Environmental Epidemiology Research Research Articles

Identifying the spatiotemporal dynamics of PM2.5 concentration and its implications for national sustainable development experimental zone of China

The accelerated pace of urbanization and industrialization in China has given rise to air pollution in the form of PM2.5. This pollution poses significant threats to the atmospheric environment and human health. Traditional statistical models often lack the required data precision for medium or small-scale epidemiological and pollutant exposure studies. Therefore, alternative approaches must be developed. This study employs the Random Forest (RF) model, utilizing measured PM2.5 concentration and auxiliary data, to simulate PM2.5 concentration with 1 km spatial resolution for 2015–2019. The results showed that: (1) The RF model was sufficiently accurate with a 10-fold CV, resulting in a coefficient of determination (R2) of 0.75, a root-mean-square error (RMSE) measuring at 13.24 μg/m3, and a mean absolute error (MAE) of 9.12 μg/m3 (2) A notable variation and dynamic pattern in the concentration of PM2.5 were observed. The geographical distribution displayed elevated levels in the northern regions and subdued levels in the southern regions, with the most elevated PM2.5 values recorded in Xinjiang and Northern China. (3) Pollution levels in the five major urban agglomerations, ranked from high to low, were as follows: the Beijing-Tianjin-Hebei (BTH), the Guanzhong Plain (GZP), the Yangtze River Delta (YRD), the Chengdu-Chongqing (CY), and the Pearl River Delta (PRD). (4) The PM2.5 concentration of the whole country generally showed a downward trend. These findings offer valuable scientific insights to support atmospheric environmental protection and epidemiological research endeavors. Finally, policy implications for the coordinated development of the economy and environment in the national sustainable development experimental zone major urban agglomerations were proposed to achieve more equitable and balanced development. This paper provides policy recommendations and empirical evidence for further promoting environmental balance in national sustainable development experimental zones.

Risk of papillary thyroid carcinoma and nodular goiter associated with exposure to semi-volatile organic compounds: A multi-pollutant assessment based on machine learning algorithms

BackgroundExposure to semi-volatile organic compounds (SVOCs) may link to thyroid nodule risk, but studies of mixed-SVOCs exposure effects are lacking. Traditional analytical methods are inadequate for dealing with mixed exposures, while machine learning (ML) seems to be a good way to fill the gaps in the field of environmental epidemiology research. ObjectivesDifferent ML algorithms were used to explore the relationship between mixed-SVOCs exposure and thyroid nodule. MethodsA 1:1:1 age- and gender-matched case-control study was conducted in which 96 serum SVOCs were measured in 50 papillary thyroid carcinoma (PTC), 50 nodular goiters (NG), and 50 controls. Different ML techniques such as Random Forest, AdaBoost were selected based on their predictive power, and variables were selected based on their weights in the models. Weighted quantile sum (WQS) regression and Bayesian kernel machine regression (BKMR) were used to assess the mixed effects of the SVOCs exposure on thyroid nodule. ResultsForty-three of 96 SVOCs with detection rate >80 % were included in the analysis. ML algorithms showed a consistent selection of SVOCs associated with thyroid nodule. Fluazifop-butyl and fenpropathrin are positively associated with PTC and NG in single compound models (all P < 0.05). WQS model shows that exposure to mixed-SVOCs was associated with an increased risk of PTC and NG, with the mixture dominated by fenpropathrin, followed by fluazifop-butyl and propham. In the BKMR model, mixtures showed a significant positive association with thyroid nodule risk at high exposure levels, and fluazifop-butyl showed positive effects associated with PTC and NG. ConclusionThis study confirms the feasibility of ML methods for variable selection in high-dimensional complex data and showed that mixed exposure to SVOCs was associated with increased risk of PTC and NG. The observed association was primarily driven by fluazifop-butyl and fenpropathrin. The findings warranted further investigation.

