Within-City Spatial Variations in Multiple Measures of PM2.5 Oxidative Potential in Toronto, Canada

Few studies have characterized within-city spatial variations in the oxidative potential of fine particulate air pollution (PM_2.5). In this study, we evaluated multiple measures of PM_2.5 oxidative potential across Toronto, Canada (2016-2017), including glutathione/ascorbate-related oxidative potential (OP^GSH and OP^AA) and dithiothreitol depletion (OP^DTT). Integrated 2-week samples were collected from 67 sites in summer and 42 sites in winter. Multivariable linear models were developed to predict OP based on various land use/traffic factors, and PM_2.5 metals and black carbon were also examined. All three measures of PM_2.5 oxidative potential varied substantially across Toronto. OP^AA and OP^DTT were primarily associated with traffic-related components of PM_2.5 (i.e., Fe, Cu, and black carbon) whereas OP^GSH was not a strong marker for traffic during either season. During summer, multivariable models performed best for OP^AA ( R_CV² = 0.48) followed by OP^DTT ( R_CV² = 0.32) and OP^GSH ( R_CV² = 0.22). During winter, model performance was best for OP^DTT ( R_CV² = 0.55) followed by OP^GSH ( R_CV² = 0.50) and OP^AA ( R_CV² = 0.23). Model parameters varied between seasons, and between-season differences in PM_2.5 mass concentrations were weakly/moderately correlated with seasonal differences in OP. Our findings highlight substantial within-city variations in PM_2.5 oxidative potential. More detailed information is needed on local sources of air pollution to improve model performance.

Associations of Long-Term Exposure to Ultrafine Particles and Nitrogen Dioxide With Increased Incidence of Congestive Heart Failure and Acute Myocardial Infarction

Although long-term exposure to traffic-related air pollutants such as nitrogen dioxide has been linked to cardiovascular disease (CVD) mortality, little is known about the association between ultrafine particles (UFPs), defined as particles less than or equal to 0.1 μm in diameter, and incidence of major CVD events. We conducted a population-based cohort study to assess the associations of chronic exposure to UFPs and nitrogen dioxide with incident congestive heart failure (CHF) and acute myocardial infarction. Our study population comprised all long-term Canadian residents aged 30-100 years who lived in Toronto, Ontario, Canada, during the years 1996-2012. We estimated annual concentrations of UFPs and nitrogen dioxide by means of land-use regression models and assigned these estimates to participants' postal-code addresses in each year during the follow-up period. We estimated hazard ratios for the associations of UFPs and nitrogen dioxide with incident CVD using random-effects Cox proportional hazards models. We controlled for smoking and obesity using an indirect adjustment method. Our cohorts comprised approximately 1.1 million individuals at baseline. In single-pollutant models, each interquartile-range increase in UFP exposure was associated with increased incidence of CHF (hazard ratio for an interquartile-range increase (HRIQR) = 1.03, 95% confidence interval (CI): 1.02, 1.05) and acute myocardial infarction (HRIQR = 1.05, 95% CI: 1.02, 1.07). Adjustment for fine particles and nitrogen dioxide did not materially change these estimated associations. Exposure to nitrogen dioxide was also independently associated with higher CHF incidence (HRIQR = 1.04, 95% CI: 1.03, 1.06).

A picture tells a thousand…exposures: Opportunities and challenges of deep learning image analyses in exposure science and environmental epidemiology

BACKGROUND

Artificial intelligence (AI) is revolutionizing our world, with applications ranging from medicine to engineering.

OBJECTIVES

Here we discuss the promise, challenges, and probable data sources needed to apply AI in the fields of exposure science and environmental health. In particular, we focus on the use of deep convolutional neural networks to estimate environmental exposures using images and other complementary data sources such as cell phone mobility and social media information.

