454 children. In the intent-to-treat analysis, children in the education arm (Odds ratio: 1.0; 95% Confidence Interval [95%CI]: 0.4, 2.7) and air filtration arm (Odds ratio: 1.2; 95%CI: 0.5, 3.3) had similar odds of LRTI versus control. Geometric mean PM₂.₅ concentrations were similar to control among education arm (14% higher; 95%CI: -11, 45) and air filtration arm (13% lower; 95%CI: -32, 11) households. In the exposure-response analysis, odds of LRTI were 1.5 times higher (95%CI: 1.0, 2.1) per interquartile range increase in mean indoor PM₂.₅. CONCLUSIONS:We did not observe meaningful differences in LRTI or indoor PM₂.₅ in the air filtration or education arms versus control. The exposure-response analysis provides further evidence that biomass air pollution adversely impacts childhood LRTI. Our results highlight the need for novel, effective intervention strategies in households heated with wood stoves. KEYWORDS: Children's environmental health, Respiratory outcomes, Environmental epidemiology

Exposure to phthalates in relation to sleep duration and social jetlag among adolescent boys and girls in Mexico City

BACKGROUND AND AIM: Emerging research has demonstrated the importance of environmental toxicants, such as phthalates, for sleep health, as they may disrupt neuronal sleep regulation circuitry. However, the impact of high-molecular-weight (HMW) versus low-molecular-weight (LMW) phthalates remains unknown. Among Mexican adolescents, we evaluated the association between exposure to HMW and LMW phthalates with sleep duration and social jetlag (a circadian marker that measures the discrepancy between weekend and weekday sleep duration). METHODS: Participants included 413 adolescents from the Early Life Exposures in Mexico to ENvironmental Toxicants (ELEMENT) study. In 2015, urinary concentrations of 13 phthalates were measured from spot urine samples, and sleep duration was assessed with wrist-actigraph devices over 7consecutive days. We examined associations between phthalates with mean weekday and weekend sleep duration and social jetlag using linear regression models, adjusted for specific-gravity, sex, age, BMI z-score, total energy intake, caffeine intake, and physical activity. RESULTS:The mean (SD) age was 15.6 (2.0) years; 53.5% were girls. The mean (SD) sleep duration was 8.4 (1.2) hr/night on weekdays and 9.0 (1.3) hr/night on weekends. In adjusted models, each IQR increase in the HMW phthalate, mono-(2-ethyl-5-oxohexyl) phthalate was associated with a 13.3 minute (95%CI: -26.3, -0.41; p 0.05) lower social jetlag, and each IQR increase in mono-benzyl phthalate was associated with an 8.0 minute (95% CI: -16.8, -0.72; p 0.05) shorter sleep duration on weekends. Among LMW phthalates, we found that each IQR increase in monoethyl phthalate was associated with an 8.6 minute (95% CI: -17.9, -0.63; p 0.05) shorter sleep duration on weekends and a 15.3 minute (95% CI: -29.0, -1.6; p 0.05) lower social jetlag. CONCLUSIONS:In contrast to other literature, higher phthalate exposure was related to shorter weekend sleep duration and lower social jetlag among Mexican adolescents, sleep characteristics that may be indicative of better sleep health. KEYWORDS: phthalates; environmental epidemiology; adolescent sleep health; social jetlag, sleep duration, wrist-actigraphy

Association between outdoor traffic air pollutants and spread of SARS-CoV-2 pandemic in Modena, Northern Italy

BACKGROUND AND AIM: Italy was the first European country severely affected by the SARS-CoV-2 pandemic, especially during the first wave in the North of the country. In particular, Modena is a city of Emilia-Romagna which is a region entirely included in the Po valley, one of European areas characterized by the heaviest air pollution levels due also to its orography. Previous studies yielded at a regional level suggested that higher air pollutant levels may increase both SARS-CoV-2 infection and mortality. METHODS: In this study, we further investigated the association between air pollutant exposure and spread of the SARS-CoV-2 using data collected from Modena municipality in the period February 2020-April 2021. We used traffic pollutant levels collected from an urban air quality monitoring station in the period January 2020-January 2021, including particulate matter (PM10), nitrogen oxides (NO2 and NOx), benzene, and black carbon (BC). We used a random-effects linear regression model within panel data analysis over the study period and we computed beta correlation coefficients (beta) with 95% confidence interval-CI between mean daily pollutant concentrations and SARS-CoV-2 daily positive cases diagnosed in Modena. RESULTS:We found a positive correlation between all traffic pollutants and SARS-CoV-2 cases, namely for PM10 beta was 1.23 (95%CI 1.00-1.46), lower for NOx (beta=0.66, 95%CI 0.56-0.75), and higher for NO2 (beta=1.95, 95%CI 1.59-2.31), benzene (beta=41.41, 95%CI 36.23-46.59), and BC (beta=5.95, 95%CI 1.19-10.72). CONCLUSIONS:Notwithstanding the limitations of use of aggregated data and potential the residual confounding, these findings seem to support the hypothesis that high levels of air pollution may favor the spread of SARS-CoV-2 infection, or alternatively that they reflect a higher mobility and number of social contacts that favor the spread of the infection. KEYWORDS: Air pollution, Traffic-related, Infectious diseases, Environmental epidemiology, Exposure assessment

