Validity of Mobility-Based Exposure Assessment of Air Pollution: A Comparative Analysis with Home-Based Exposure Assessment.

Air pollution exposure is typically assessed at the front door where people live in large-scale epidemiological studies, overlooking individuals' daily mobility out-of-home. However, there is limited evidence that incorporating mobility data into personal air pollution assessment improves exposure assessment compared to home-based assessments. This study aimed to compare the agreement between mobility-based and home-based assessments with personal exposure measurements. We measured repeatedly particulate matter (PM2.5) and black carbon (BC) using a sample of 41 older adults in the Netherlands. In total, 104 valid 24 h average personal measurements were collected. Home-based exposures were estimated by combining participants' home locations and temporal-adjusted air pollution maps. Mobility-based estimates of air pollution were computed based on smartphone-based tracking data, temporal-adjusted air pollution maps, indoor-outdoor penetration, and travel mode adjustment. Intraclass correlation coefficients (ICC) revealed that mobility-based estimates significantly improved agreement with personal measurements compared to home-based assessments. For PM2.5, agreement increased by 64% (ICC: 0.39-0.64), and for BC, it increased by 21% (ICC: 0.43-0.52). Our findings suggest that adjusting for indoor-outdoor pollutant ratios in mobility-based assessments can provide more valid estimates of air pollution than the commonly used home-based assessments, with no added value observed from travel mode adjustments.

Variability of the Human Serum Metabolome over 3 Months in the EXPOsOMICS Personal Exposure Monitoring Study.

Liquid chromatography coupled to high-resolution mass spectrometry (LC-HRMS) and untargeted metabolomics are increasingly used in exposome studies to study the interactions between nongenetic factors and the blood metabolome. To reliably and efficiently link detected compounds to exposures and health phenotypes in such studies, it is important to understand the variability in metabolome measures. We assessed the within- and between-subject variability of untargeted LC-HRMS measurements in 298 nonfasting human serum samples collected on two occasions from 157 subjects. Samples were collected ca. 107 (IQR: 34) days apart as part of the multicenter EXPOsOMICS Personal Exposure Monitoring study. In total, 4294 metabolic features were detected, and 184 unique compounds could be identified with high confidence. The median intraclass correlation coefficient (ICC) across all metabolic features was 0.51 (IQR: 0.29) and 0.64 (IQR: 0.25) for the 184 uniquely identified compounds. For this group, the median ICC marginally changed (0.63) when we included common confounders (age, sex, and body mass index) in the regression model. When grouping compounds by compound class, the ICC was largest among glycerophospholipids (median ICC 0.70) and steroids (0.67), and lowest for amino acids (0.61) and the O-acylcarnitine class (0.44). ICCs varied substantially within chemical classes. Our results suggest that the metabolome as measured with untargeted LC-HRMS is fairly stable (ICC > 0.5) over 100 days for more than half of the features monitored in our study, to reflect average levels across this time period. Variance across the metabolome will result in differential measurement error across the metabolome, which needs to be considered in the interpretation of metabolome results.

Short-term personal and outdoor exposure to ultrafine and fine particulate air pollution in association with blood pressure and lung function in healthy adults.