Long-term spatiotemporal variations in surface NO2 for Beijing reconstructed from surface data and satellite retrievals

Remote sensing data from the Ozone Monitoring Instrument (OMI) and the TROPOspheric Monitoring Instrument (TROPOMI) play important roles in estimating surface nitrogen dioxide (NO2), but few studies have compared their differences for application in surface NO2 reconstruction. This study aims to explore the effectiveness of incorporating the tropospheric NO2 vertical column density (VCD) from OMI and TROPOMI (hereafter referred to as OMI and TROPOMI, respectively, for conciseness) for deriving surface NO2 and to apply the resulting data to revisit the spatiotemporal variations in surface NO2 for Beijing over the 2005–2020 period during which there were significant reductions in nitrogen oxide emissions. In the OMI versus TROPOMI performance comparison, the cross-validation R2 values were 0.73 and 0.72, respectively, at 1 km resolution and 0.69 for both at 100 m resolution. The comparisons between satellite data sources indicate that even though TROPOMI has a finer resolution it does not improve upon OMI for deriving surface NO2 at 1 km resolution, especially for analyzing long-term trends. In light of the comparison results, we used a hybrid approach based on machine learning to derive the spatiotemporal distribution of surface NO2 during 2005–2020 based on OMI. We had novel, independent passive sampling data collected weekly from July to September of 2008 for hindcasting validation and found a spatiotemporal R2 of 0.46 (RMSE = 7.0 ppb). Regarding the long-term trend of surface NO2, the level in 2008 was obviously lower than that in 2007 and 2009, as expected, which was attributed to pollution restrictions during the Olympic Games. The NO2 level started to steadily decline from 2015 and fell below 2008's level after 2017. Based on OMI, a long-term and fine-resolution surface NO2 dataset was developed for Beijing to support future environmental management questions and epidemiological research.

Spatiotemporal variability and measurement uncertainty in atmospheric polycyclic aromatic hydrocarbons (PAHs): A quantitative variance component analysis

A comprehensive understanding of the spatial and temporal variability in air pollution facilitates air monitoring designs, exposure assessment, and environmental epidemiology research. Polycyclic aromatic hydrocarbons (PAHs) are ubiquitous in the air but are challenging to measure due to their low concentrations. To date, no study has been specifically designed to quantify the spatial and temporal aspects of variability in ambient PAH concentrations. This study aims to decompose the variability in PAH concentrations measured in the ambient air in a metropolitan area into different components. Air samples were collected using high-volume samplers at 19 stations in the Memphis Tri-state Area across Tennessee, Arkansas, and Mississippi, USA, every 12 days from March 2018 to May 2019. Samples were collected on polyurethane forms and quartz fiber filters and analyzed using gas chromatography/mass spectrometry (GC/MS) for 32 PAHs. Variance components of the total variance were computed using mixed models for individual PAHs and sum concentrations of the 16 priority PAHs (Σ16PAHs). For Σ16PAHs, spatial and temporal variability represented 16% and 84% of the total variance. The total variance could be further apportioned to between-urbanicity (0.5%), between-site (5.5%), between-season (41%), and within-season (53%) variances. Sampling and analytical uncertainties represented 39% and 0.01% of the total variance, when monitoring uncertainties were considered. Concentrations of individual PAHs followed similar variance structures. These findings confirmed the validity of centralized monitoring stations to represent a large area, suggested the need for repeated samples across seasons and at multiple time points within a season, and warranted improving quality control for field sample collections.

Decoupling race/ethnicity from pediatric lung function metrics to investigate potential misclassification bias in environmental epidemiologic research