DISCUSSION

Characterizing the health impacts of multiple spatially-correlated exposures remains a challenge in environmental epidemiology. A shift toward integrated measures that simultaneously capture multiple aspects of the urban built environment could improve efficiency and provide important insights into how our collective environments influence population health. The widespread adoption of AI in exposure science is on the frontier. This will likely result in new ways of understanding environmental impacts on health and may allow for analyses to be efficiently scaled for broad coverage. Image-based convolutional neural networks may also offer a cost-effective means of estimating local environmental exposures in low and middle-income countries where monitoring and surveillance infrastructure is limited. However, suitable databases must first be assembled to train and evaluate these models and these novel approaches should be complemented with traditional exposure metrics.

CONCLUSIONS

The promise of deep learning in environmental health is great and will complement existing measurements for data-rich settings and could enhance the resolution and accuracy of estimates in data poor scenarios. Interdisciplinary partnerships will be needed to fully realize this potential.

Accommodating exogenous variable and decision rule heterogeneity in discrete choice models: Application to bicyclist route choice

The proposed research contributes to our understanding of incorporating heterogeneity in discrete choice models with respect to exogenous variables and decision rules. Specifically, the proposed latent segmentation based mixed models segment population to different classes with their own decision rules while also incorporating unobserved heterogeneity within the segment level models. In our analysis, we choose to consider both random utility and random regret theories. Further, instead of assuming the number of segments (as 2), we conduct an exhaustive exploration with multiple segments across the two decision rules. The model estimation is conducted using a stated preference data from 695 commuter cyclists compiled through a web-based survey. The probabilistic allocation of respondents to different segments indicates that female commuter cyclists are more utility oriented; however, the majority of the commuter cyclist’s choice pattern is consistent with regret minimization mechanism. Overall, cyclists’ route choice decisions are influenced by roadway attributes, cycling infrastructure availability, pollution exposure, and travel time. The analysis approach also allows us to investigate time based trade-offs across cyclists belonging to different classes. Interestingly, we observe that the trade-off values in regret and utility based segments for roadway attributes are similar in magnitude; but the values differ greatly for cycling infrastructure and pollution exposure attributes, particularly for maximum exposure levels.

Comparing emission rates derived from a model with those estimated using a plume-based approach and quantifying the contribution of vehicle classes to on-road emissions and air quality

This study presents a comparison of fleet average emission factor (s) derived from a traffic emission model with EFs estimated using plume-based measurements, including an investigation of the contribution of vehicle classes to carbon monoxide (CO), nitrogen oxides (NO_x), and elemental carbon (EC) along an urban corridor. To this end, a field campaign was conducted over one week in June 2016 on an arterial road in Toronto, Canada. Traffic data were collected using a traffic camera and a radar, whereas air quality was characterized using two monitoring stations: one located at ground level and another at the rooftop of a four-story building. A traffic simulation model was calibrated and validated, and second-by-second speed profiles for all vehicle trajectories were extracted to model emissions. In addition, dispersion modeling was conducted to identify the extent to which differences in emissions translate to differences in near-road concentrations. The results indicate that modeled EFs for CO and NO_x are twice as high as plume-based EFs. Besides, modeled results indicate that transit bus emissions accounted for 60% and 70% of the total emissions of NO_x and EC, respectively. Transit bus emission rates in g/passenger·km for NO_x and EC were up to 8 and 22 times, respectively, the emission rates of passenger cars. In contrast, the Toronto streetcars, which are electrically fueled, were found to improve near-road air quality despite their negative impact on traffic speeds. Finally, we observe that the difference in estimated concentrations derived from the two methods is not as large as the difference in estimated emissions due to the influence of meteorology and of the urban background given that the study network is located in a busy downtown area. Implications: This study presents a comparison of fleet average emission factor (s) derived from a traffic emission model with EFs estimated using plume-based measurements, including an investigation of the contribution of vehicle classes to various pollutants. Besides, dispersion modeling was conducted to identify the extent to which differences in emissions translate to differences in near-road concentrations. It was observed that the difference in estimated concentrations derived from the two methods is not as large as the difference in estimated emissions due to the influence of meteorology and of the urban background, as the study network is located in a busy downtown area.