Studies reporting on associations between short-term exposure to outdoor fine (PM2.5), and ultrafine particles (UFP) and blood pressure and lung function have been inconsistent. Few studies have characterized exposure by personal monitoring, which especially for UFP may have resulted in substantial exposure measurement error. We investigated the association between 24-h average personal UFP, PM2.5, and soot exposure and dose and the health parameters blood pressure and lung function. We further assessed the short-term associations between outdoor concentrations measured at a central monitoring site and near the residences and these health outcomes. We performed three 24-h personal exposure measurements for UFP, PM2.5, and soot in 132 healthy adults from Basel (Switzerland), Amsterdam and Utrecht (the Netherlands), and Turin (Italy). Monitoring of each subject was conducted in different seasons in a one-year study period. Subject's activity levels and associated ventilation rates were measured using actigraphy to calculate the inhaled dose. After each 24-h monitoring session, blood pressure and lung function were measured. Contemporaneously with personal measurements, UFP, PM2.5 and soot were measured outdoor at the subject's residential address and at a central site in the research area. Associations between short-term personal and outdoor exposure and dose to UFP, PM2.5, and soot and health outcomes were tested using linear mixed effect models. The 24-h mean personal, residential and central site outdoor UFP exposures were not associated with blood pressure or lung function. UFP mean exposures in the 2-h prior to the health test was also not associated with blood pressure and lung function. Personal, central site and residential PM2.5 exposure were positively associated with systolic blood pressure (about 1.4 mmHg increase per Interquartile range). Personal soot exposure and dose were positively associated with diastolic blood pressure (1.2 and 0.9 mmHg increase per Interquartile range). No consistent associations between PM2.5 or soot exposure and lung function were observed. Short-term personal, residential outdoor or central site exposure to UFP was not associated with blood pressure or lung function. Short-term personal PM2.5 and soot exposures were associated with blood pressure, but not lung function.

Land use regression models for the oxidative potential of fine particles (PM2.5) in five European areas.

Oxidative potential (OP) of particulate matter (PM) is proposed as a biologically-relevant exposure metric for studies of air pollution and health. We aimed to evaluate the spatial variability of the OP of measured PM2.5 using ascorbate (AA) and (reduced) glutathione (GSH), and develop land use regression (LUR) models to explain this spatial variability. We estimated annual average values (m-3) of OPAA and OPGSH for five areas (Basel, CH; Catalonia, ES; London-Oxford, UK (no OPGSH); the Netherlands; and Turin, IT) using PM2.5 filters. OPAA and OPGSH LUR models were developed using all monitoring sites, separately for each area and combined-areas. The same variables were then used in repeated sub-sampling of monitoring sites to test sensitivity of variable selection; new variables were offered where variables were excluded (p > .1). On average, measurements of OPAA and OPGSH were moderately correlated (maximum Pearson's maximum Pearson's R = = .7) with PM2.5 and other metrics (PM2.5absorbance, NO2, Cu, Fe). HOV (hold-out validation) R2 for OPAA models was .21, .58, .45, .53, and .13 for Basel, Catalonia, London-Oxford, the Netherlands and Turin respectively. For OPGSH, the only model achieving at least moderate performance was for the Netherlands (R2 = .31). Combined models for OPAA and OPGSH were largely explained by study area with weak local predictors of intra-area contrasts; we therefore do not endorse them for use in epidemiologic studies. Given the moderate correlation of OPAA with other pollutants, the three reasonably performing LUR models for OPAA could be used independently of other pollutant metrics in epidemiological studies.

Land Use Regression Models for Ultrafine Particles in Six European Areas.

Long-term ultrafine particle (UFP) exposure estimates at a fine spatial scale are needed for epidemiological studies. Land use regression (LUR) models were developed and evaluated for six European areas based on repeated 30 min monitoring following standardized protocols. In each area; Basel (Switzerland), Heraklion (Greece), Amsterdam, Maastricht, and Utrecht ("The Netherlands"), Norwich (United Kingdom), Sabadell (Spain), and Turin (Italy), 160-240 sites were monitored to develop LUR models by supervised stepwise selection of GIS predictors. For each area and all areas combined, 10 models were developed in stratified random selections of 90% of sites. UFP prediction robustness was evaluated with the intraclass correlation coefficient (ICC) at 31-50 external sites per area. Models from Basel and The Netherlands were validated against repeated 24 h outdoor measurements. Structure and model R2 of local models were similar within, but varied between areas (e.g., 38-43% Turin; 25-31% Sabadell). Robustness of predictions within areas was high (ICC 0.73-0.98). External validation R2 was 53% in Basel and 50% in The Netherlands. Combined area models were robust (ICC 0.93-1.00) and explained UFP variation almost equally well as local models. In conclusion, robust UFP LUR models could be developed on short-term monitoring, explaining around 50% of spatial variance in longer-term measurements.