Background and aim: Electronic health records (EHR) are increasingly being used in environmental epidemiologic research. Respiratory studies of asthma commonly use derived spirometry metrics, such as percent predicted forced expiratory volume in one second (FEV1%predicted), that are calculated from age, height, sex, and – controversially – race/ethnicity. We examine potential misclassification bias introduced by historically race/ethnicity-adjusted pediatric lung function metrics. Methods: Spirometry results for children with asthma ages 5-17 years were obtained from a clinical EHR data repository at Boston Medical Center (BMC) from 2003 to 2019. BMC is New England’s largest safety net hospital, serving a racio-ethnically diverse pediatric population. Recorded lung function metrics included FEV1 (liters), forced vital capacity (liters), and previously calculated FEV1%predicted (%) that adjusted for race/ethnicity. We recalculated FEV1%predicted values by standardizing to a global, multiethnic population using the Global Lung Initiative equations and compared to original FEV1%predicted from the EHR data repository. Results: The EHR included 1141 children that had 2651 spirometry measurements. Mean FEV1%predicted was 91.6% from the EHR and 87.2% when recalculated. Original FEV1%predicted values were higher compared to recalculated values for Black children (mean=89.3% [standard error=0.56] vs. 82.5% [0.42] , n=1526, p&amp;#x3c;0.01), but not for Asian (93.8% [3.72] vs. 92.0% [3.59], n=32, p=0.72), Hispanic (96.5% [1.65] vs. 94.4% [1.56], n=129, p=0.37), Other (93.8% [3.55] vs. 91.3% [2.52], n=47, p=0.57), Not Recorded (94.7% [0.73] vs. 94.1% [0.66], n=720, p=0.25), and White (94.1% [1.72] vs. 94.1% [1.42], n=195, p=0.97). Conclusions: Race/ethnicity-adjustment of a commonly used &amp;#x22;objective” lung function metric in EHR may artificially improve lung function test results, particularly for Black children with asthma, resulting in a potential source of misclassification bias in environmental epidemiologic research. Continued use of historically race/ethnicity-adjusted lung function metrics warrants critical reevaluation, especially in research aimed at disentangling the environmental drivers of health disparities driven by structural racism.

Comparing greenspace metrics and buffer methods in Metro Vancouver, Canada: Implications for environmental health research

BACKGROUND &amp;#x26; AIM: Access and exposure to greenspace has been shown to promote mental and physical health; however, not all studies have found these benefits, with some studies reporting no association and others reporting associations with adverse health outcomes. It has been suggested that the use of diverse greenspace metrics, the varied quality of these metrics, and the specific methods used to quantify exposure or access, may contribute to these inconsistent findings. This project sought to compare the results of using different metrics in Metro Vancouver, Canada. METHODS: Greenspace metrics were calculated from the centroid of each 6-digit postal code. Average buffer normalized differences vegetation index (NDVI), green landcover, tree canopy cover, and park count were calculated using circular and network buffers, at both 400m and 1000m. Pearson’s correlation coefficients were calculated to compare the results for each metric and method. RESULTS: The number of parks was only weakly positively correlated with each of the other three greenspace metrics for 1000m circular buffers (NDVI r = 0.12; green land cover r = 0.07; tree canopy r = 0.14, p &amp;#x3c; .01). For 1000m network buffers (NDVI r = -0.34; green landcover r = -0.12; tree canopy r = 0.05, p &amp;#x3c; .01) two of the correlations became negative. Circular and network buffer were most highly correlated for the tree canopy (1000m r = 0.92; 400m r = 0.86, p &amp;#x3c; .01), whereas park count had the weakest correlation (1000m r = 0.66; 400m r = 0.66, p &amp;#x3c; .01). CONCLUSIONS: The associations between greenspace metrics vary. Selection of buffering method may be particularly influential when using park count data. These findings support the importance of intentional selection of greenspace metric and measurement method in environmental epidemiological research. KEYWORDS: greenspace metrics, NDVI, tree canopy cover, parks, network buffers, and circular buffers

Strengthening the evidence base on global air pollution epidemiology: the case for collaboration and engagement

Increasing interest in expanding the scope of environmental epidemiology to more adequately reflect the diversity of global contexts is converging with calls for reforming global health. This presentation will examine some of the drivers of these converging trends and offer reflections on the key components of sustainable, mutually beneficial research collaborations to support global air pollution epidemiology. I will discuss the important role of capacity building at the organizational, project, and individual-level for strengthening the evidence base on air pollution epidemiology globally, drawing from the experience of the ISEE Capacity Building and Education Committee (CAPE) and specific projects including the Cardiovascular health effects of Air pollution in Telangana, India (CHAI) project. I will provide an overview of the guiding principles applied in the CHAI project to build capacity of new researchers in India, and increase their participation in international publications and professional networks. I will reflect on current barriers to publication of environmental exposure science and epidemiology research from low- and middle-income country (LMIC) authors in journals. Finally, I will provide an overview of innovative strategies to facilitate publication of high quality LMIC-led research, including as an example, the CAPE Peer Editing Program.