Capturing the spatial variability of noise levels based on a short-term monitoring campaign and comparing noise surfaces against personal exposures collected through a panel study

Environmental noise can cause important cardiovascular effects, stress and sleep disturbance. The development of appropriate methods to estimate noise exposure within a single urban area remains a challenging task, due to the presence of various transportation noise sources (road, rail, and aircraft). In this study, we developed a land-use regression (LUR) approach using a Generalized Additive Model (GAM) for LA_eq (equivalent noise level) to capture the spatial variability of noise levels in Toronto, Canada. Four different model formulations were proposed based on continuous 20-min noise measurements at 92 sites and a leave one out cross-validation (LOOCV). Models where coefficients for variables considered as noise sources were forced to be positive, led to the development of more realistic exposure surfaces. Three different measures were used to assess the models; adjusted R² (0.44-0.64), deviance (51-72%) and Akaike information criterion (AIC) (469.2-434.6). When comparing exposures derived from the four approaches to personal exposures from a panel study, we observed that all approaches performed very similarly, with values for the Fractional mean bias (FB), normalized mean square error (NMSE), and normalized absolute difference (NAD) very close to 0. Finally, we compared the noise surfaces with data collected from a previous campaign consisting of 1-week measurements at 200 fixed sites in Toronto and observed that the strongest correlations occurred between our predictions and measured noise levels along major roads and highway collectors. Our validation against long-term measurements and panel data demonstrates that manual modifications brought to the models were able to reduce bias in model predictions and achieve a wider range of exposures, comparable with measurement data.

Evaluating the Impact of Neighborhood Characteristics on Differences between Residential and Mobility-Based Exposures to Outdoor Air Pollution

Epidemiological studies often assign outdoor air pollution concentrations to residential locations without accounting for mobility patterns. In this study, we examined how neighborhood characteristics may influence differences in exposure assessments between outdoor residential concentrations and mobility-based exposures. To do this, we linked residential location and mobility data to exposure surfaces for NO₂, PM_2.5, and ultrafine particles in Montreal, Canada for 5452 people in 2016. Mobility data were collected using the MTL Trajet smartphone application (mean: 16 days/subject). Generalized additive models were used to identify important neighborhood predictors of differences between residential and mobility-based exposures and included residential distances to highways, traffic counts within 500 m of the residence, neighborhood walkability, median income, and unemployment rate. Final models including these parameters provided unbiased estimates of differences between residential and mobility-based exposures with small root-mean-square error values in 10-fold cross validation samples. In general, our findings suggest that differences between residential and mobility-based exposures are not evenly distributed across cities and are greater for pollutants with higher spatial variability like NO₂. It may be possible to use neighborhood characteristics to predict the magnitude and direction of this error to better understand its likely impact on risk estimates in epidemiological analyses.

Associations between ambient air pollution and daily mortality in a cohort of congestive heart failure: Case-crossover and nested case-control analyses using a distributed lag nonlinear model

BACKGROUND

Persons with congestive heart failure may be at higher risk of the acute effects related to daily fluctuations in ambient air pollution. To meet some of the limitations of previous studies using grouped-analysis, we developed a cohort study of persons with congestive heart failure to estimate whether daily non-accidental mortality were associated with spatially-resolved, daily exposures to ambient nitrogen dioxide (NO₂) and ozone (O₃), and whether these associations were modified according to a series of indicators potentially reflecting complications or worsening of health.

METHODS

We constructed the cohort from the linkage of administrative health databases. Daily exposure was assigned from different methods we developed previously to predict spatially-resolved, time-dependent concentrations of ambient NO₂ (all year) and O₃ (warm season) at participants' residences. We performed two distinct types of analyses: a case-crossover that contrasts the same person at different times, and a nested case-control that contrasts different persons at similar times. We modelled the effects of air pollution and weather (case-crossover only) on mortality using distributed lag nonlinear models over lags 0 to 3 days. We developed from administrative health data a series of indicators that may reflect the underlying construct of "declining health", and used interactions between these indicators and the cross-basis function for air pollutant to assess potential effect modification.