Robustness of intra urban land-use regression models for ultrafine particles and black carbon based on mobile monitoring.

Land-use regression (LUR) models for ultrafine particles (UFP) and Black Carbon (BC) in urban areas have been developed using short-term stationary monitoring or mobile platforms in order to capture the high variability of these pollutants. However, little is known about the comparability of predictions of mobile and short-term stationary models and especially the validity of these models for assessing residential exposures and the robustness of model predictions developed in different campaigns. We used an electric car to collect mobile measurements (n = 5236 unique road segments) and short-term stationary measurements (3 × 30min, n = 240) of UFP and BC in three Dutch cities (Amsterdam, Utrecht, Maastricht) in 2014-2015. Predictions of LUR models based on mobile measurements were compared to (i) measured concentrations at the short-term stationary sites, (ii) LUR model predictions based on short-term stationary measurements at 1500 random addresses in the three cities, (iii) externally obtained home outdoor measurements (3 × 24h samples; n = 42) and (iv) predictions of a LUR model developed based upon a 2013 mobile campaign in two cities (Amsterdam, Rotterdam). Despite the poor model R2 of 15%, the ability of mobile UFP models to predict measurements with longer averaging time increased substantially from 36% for short-term stationary measurements to 57% for home outdoor measurements. In contrast, the mobile BC model only predicted 14% of the variation in the short-term stationary sites and also 14% of the home outdoor sites. Models based upon mobile and short-term stationary monitoring provided fairly high correlated predictions of UFP concentrations at 1500 randomly selected addresses in the three Dutch cities (R2 = 0.64). We found higher UFP predictions (of about 30%) based on mobile models opposed to short-term model predictions and home outdoor measurements with no clear geospatial patterns. The mobile model for UFP was stable over different settings as the model predicted concentration levels highly correlated to predictions made by a previously developed LUR model with another spatial extent and in a different year at the 1500 random addresses (R2 = 0.80). In conclusion, mobile monitoring provided robust LUR models for UFP, valid to use in epidemiological studies.

microRNA expression profiles and personal monitoring of exposure to particulate matter.

An increasing number of findings from epidemiological studies support associations between exposure to air pollution and the onset of several diseases, including pulmonary, cardiovascular and neurodegenerative diseases, and malignancies. However, intermediate, and potentially mediating, biological mechanisms associated with exposure to air pollutants are largely unknown. Previous studies on the human exposome have shown that the expression of certain circulating microRNAs (miRNAs), regulators of gene expression, are altered upon exposure to traffic-related air pollutants. In the present study, we investigated the relationship between particulate matter (PM) smaller than 2.5 µm (PM2.5), PM2.5 absorbance (as a proxy of black carbon and soot), and ultrafine-particles (UFP, smaller than 0.1 µm), measured in healthy volunteers by 24 h personal monitoring (PEM) sessions and global expression levels of peripheral blood miRNAs. The PEM sessions were conducted in four European countries, namely Switzerland (Basel), United Kingdom (Norwich), Italy (Turin), and The Netherlands (Utrecht). miRNAs expression levels were analysed using microarray technology on blood samples from 143 participants. Seven miRNAs, hsa-miR-24-3p, hsa-miR-4454, hsa-miR-4763-3p, hsa-miR-425-5p, hsa-let-7d-5p, hsa-miR-502-5p, and hsa-miR-505-3p were significantly (FDR corrected) expressed in association with PM2.5 personal exposure, while no significant association was found between miRNA expression and the other pollutants. The results obtained from this investigation suggest that personal exposure to PM2.5 is associated with miRNA expression levels, showing the potential for these circulating miRNAs as novel biomarkers for air pollution health risk assessment.