The Nigerian Environmental Epidemiology Accelerated Research Program (NEEAR): Evaluation of an Environmental Epidemiology Short Course for Nigerian Researchers

BACKGROUND AND AIM The Nigerian Environmental Epidemiology Accelerated Research (NEEAR) Program seeks to promote collaborative capacity and research development in environmental epidemiology in Nigeria. In 2021, a short course on environmental epidemiology was developed to strengthen the ability of Nigerian researchers to conduct environmental epidemiology research and thereby, to fill gaps in research capacity. METHODS The course was extensively advertised through social media and professional networks. Eligible applicants included Nigerian graduate students and scientists with degrees in Environment and Health related disciplines. Training was weekly 2-hour online sessions for 2 weeks and a 1-day in-person session in Nigeria. It included tutorial, videos, group work and hands-on exercises with faculty from Emory University USA and the Nigerian Institute of Medical Research. Session topics included study design, exposure assessment, data analysis, research ethics, and dissemination. A career talk and transferable skills sessions were also included. Pre- and post-session quizzes were administered during each session to assess whether the training improved participants’ understanding of the session topics. RESULTS A total of 240 Nigerian researchers applied to the program, of which 35 (15%) top applicants were invited. Short course participants were 47% female and 53% male, with 87% possessing postgraduate qualifications. The participants came from 8 states within Nigeria. There was a marked improvement in participants post-session quiz scores compared to scores from the beginning of the sessions. Participants feedback was very positive. There was clarity on how the training will be applied in their current studies or work. CONCLUSIONS The short course was impactful and fostered collaborative research capacity among Nigerian researchers. There is potential for more of such courses to be held due to the large number of eligible applications received. Future courses on intermediate/advanced epidemiology and biostatistics methods, internships and post-training connects are recommended. KEYWORDS Capacity development, Nigeria, collaborative research

Mapping Seasonal High-Resolution PM2.5 Concentrations with Spatiotemporal Bagged-Tree Model across China

High concentrations of fine particulate matter (PM2.5) are well known to reduce environmental quality, visibility, atmospheric radiation, and damage the human respiratory system. Satellite-based aerosol retrievals are widely used to estimate surface PM2.5 levels because satellite remote sensing can break through the spatial limitations caused by sparse observation stations. In this work, a spatiotemporal weighted bagged-tree remote sensing (STBT) model that simultaneously considers the effects of aerosol optical depth, meteorological parameters, and topographic factors was proposed to map PM2.5 concentrations across China that occurred in 2018. The proposed model shows superior performance with the determination coefficient (R2) of 0.84, mean-absolute error (MAE) of 8.77 μg/m3 and root-mean-squared error (RMSE) of 15.14 μg/m3 when compared with the traditional multiple linear regression (R2 = 0.38, MAE = 18.15 μg/m3, RMSE = 29.06 μg/m3) and linear mixed-effect (R2 = 0.52, MAE = 15.43 μg/m3, RMSE = 25.41 μg/m3) models by the 10-fold cross-validation method. The results collectively demonstrate the superiority of the STBT model to other models for PM2.5 concentration monitoring. Thus, this method may provide important data support for atmospheric environmental monitoring and epidemiological research.