RESULTS

The magnitude of the cumulative as well as the lag-specific estimates of association differed in many instances according to the metric of exposure. Using the back-extrapolation method, which is our preferred exposure model, we found for the case-crossover design a cumulative mean percentage changes (MPC) in daily mortality per interquartile increment in NO₂ (8.8 ppb) of 3.0% (95% CI: -0.4, 6.6%) and for O₃ (16.5 ppb) 3.5% (95% CI: -4.5, 12.1). For O₃ there was strong confounding by weather (unadjusted MPC = 7.1%; 95% CI: 1.7, 12.7%). For the nested case-control approach the cumulative MPC for NO₂ in daily mortality was 2.9% (95% CI: -0.9, 6.9%) and for O₃ 7.3% (95% CI: 3.0, 11.9%). We found evidence of effect modification between daily mortality and cumulative NO₂ and O₃ according to the prescribed dose of furosemide in the nested case-control analysis, but not in the case-crossover analysis.

CONCLUSIONS

Mortality in congestive heart failure was associated with exposure to daily ambient NO₂ and O₃ predicted from a back-extrapolation method using a land use regression model from dense sampling surveys. The methods used to assess exposure can have considerable influence on the estimated acute health effects of the two air pollutants.

Development and Comparison of Air Pollution Exposure Surfaces Derived from On-Road Mobile Monitoring and Short-Term Stationary Sidewalk Measurements

Land-use regression (LUR) models of air pollutants are frequently developed on the basis of short-term stationary or mobile monitoring approaches, which raises the question of whether these two data collection protocols lead to similar exposure surfaces. In this study, we measured ultrafine particles (UFP) and black carbon (BC) concentrations in Toronto during summer 2016, using two short-term data collection approaches: mobile, involving 3023 road segments sampled on bicycles, and stationary, involving 92 sidewalk locations. We developed four LUR models and exposure surfaces, for the two pollutants and measurement protocols. Coefficients of determination ( R²) varied from 0.434 to 0.525. Various small-scale traffic variables were included in the mobile LUR. Pearson correlation coefficients between the mobile and stationary surfaces were 0.23 for UFP and 0.49 for BC. We also compared the two surfaces using personal exposures from a panel study in Toronto conducted during the same period. The personal exposures differed from the outdoor exposures derived from the combination of GPS information and exposure surfaces. For UFP, the median for personal outdoor exposure was 26 344 part/cm³, while the cycling and stationary surfaces predicted medians of 31 201 and 19 057 part/cm³. Similar trends were observed for BC, with median exposures of 1764 (personal), 1799 (cycling), and 1469 ng/m³ (stationary).

Estimating the health benefits of planned public transit investments in Montreal

BACKGROUND

Since public transit infrastructure affects road traffic volumes and influences transportation mode choice, which in turn impacts health, it is important to estimate the alteration of the health burden linked with transit policies.

OBJECTIVE

We quantified the variation in health benefits and burden between a business as usual (BAU) and a public transit (PT) scenarios in 2031 (with 8 and 19 new subway and train stations) for the greater Montreal region.

METHOD

Using mode choice and traffic assignment models, we predicted the transportation mode choice and traffic assignment on the road network. Subsequently, we estimated the distance travelled in each municipality by mode, the minutes spent in active transportation, as well as traffic emissions. Thereafter we estimated the health burden attributed to air pollution and road traumas and the gains associated with active transportation for both the BAU and PT scenarios.