ExpoApp: An integrated system to assess multiple personal environmental exposures.

To assess environmental exposures at the individual level, new assessment methods and tools are required. We developed an exposure assessment system (ExpoApp) for smartphones. ExpoApp integrates: (i) geo-location and accelerometry measurements from a waist attached smartphone, (ii) data from portable monitors, (iii) geographic information systems, and (iv) individual's information. ExpoApp calculates time spent in microenvironments, physical activity level, inhalation rate, and environmental exposures and doses (e.g., green spaces, inhaled ultrafine particles- UFP). We deployed ExpoApp in a panel study of 158 adults from five cities (Amsterdam and Utrecht- the Netherlands, Basel- Switzerland, Norwich- UK, and Torino- Italy) with an UFP monitor. To evaluate ExpoApp, participants also carried a reference accelerometer (ActiGraph) and completed a travel-activity diary (TAD). System reliability and validity of measurements were evaluated by comparing the monitoring failure rate and the agreement on time spent in microenvironments and physical activity with the reference tools. There were only significant failure rate differences between ExpoApp and ActiGraph in Norwich. Agreement on time in microenvironments and physical activity level between ExpoApp and reference tools was 86.6% (86.5-86.7) and 75.7% (71.5-79.4), respectively. ExpoApp estimated that participants inhaled 16.5 × 1010 particles/day of UFP and had almost no contact with green spaces (24% of participants spent ≥30 min/day in green spaces). Participants with more contact with green spaces had higher inhaled dose of UFP, except for the Netherlands, where the relationship was the inverse. ExpoApp is a reliable system and provides accurate individual's measurements, which may help to understand the role of environmental exposures on the origin and course of diseases.

Spatial Variation of Endotoxin Concentrations Measured in Ambient PM10 in a Livestock-Dense Area: Implementation of a Land-Use Regression Approach.

BACKGROUND: Results from studies on residential health effects of livestock farming are inconsistent, potentially due to simple exposure proxies used (e.g., livestock density). Accuracy of these proxies compared with measured exposure concentrations is unknown. OBJECTIVES: We aimed to assess spatial variation of endotoxin in PM10 (particulate matter ≤10µm) at residential level in a livestock-dense area, compare simple livestock exposure proxies to measured endotoxin concentrations, and evaluate whether land-use regression (LUR) can be used to explain spatial variation of endotoxin. METHODS: The study area (3,000 km2) was located in Netherlands. Ambient PM10 was collected at 61 residential sites representing a variety of surrounding livestock-related characteristics. Three to four 2-wk averaged samples were collected at each site. A local reference site was used for temporal variation adjustment. Samples were analyzed for PM10 mass by weighing and for endotoxin by using the limulus amebocyte lysate assay. Three LUR models were developed, first a model based on general livestock-related GIS predictors only, followed by models that also considered species-specific predictors and farm type-specific predictors. RESULTS: Variation in concentrations measured between sites was substantial for endotoxin and more limited for PM10 (coefficient of variation: 43%, 8%, respectively); spatial patterns differed considerably. Simple exposure proxies were associated with endotoxin concentrations although spatial variation explained was modest (R2<26%). LUR models using a combination of animal-specific livestock-related characteristics performed markedly better, with up to 64% explained spatial variation. CONCLUSION: The considerable spatial variation of ambient endotoxin concentrations measured in a livestock-dense area can largely be explained by LUR modeling based on livestock-related characteristics. Application of endotoxin LUR models seems promising for residential exposure estimation within health studies. https://doi.org/10.1289/EHP2252.

Long-Term Exposure to Ultrafine Particles and Incidence of Cardiovascular and Cerebrovascular Disease in a Prospective Study of a Dutch Cohort.