New directions in global environmental health

BACKGROUND AND AIM: The COVID-19 pandemic is challenging the world’s economic and health systems and exemplifies the degree of global interdependencies, the problems in the governance of global crises and the need of preparedness for global health threats. The outbreak of COVID-19 highlighted the links between the occurrence of new infections, the environment and climate at a global scale and the need for a planetary health perspective and trans-disciplinary approaches. The pandemic brought forward considerable health inequities globally METHODS: The new post-pandemic global settings bring forward overarching topics that have not been efficiently addressed: research needs for policy development, locally and globally; development of transdisciplinary research; implementation science for long-standing global problems; global ethics; new approaches in global public health governance. RESULTS:The pandemic has profoundly affected global society in multiple ways, introducing entirely new challenges and accelerating ongoing trends. The pandemic has highlighted the deep interconnections between problems typically studied in isolation or that previously were situated at the periphery of environmental epidemiology. There is a need to develop research and action on multiple levels. For example, air pollution, chemical pollution, or nature-based solutions must be addressed locally in cities. We should also address them at a global level: global air-pollution in relation to decarbonisation, the effects of global pollution and loss of biodiversity on ecosystems and human health. LMICs are disproportionately impacted by environmental pollution because of the lack of environmental policies and export of contaminants from high-income countries. The urgency of responding to the pandemic brought into prominence the need for international collaboration to develop actionable knowledge to emergencies that will require research in understudied populations and understudied thematic areas. CONCLUSIONS:The dramatic changes occurring in a short time will affect research in environmental epidemiology in the long term and the role of environmental epidemiology on a global scale. KEYWORDS: global epidemiology, covid-19, international collaboration

Scientist Civic Engagement and the US Ozone Standards in 2020

BACKGROUND AND AIM: In 2020, the United States Environmental Protection Agency (EPA) proposed to retain, without revision, its national primary and secondary ozone National Ambient Air Quality Standards (NAAQS). As exposure scientists and environmental epidemiologists, we recognized that this decision may not have reflected the most updated research on ozone and population health effects. We thus aimed to inform the EPA of the state-of-the-science through official channels. METHODS: Within the North American Chapter (NAC) of the International Society for Environmental Epidemiology (ISEE), the Policy Committee worked as a team to gather, synthesize, and communicate evidence on ozone effects on population health. For the ozone NAAQS, one member attended the oral comment sessions and led the effort to draft and submit a written comment to the EPA. RESULTS:At the ISEE NAC Policy Committee, we recognized a proposed ruling that would have affected air quality in the US. We effectively organized to engage government stakeholders and communicate the latest research on ozone and population health. These actions entered into the public record a statement aimed at a general audience on the effects of ozone along with supporting scientific articles. Such materials could be considered in a court of law if and when a decision on the ozone NAAQS is contested. CONCLUSIONS:Environmental exposure and epidemiology research have the potential to inform policy decisions. The ISEE NAC Policy Committee provides an avenue for scientists to gain experience in communicating the research implications to decision-makers. KEYWORDS: Ozone, Research translation, Science communication, Policy

Changes in environmental epidemiology research methods due to COVID-19

BACKGROUND AND AIM: The obligatory pivot to digital research methods during the pandemic has illustrated its advantages in terms of flexibility, efficiency, broader geographic coverage, and lower barriers to participation. Digital research methods are likely to continue to expand in the post-COVID-19 era, although not without challenges and tradeoffs. The digital divide will become an increasingly complex source of selection bias, with important implications for participation of older individuals or those with lower socioeconomic position. For example, online questionnaires used heavily in the early stages of the pandemic, often had limited researcher control over who participated and little information on potentially large selection biases. METHODS: The pandemic has highlighted the potential of big data for public health protection. High-resolution geo-located data, increasingly available through either top-down (i.e. government surveillance) or bottom-up (i.e. volunteered geographic information) sources, paired with spatial analysis, are providing unprecedented opportunities to monitor complex phenomena such as human mobility and interactions. COVID-19 is proving an important impetus for new methods development to deal with complex spatio-temporal data structures with promising applications in environmental epidemiology. RESULTS:Research methods during COVID-19 shifted heavily towards remote biological sample collection, influencing the type and volume of samples collected. This shift is pushing development of improved analytical techniques for small volume samples. While remote collection offers several advantages over laboratory tests that could reduce selection bias (e.g. increased convenience and accessibility for participants living far from a lab), there are important challenges in ensuring data quality during the collection and handling of samples. CONCLUSIONS:In conclusion, the application of methods for conducting research that were a necessity during the COVID-19 pandemic because of the risk of infection can be expected to accelerate. The pandemic has triggered the development of creative, flexible research methods that offer advantages outside of the pandemic context; however, careful consideration of drawbacks is essential. KEYWORDS: COVID-19, big data, environmental epidemiology

Measuring the Outcomes of a Participatory Research Study: Findings from an Environmental Epidemiological Study in Kaunas City