RESULTS

We predicted a slight decrease of overall trips and kilometers travelled by car as well as an increase of active transportation for the PT in 2031 vs the BAU. Our analysis shows that new infrastructure will reduce the overall burden of transportation by 2.5 DALYs per 100,000 persons. This decrease is caused by the reduction of road traumas occurring in the inner suburbs and central Montreal region as well as gains in active transportation in the inner suburbs.

CONCLUSION

Based on the results of our study, transportation planned public transit projects for Montreal are unlikely to reduce drastically the burden of disease attributable to road vehicles and infrastructures in the Montreal region. The impact of the planned transportation infrastructures seems to be very low and localized mainly in the areas where new public transit stations are planned.

Capturing the sensitivity of land-use regression models to short-term mobile monitoring campaigns using air pollution micro-sensors

The development of reliable measures of exposure to traffic-related air pollution is crucial for the evaluation of the health effects of transportation. Land-use regression (LUR) techniques have been widely used for the development of exposure surfaces, however these surfaces are often highly sensitive to the data collected. With the rise of inexpensive air pollution sensors paired with GPS devices, we witness the emergence of mobile data collection protocols. For the same urban area, can we achieve a 'universal' model irrespective of the number of locations and sampling visits? Can we trade the temporal representation of fixed-point sampling for a larger spatial extent afforded by mobile monitoring? This study highlights the challenges of short-term mobile sampling campaigns in terms of the resulting exposure surfaces. A mobile monitoring campaign was conducted in 2015 in Montreal; nitrogen dioxide (NO₂) levels at 1395 road segments were measured under repeated visits. We developed LUR models based on sub-segments, categorized in terms of the number of visits per road segment. We observe that LUR models were highly sensitive to the number of road segments and to the number of visits per road segment. The associated exposure surfaces were also highly dissimilar.

The association between the incidence of postmenopausal breast cancer and concentrations at street-level of nitrogen dioxide and ultrafine particles

BACKGROUND

There is scant information as to whether traffic-related air pollution is associated with the incidence of breast cancer. Nitrogen dioxide (NO₂) and ultrafine particles (UFPs, <0.1µm), are two pollutants that capture intra-urban variations in traffic-related air pollution and may also be associated with incidence.

METHODS

We conducted a population-based, case-control study of street-level concentrations of NO₂ and UFPs and incident postmenopausal breast cancer in Montreal, Canada. Incident cases were identified between 2008 and 2011 from all but one hospital that treated breast cancer in the Montreal area. Population controls were identified from provincial electoral lists of Montreal residents and frequency-matched to cases using 5-year age groups. Concentrations of NO₂ and UFPs were estimated using two separate land-use regression models. Exposures were assigned to residential locations at the time of recruitment, and we identified residential histories of women who had lived in these residences for 10 years or more. Odds ratios (OR) and 95% confidence intervals (CI) were estimated using logistic regression models adjusting for individual-level and ecological covariates. We assessed the functional form of NO₂ and UFP exposures using natural cubic splines.

RESULTS

We found that the functional form of the response functions between incident postmenopausal breast cancer and concentrations of NO₂ and UFPs were consistent with linearity. For NO₂, we found increasing risks of breast cancer for all subjects combined and stronger associations when analyses were restricted to those women who had lived at their current address for 10 years or more. Specifically, the OR, adjusted for personal covariates, per increase in the interquartile range (IQR=3.75 ppb) of NO₂ was 1.08 (95%CI: 0.92-1.27). For women living in their homes for 10 years or more, the adjusted OR was 1.17 (95%CI: 0.93-1.46; IQR=3.84 ppb); for those not living at that home 10 years before the study, it was 0.93 (95%CI: 0.64, 1.36; IQR=3.65 ppb). For UFPs, the ORs were lower than for NO₂, with little evidence of association in any of the models or sub-analyses and little variability in the ORs (about 1.02 for an IQR of ~3500cm^-3). On the other hand, we found higher ORs amongst cases with positive oestrogen and progesterone receptor status; namely for NO₂, the OR was 1.13 (95%CI: 0.94-1.35) and for UFPs it was 1.05 (95%CI: 0.96-1.14).