BACKGROUND: There is growing evidence that exposure to ultrafine particles (UFP; particles smaller than [Formula: see text]) may play an underexplored role in the etiology of several illnesses, including cardiovascular disease (CVD). OBJECTIVES: We aimed o investigate the relationship between long-term exposure to ambient UFP and incident cardiovascular and cerebrovascular disease (CVA). As a secondary objective, we sought to compare effect estimates for UFP with those derived for other air pollutants, including estimates from two-pollutant models. METHODS: Using a prospective cohort of 33,831 Dutch residents, we studied the association between long-term exposure to UFP (predicted via land use regression) and incident disease using Cox proportional hazard models. Hazard ratios (HR) for UFP were compared to HRs for more routinely monitored air pollutants, including particulate matter with aerodynamic diameter [Formula: see text] ([Formula: see text]), PM with aerodynamic diameter [Formula: see text] ([Formula: see text]), and [Formula: see text]. RESULTS: Long-term UFP exposure was associated with an increased risk for all incident CVD [[Formula: see text] per [Formula: see text]; 95% confidence interval (CI): 1.03, 1.34], myocardial infarction (MI) ([Formula: see text]; 95% CI: 1.00, 1.79), and heart failure ([Formula: see text]; 95% CI: 1.17, 2.66). Positive associations were also estimated for [Formula: see text] ([Formula: see text]; 95% CI: 1.01, 1.48 per [Formula: see text]) and coarse PM ([Formula: see text]; HR for all [Formula: see text]; 95% CI: 1.01, 1.45 per [Formula: see text]). CVD was not positively associated with [Formula: see text] (HR for all [Formula: see text]; 95% CI: 0.75, 1.28 per [Formula: see text]). HRs for UFP and CVAs were positive, but not significant. In two-pollutant models ([Formula: see text] and [Formula: see text]), positive associations tended to remain for UFP, while HRs for [Formula: see text] and [Formula: see text] generally attenuated towards the null. CONCLUSIONS: These findings strengthen the evidence that UFP exposure plays an important role in cardiovascular health and that risks of ambient air pollution may have been underestimated based on conventional air pollution metrics. https://doi.org/10.1289/EHP3047.

Acute changes in DNA methylation in relation to 24 h personal air pollution exposure measurements: A panel study in four European countries.

BACKGROUND: One of the potential mechanisms linking air pollution to health effects is through changes in DNA-methylation, which so far has mainly been analyzed globally or at candidate sites. OBJECTIVE: We investigated the association of personal and ambient air pollution exposure measures with genome-wide DNA-methylation changes. METHODS: We collected repeated 24-hour personal and ambient exposure measurements of particulate matter (PM2.5), PM2.5 absorbance, and ultrafine particles (UFP) and peripheral blood samples from a panel of 157 healthy non-smoking adults living in four European countries. We applied univariate mixed-effects models to investigate the association between air pollution and genome-wide DNA-methylation perturbations at single CpG (cytosine-guanine dinucleotide) sites and in Differentially Methylated Regions (DMRs). Subsequently, we explored the association of air pollution-induced methylation alterations with gene expression and serum immune marker levels measured in the same subjects. RESULTS: Personal exposure to PM2.5 was associated with methylation changes at 13 CpG sites and 69 DMRs. Two of the 13 identified CpG sites (mapped to genes KNDC1 and FAM50B) were located within these DMRs. In addition, 42 DMRs were associated with personal PM2.5 absorbance exposure, 16 DMRs with personal exposure to UFP, 4 DMRs with ambient exposure to PM2.5, 16 DMRs with ambient PM2.5 absorbance exposure, and 15 DMRs with ambient UFP exposure. Correlation between methylation levels at identified CpG sites and gene expression and immune markers was generally moderate. CONCLUSION: This study provides evidence for an association between 24-hour exposure to air pollution and DNA-methylation at single sites and regional clusters of CpGs. Analysis of differentially methylated regions provides a promising avenue to further explore the subtle impact of environmental exposures on DNA-methylation.