The achievement of a sustainable urban environment and health for all requires the engagement and greater awareness of local communities on issues of environment and health. This HORIZON2020 CitieS-Health study presents the outcomes of the environmental epidemiological research on the participants’ acquisition of new skills and knowledge as well as on health behaviour. We conducted a cross-sectional study of 1062 residents of Kaunas city, Lithuania, from 2019 to 2021. We analysed the associations between the neighbourhood environmental quality scores and health issues, and the self-reported ratings on the acquired knowledge measured using a Likert rating scale. About 42.7% of the participants acknowledged that participation in the research study improved their data collection and interpretation skills, and 58.8% of them stated that the participation improved their knowledge on the links between environmental quality and health. The participants with increased knowledge more often rated their health as “good”, had a significantly lower diastolic blood pressure, and regularly visited the natural environment. The high impact of participatory research was associated with a higher scoring of the neighbourhood environmental quality, higher physical activity, and a beneficial effect on health. The study provides scientific evidence that improving the neighbourhood environment would promote increased physical activity, such as reaching green spaces by walking, and might benefit the society.

A robust approach to deriving long-term daily surface NO2 levels across China: Correction to substantial estimation bias in back-extrapolation

BackgroundLong-term surface NO2 data are essential for retrospective policy evaluation and chronic human exposure assessment. In the absence of NO2 observations for Mainland China before 2013, training a model with 2013–2018 data to make predictions for 2005–2012 (back-extrapolation) could cause substantial estimation bias due to concept drift. ObjectiveThis study aims to correct the estimation bias in order to reconstruct the spatiotemporal distribution of daily surface NO2 levels across China during 2005–2018. MethodsOn the basis of ground- and satellite-based data, we proposed the robust back-extrapolation with a random forest (RBE-RF) to simulate the surface NO2 through intermediate modeling of the scaling factors. For comparison purposes, we also employed a random forest (Base-RF), as a representative of the commonly used approach, to directly model the surface NO2 levels. ResultsThe validation against Taiwan’s NO2 observations during 2005–2012 showed that RBE-RF adequately corrected the substantial underestimation by Base-RF. The RMSE decreased from 10.1 to 8.2 µg/m3, 7.1 to 4.3 µg/m3, and 6.1 to 2.9 µg/m3 in predicting daily, monthly, and annual levels, respectively. For North China with the most severe pollution, the population-weighted NO2 ([NO2]pw) during 2005–2012 was estimated as 40.2 and 50.9 µg/m3 by Base-RF and RBE-RF, respectively, i.e., 21.0% difference. While both models predicted that the national annual [NO2]pw increased during 2005–2011 and then decreased, the interannual trends were underestimated by >50.2% by Base-RF relative to RBE-RF. During 2005–2018, the nationwide population that lived in the areas with NO2 > 40 µg/m3 were estimated as 259 and 460 million by Base-RF and RBE-RF, respectively. ConclusionWith RBE-RF, we corrected the estimation bias in back-extrapolation and obtained a full-coverage dataset of daily surface NO2 across China during 2005–2018, which is valuable for environmental management and epidemiological research.

Centralizing environmental datasets to support (inter)national chronic disease research: Values, challenges, and recommendations.

Background:Whereas environmental data are increasingly available, it is often not clear how or if datasets are available for health research. Exposure metrics are typically developed for specific research initiatives using disparate exposure assessment methods and no mechanisms are put in place for centralizing, archiving, or distributing environmental datasets. In parallel, potentially vast amounts of environmental data are emerging due to new technologies such as high resolution imagery and machine learning.Objectives:The Canadian Urban Environmental Health Research Consortium (CANUE) and the Geoscience and Health Cohort Consortium (GECCO) provide a proof of concept that centralizing and disseminating environmental data for health research is valuable and can accelerate discovery. In this essay, we argue that more efficient use of exposure data for environmental epidemiological research over the next decade requires progress in four key areas: metadata and data access portals, linkage with health databases, harmonization of exposure measures and models over large areas, and leveraging “big data” streams for exposure characterization and evaluation of temporal changes.Discussion:Optimizing the use of existing environmental data and exploiting emerging data streams can provide unprecedented research opportunities in environmental epidemiology through a better characterization of individuals’ exposures and the ability to study the intersecting impacts of multiple environmental features or urban attributes across different populations around the world. Proper documentation, linkage, and dissemination of new and emerging exposure data leads to a better awareness of data availability, a reduction of duplication of effort and increases research output.