Having your cake (mix) and eating it too: Independent, interaction, and group effects of mixtures using Bayesian Hierarchical Regression Modelling

BACKGROUND AND AIM: Mixtures analysis methods are increasingly being applied in environmental epidemiology, but current methods are often limited by either only providing a group effect or independent exposure effects without a group effect. We investigate Bayesian Hierarchical Regression Modelling (BHRM) and show how it can be adapted to handle highly correlated exposures to estimate: 1) independent exposure effects; 2) interactions between exposures; and 3) combined effects for a mixture exposure. METHODS: BHRM is a flexible framework that can provide robust estimation for highly correlated exposures (via g-prior specification), yield conditional exposure-specific estimates, and include interactions effects. We demonstrate how these general regression models can provide additional inference on the combined effect for a multi-pollutant exposure. To demonstrate potential advantages of certain specifications, we applied BHRM to an analysis of liver injury and exposure to perfluoroalkyl substances (PFAS), including PFOS, PFOA, PFHxS, PFNA, and PFUnDA, in 1105 children from the Human Early Life Exposome (HELIX) project. Liver injury was defined as 90th percentile for any serum liver injury biomarker (ALT, AST, GGT). We used BHRM to estimate the mixture effect and pollutant-specific effects. For comparison, we also applied Bayesian Weighted Quantile Sum regression (BWQS)—an alternative specification within Bayesian Hierarchical Modeling. RESULTS:PFAS mixture was associated with childhood liver injury: OR=1.64 (95%CI:1.34-2.00) per increased exposure equivalent to one standard deviation for all PFAS. Within this mixture effect, PFNA was the predominant exposure driving the association with OR=1.62 (95%CI:1.11-1.97) and a posterior inclusion probability of 0.969 (Bayes factor, BF=125). No strong evidence of interactions. Although BWQS estimated a mixture effect of OR=1.85 (95%CI:1.50-2.25) and indicated PFNA had the most substantial estimated weight of 0.513 (BF=4.21), the approach lacks pollutant-specific effects. CONCLUSIONS:BHRM is an efficient method for mixtures analysis. Estimation of pollutant-specific effects with group effects provide critical data for identifying causal agents in mixtures of environmental contaminants. KEYWORDS: mixtures, methods, Bayesian Hierarchical Regression Modelling, multi-pollutant

A Prospective Study of Arsenic and Manganese Exposure and Maternal Blood Pressure During Gestation

BACKGROUND AND AIM: Pregnancy is a sensitive time for maternal cardiovascular functioning. This study examines the associations between arsenic and manganese exposures to maternal blood pressure (BP) measured during pregnancy. METHODS: Pregnant women (N=1,522) were recruited for a prospective cohort in Bangladesh (2008-2011). Drinking water arsenic and manganese were measured at 16 weeks gestation and toenail measurements were made one-month post-partum. BP (systolic and diastolic) levels were measured during monthly home visits. Linear mixed models estimated the associations between repeated BP measurements, arsenic, and manganese. RESULTS:Drinking water and toenail arsenic showed a positive dose-response association with maternal BP. Each natural log increase in drinking water arsenic was associated with 1.27 mmHg systolic (95% CI: 1.20 – 1.34) and 1.16 mmHg diastolic (95% CI: 1.11-1.20) increase in BP per month of gestation after adjusting for covariates. Each natural log increase in toenail arsenic levels was associated with a 1.28 mmHg systolic (95% CI: 1.15 – 1.44) and 1.14 mmHg diastolic (95% CI: 1.05 – 1.23) increase in BP per month of gestation after adjusting for covariates. Drinking water manganese exposure had a non-linear association with maternal BP. Drinking water manganese in the 2nd quartile (160-589 µg/L) was associated with a 1.70 mmHg increase in systolic (95% CI: 1.21-2.38) and 1.39 mmHg in diastolic (95% CI: 1.09-1.76) BP per month of gestation after adjusting for covariates. Each natural log increase of toenail manganese was associated with a 1.21 mmHg of systolic and 1.54 mmHg of diastolic increase in BP over gestation (systolic 95% CI: 1.06-1.37, diastolic 95% CI: 1.15-2.05), after adjusting for covariates. CONCLUSIONS:Arsenic and manganese exposure during pregnancy was associated with increases in maternal systolic and diastolic BP. Elevated gestational BP can increase risks of adverse birth outcomes and CVD later in life for the mother. KEYWORDS: Cardiovascular diseases, Pregnancy outcomes, Water quality, Heavy metals, Environmental Epidemiology