CONCLUSIONS

Our findings suggest that exposure to ambient NO₂ and UFPs may increase the risk of incident postmenopausal breast cancer especially amongst cases with positive oestrogen and progesterone receptor status.

Spatial variations in ambient ultrafine particle concentrations and the risk of incident prostate cancer: A case-control study

BACKGROUND

Diesel exhaust contains large numbers of ultrafine particles (UFPs, <0.1µm) and is a recognized human carcinogen. However, epidemiological studies have yet to evaluate the relationship between UFPs and cancer incidence.

METHODS

We conducted a case-control study of UFPs and incident prostate cancer in Montreal, Canada. Cases were identified from all main Francophone hospitals in the Montreal area between 2005 and 2009. Population controls were identified from provincial electoral lists of French Montreal residents and frequency-matched to cases using 5-year age groups. UFP exposures were estimated using a land use regression model. Exposures were assigned to residential locations at the time of diagnosis/recruitment as well as approximately 10-years earlier to consider potential latency between exposure and disease onset. Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated per interquartile range (IQR) increase in UFPs (approximately 4000 particles/cm³) using logistic regression models adjusting for individual-level and ecological covariates.

RESULTS

Ambient UFP concentrations were associated with an increased risk of prostate cancer (OR=1.10, 95% CI: 1.01, 1.19) in fully adjusted models when exposures were assigned to residences 10-years prior to diagnosis. This risk estimate increased slightly (OR=1.17, 95% CI; 1.01, 1.35) when modeled as a non-linear natural spline function. A smaller increased risk (OR=1.04, 95% CI: 0.97, 1.11) was observed when exposures were assigned to residences at the time of diagnosis.

CONCLUSIONS

Exposure to ambient UFPs may increase the risk of prostate cancer. Future studies are needed to replicate this finding as this is the first study to evaluate this relationship.

Comparison of spatiotemporal prediction models of daily exposure of individuals to ambient nitrogen dioxide and ozone in Montreal, Canada

BACKGROUND

In previous studies investigating the short-term health effects of ambient air pollution the exposure metric that is often used is the daily average across monitors, thus assuming that all individuals have the same daily exposure. Studies that incorporate space-time exposures of individuals are essential to further our understanding of the short-term health effects of ambient air pollution.

OBJECTIVES

As part of a longitudinal cohort study of the acute effects of air pollution that incorporated subject-specific information and medical histories of subjects throughout the follow-up, the purpose of this study was to develop and compare different prediction models using data from fixed-site monitors and other monitoring campaigns to estimate daily, spatially-resolved concentrations of ozone (O₃) and nitrogen dioxide (NO₂) of participants' residences in Montreal, 1991-2002.

METHODS

We used the following methods to predict spatially-resolved daily concentrations of O₃ and NO₂ for each geographic region in Montreal (defined by three-character postal code areas): (1) assigning concentrations from the nearest monitor; (2) spatial interpolation using inverse-distance weighting; (3) back-extrapolation from a land-use regression model from a dense monitoring survey, and; (4) a combination of a land-use and Bayesian maximum entropy model. We used a variety of indices of agreement to compare estimates of exposure assigned from the different methods, notably scatterplots of pairwise predictions, distribution of differences and computation of the absolute agreement intraclass correlation (ICC). For each pairwise prediction, we also produced maps of the ICCs by these regions indicating the spatial variability in the degree of agreement.