Geospatial analysis of individual-based Parkinson's disease data supports a link with air pollution: A case-control study

BackgroundThe etiology of Parkinson's disease (PD) remains unknown. To approach the issue of PD's risk factors from a new perspective, we hypothesized that coupling the geographic distribution of PD with spatial statistics may provide new insights into environmental epidemiology research. The aim of this case-control study was to examine the spatial dependence of PD prevalence in the Canton of Geneva, Switzerland (population = 474,211). MethodsPD cases were identified through Geneva University Hospitals, private neurologists and nursing homes medical records (n = 1115). Controls derived from a population-based study (n = 12,614) and a comprehensive population census dataset (n = 237,771). All individuals were geographically localized based on their place of residence. Spatial Getis-Ord Gi* statistics were used to identify clusters of high versus low disease prevalence. Confounder-adjustment was performed for age, sex, nationality and income. Tukey's honestly significant difference was used to determine whether nitrogen dioxide and particulate matters PM10 concentrations were different within PD hotspots, coldspots or neutral areas. ResultsConfounder-adjustment greatly reduced greatly the spatial association. Characteristics of the geographic space influenced PD prevalence in 6% of patients. PD hotspots were concentrated in the urban centre. There was a significant difference in mean annual nitrogen dioxide and PM10 levels (+3.6 μg/m3 [p < 0.001] and +0.63 μg/m3 [p < 0.001] respectively) between PD hotspots and coldspots. ConclusionPD prevalence exhibited a spatial dependence for a small but significant proportion of patients. A positive association was detected between PD clusters and air pollution. Our data emphasize the multifactorial nature of PD and support a link between PD and air pollution.

Capacity development in environmental epidemiology research in Africa: experience from ISEE sponsored short course in Nigeria

A short course on environmental epidemiology and exposure assessment sponsored by ISEE Capacity and Education committee in 2019 was aimed to strengthen the ability of African researchers to conduct more studies in Africa to fill data gaps. This complements the mission of ISEE Africa chapter. A local institution (University of Ibadan, Nigeria) was sourced to partner with ISEE Africa chapter to host the 3-day introductory training. The course was advertised extensively to all parts of Africa through social media and focus groups. Eligibility criteria for applicants are that they must be African, graduate students and scientists with basic degree in environment, health or a related discipline and ability to meet own transportation and accommodation. Applicants submitted online essay on African environmental issues, research interests and how the training will benefit own career. Training was face-to-face tutorial, videos, group work and hands-on exercises. The course faculty were four experienced tutors from 1SEE and partner institution. 213 (39%) out of 545 applications received from 15 African countries, passed the eligibility and screening criteria. Seventy-three (13%) successful applicants were invited but only 59 from Nigeria, Uganda and Guinea attended. Participants comprised 46% female and 54% male with 97% possessing a postgraduate qualification. Participants came from a heterogeneous background and for many, this was their first epidemiology course. Forty-eight participants joined ISEE. Participants feedback was very positive. There was potential for continuation of work locally. For future courses, location selection criteria should include a place that may attract participants from other countries, collaboration with reputable local institution, consideration for post-training sustainability, local organizing committee that can mobilize high number of participants and members for ISEE. Increased knowledge of environmental health risks in Africa is important to improve global public health. Therefore, capacity development in environmental epidemiology in Africa deserves priority in global scientific agenda.