Birth Outcomes Associated with Paternal Polybrominated and Polychlorinated Biphenyl Exposure

BACKGROUND AND AIM: In 1973, polybrominated biphenyl (PBB) flame retardant was shipped to Michigan feed mills in place of a nutritional supplement and contaminated the food supply. Following the accident, the Michigan PBB Registry was established to study the long-term health effects of halogenated compounds, and is now led by a community-academic partnership. PBB exposure is associated with decreased DNA methylation in sperm regulatory elements controlling imprinted genes, which may lead to adverse birth outcomes in children whose fathers have increased levels of serum PBB or polychlorinated biphenyl (PCB), an analogue of PBB. To further understand the impact of paternal exposures on children, we analyzed paternal PBB and PCB levels and premature birth and low birthweight. METHODS: We investigated the outcomes of birthweight and gestational age of offspring whose fathers enrolled in the PBB Registry and provided serum to measure PBB and PCB concentrations. Birth outcomes were determined by linking paternal PBB records to the Michigan vital records. Birthweight and gestational age were dichotomized at the 25th percentile and 37 weeks respectively, and paternal PBB and PCB levels were divided into tertiles. Offspring birthweight and gestational age were modeled against paternal PBB and PCB serum concentrations using log-risk regression, adjusting for clustering, paternal health and lifestyle factors, and offspring sex. RESULTS:Elevated PBB and PCB levels are associated with an increased risk of low offspring birthweight, but not preterm birth. Fathers in the middle and upper PBB and PCB tertiles illustrated increased risk ratios (RR) for low offspring birthweight, RR=1.42 (CI=0.93-2.17) and RR=1.47 (CI=1.03-2.10) for PBB, and RR=1.73 (CI=1.12-2.67) and RR=1.42 (CI=0.84-2.41) for PCB respectively. Elevated PBB and PCB levels illustrated no significant increased risk for preterm birth in the middle and upper tertiles. CONCLUSIONS:These findings suggest increased paternal PBB and PCB levels may negatively impact offspring birthweight. KEYWORDS: Endocrine disrupting chemicals, Birth outcomes, Children's environmental health, Community outreach, Environmental epidemiology

Drinking Water Sources and Water Quality in the Agricultural Health Study

BACKGROUND AND AIM: Drinking water contaminants have been associated with adverse health effects. However, little is known about exposures in agricultural communities where farm practices and well use may enhance exposures to specific contaminants. We describe the exposure assessment and exposure distributions for common drinking water contaminants by water source in an agricultural population. METHODS: We used questionnaire data from the Agricultural Health Study (AHS), a cohort of licensed pesticide applicators (mostly farmers) and their spouses in Iowa (IA) and North Carolina (NC) to ascertain drinking water source at enrollment (1993-1997). For users of public water supplies (PWS), we linked participants’ geocoded addresses to monitoring data for regulated contaminants, including five haloacetic acids (HAA5), total trihalomethanes (TTHM), and nitrate-nitrogen (NO₃-N). For private well users, we estimated nitrate levels using previously developed random forest models based on well depth, land use, soil characteristics, nitrogen inputs, and other relevant factors. RESULTS:We assigned drinking water source for 84% (N=74,919) of AHS participants; approximately 70% in IA and 75% in NC used private wells. Median nitrate concentrations (mg/L NO₃-N) in PWS were higher in IA (0.8, interquartile range (IQR): 0.2-3.1) than NC (0.5, IQR: 0.5-0.6), while median concentrations of HAA5 (µg/L) and TTHM (µg/L) were higher in NC (HAA5: 12.1, IQR: 3.8-33.4; TTHM: 35.4, IQR: 8.8-54.7) than IA (HAA5: 4.3, IQR: 3.0-10.5; TTHM: 13.0, IQR: 3.7-32.3). Private well nitrate concentrations in IA (1.5, IQR: 0.8-4.9) and NC (1.9, IQR: 1.4-2.5) were higher than in PWS. A larger proportion of private wells exceeded the 10mg/L regulatory limit in IA (12%) than NC (1%). CONCLUSIONS:Drinking water nitrate levels were higher for private well users and in IA PWS, while TTHM and HAA5 levels were higher in NC PWS. This work will facilitate future analyses of the health effects of drinking water contaminants in the AHS. KEYWORDS: Water quality, Exposure assessment, Environmental epidemiology