Disentangling multiple environmental stressors in relation to incident cardiovascular disease in UK Biobank: a sparse principal component analysis

Background: Multicollinearity is a common challenge in environmental epidemiology research. However, existing methods are mostly difficult to implement or produce results that are not readily interpretable. Objectives: Our aim is to use sparse principal component analysis (SPCA) to address the multicollinearity issue and improve interpretability of the prospective association between environmental exposures and risk of incident cardiovascular disease (CVD) using data from a large population-based cohort. Methods: We derived principal components using SPCA techniques on 24 different environmental exposures from the UK Biobank (405,583 participants) and built Cox regression models to estimate the hazard ratio (HR) for incident cardiovascular disease (CVD) associated with each component. We then back-calculated the hazard ratio (HR) to each individual environmental exposure from the coefficients derived from the SPCA Cox model and loading scores using a commonly adopted method for instrumental variable analysis. Results: During an average follow-up of 9.7 years, 25,655 participants developed CVD. Components characterised by higher exposure to traffic intensity on the nearest major road or traffic-related air pollutants (NO2, NOX and PM2.5) were significantly associated with incident CVD risk, as identified by the SPCA Cox model. For each 1-SD (1.05 ug/m3) higher PM2.5, the HR was 1.09 (95% confidence interval 1.06-1.12); risk estimates were similar in magnitude for NO2 or NOx. Results for other exposures were variable and inconsistent. Conclusion: Using SPCA, our study consistently identified traffic-related air pollution as an important risk factor of CVD. SPCA offers a potential solution to address multicollinearity and the advantage for to identify relatively more important environmental exposures in relation to incident CVD outcome.

BKMR-CMA: A Novel R Command for Mediation Analysis in Environmental Mixture Studies

Background: Bayesian kernel machine regression (BKMR) is becoming a popular approach for studying the joint effect of environmental mixtures on health outcomes allowing for variable selection, which is particularly useful when continuous exposures display moderate correlations. An R package, bkmr, has been developed to implement this method. Recently, BKMR-causal mediation analysis (BKMR-CMA) has been proposed to estimate direct and indirect effects of mixtures through a hypothesized mediator. However, the current bkmr R package only allows the output of total effects of the mixtures. Method: We introduce two new commands within the bkmr package 1) BKMR-CMA: a command that allows the estimation of direct and indirect health effects of multiple environmental exposures through a single mediator; 2) BKMR-MI: a command that is used for valid estimation of environmental mixture effects and evaluation of uncertainty in the presence of missing data, which are imputed using multiple imputation techniques. The new commands also produce effective visualizations of the estimated causal effects and dose-response relationships. Results: The new computational approaches complement the existing bkmr package by allowing the study of direct and indirect effects and accounting for missing data in the estimation of the exposure effects. We illustrate the use of these novel commands in an environmental epidemiology study investigating the effects of metal mixtures on cardiovascular outcomes through a cardiovascular biomarker. Conclusion: Both BKMR-CMA and BKMR-MI are available on Github, and are expected to facilitate the conduct of reproducibility of mediation analysis in environmental epidemiology research.

Lessons Learned for Obtaining Water Data at the State Level and Linkage to Health Outcomes Data

Water quality remains an ongoing public health issue. Contaminants such as lead and arsenic are enduring water quality contaminants that threaten public health, but emerging contaminants also add to the growing health concerns of water quality. Water quality data is critical to understanding the health impacts and identifying additional contaminants resulting in adverse health outcomes. However, challenges remain in identifying and obtaining water data in the United States (US). No national, standardized database exists for water data in the US. Local public water systems are required to report regulatory violations to the state who then reports to these violations to federal regulators. Reporting violations may meet the regulatory requirements, but limitations are encountered in the utility of these data for environmental epidemiological research. As only violations are required to be reported, contaminant concentrations are only available at the state-level which requires laborious, time-consuming processes to obtain these data and/or costs associated with obtaining the data. In some cases, the states are unable or unwilling to release the data, which results in gaps for developing a national profile of water quality. If water quality data is obtained, there are issues in how to relate the information to health outcomes, as it typically is on a different geographical scale than health outcome data. Alas, even given these limitations of data availability, these data are often the only consistent water data available for epidemiological analyses at a national level. Therefore, it is critical to find methods to utilize these water quality data for health analyses. We will present several examples of the water data available at the different levels – local, state, and federal – as well as the challenges and strategies for linking these data to health outcomes data, including individual and county-level outcomes. This abstract does not reflect EPA policy. Keywords: Water, water quality, big data, environmental epidemiology