RESULTS

We found some substantial differences in agreement across pairs of methods in daily mean predicted concentrations of O₃ and NO₂. On a given day and postal code area the difference in the concentration assigned could be as high as 131ppb for O₃ and 108ppb for NO₂. For both pollutants, better agreement was found between predictions from the nearest monitor and the inverse-distance weighting interpolation methods, with ICCs of 0.89 (95% confidence interval (CI): 0.89, 0.89) for O₃ and 0.81 (95%CI: 0.80, 0.81) for NO₂, respectively. For this pair of methods the maximum difference on a given day and postal code area was 36ppb for O₃ and 74ppb for NO₂. The back-extrapolation method showed a higher degree of disagreement with the nearest monitor approach, inverse-distance weighting interpolation, and the Bayesian maximum entropy model, which were strongly constrained by the sparse monitoring network. The maps showed that the patterns of agreement differed across the postal code areas and the variability depended on the pair of methods compared and the pollutants. For O₃, but not NO₂, postal areas showing greater disagreement were mostly located near the city centre and along highways, especially in maps involving the back-extrapolation method.

CONCLUSIONS

In view of the substantial differences in daily concentrations of O₃ and NO₂ predicted by the different methods, we suggest that analyses of the health effects from air pollution should make use of multiple exposure assessment methods. Although we cannot make any recommendations as to which is the most valid method, models that make use of higher spatially resolved data, such as from dense exposure surveys or from high spatial resolution satellite data, likely provide the most valid estimates.

Long-term exposure to ambient ultrafine particles and respiratory disease incidence in in Toronto, Canada: a cohort study

BACKGROUND

Little is known about the long-term health effects of ambient ultrafine particles (<0.1 μm) (UFPs) including their association with respiratory disease incidence. In this study, we examined the relationship between long-term exposure to ambient UFPs and the incidence of lung cancer, adult-onset asthma, and chronic obstructive pulmonary disease (COPD).

METHODS

Our study cohort included approximately 1.1 million adults who resided in Toronto, Canada and who were followed for disease incidence between 1996 and 2012. UFP exposures were assigned to residential locations using a land use regression model. Random-effect Cox proportional hazard models were used to estimate hazard ratios (HRs) describing the association between ambient UFPs and respiratory disease incidence adjusting for ambient fine particulate air pollution (PM_2.5), NO₂, and other individual/neighbourhood-level covariates.

RESULTS

In total, 74,543 incident cases of COPD, 87,141 cases of asthma, and 12,908 cases of lung cancer were observed during follow-up period. In single pollutant models, each interquartile increase in ambient UFPs was associated with incident COPD (HR = 1.06, 95% CI: 1.05, 1.09) but not asthma (HR = 1.00, 95% CI: 1.00, 1.01) or lung cancer (HR = 1.00, 95% CI: 0.97, 1.03). Additional adjustment for NO₂ attenuated the association between UFPs and COPD and the HR was no longer elevated (HR = 1.01, 95% CI: 0.98, 1.03). PM_2.5 and NO₂ were each associated with increased incidence of all three outcomes but risk estimates for lung cancer were sensitive to indirect adjustment for smoking and body mass index.

CONCLUSIONS

In general, we did not observe clear evidence of positive associations between long-term exposure to ambient UFPs and respiratory disease incidence independent of other air pollutants. Further replication is required as few studies have evaluated these relationships.

Robustness of Land-Use Regression Models Developed from Mobile Air Pollutant Measurements

Land-use regression (LUR) models are useful for resolving fine scale spatial variations in average air pollutant concentrations across urban areas. With the rise of mobile air pollution campaigns, characterized by short-term monitoring and large spatial extents, it is important to investigate the effects of sampling protocols on the resulting LUR. In this study a mobile lab was used to repeatedly visit a large number of locations (∼1800), defined by road segments, to derive average concentrations across the city of Montreal, Canada. We hypothesize that the robustness of the LUR from these data depends upon how many independent, random times each location is visited (N_vis) and the number of locations (N_loc) used in model development and that these parameters can be optimized. By performing multiple LURs on random sets of locations, we assessed the robustness of the LUR through consistency in adjusted R² (i.e., coefficient of variation, CV) and in regression coefficients among different models. As N_loc increased, R²_adj became less variable; for N_loc = 100 vs N_loc = 300 the CV in R²_adj for ultrafine particles decreased from 0.088 to 0.029 and from 0.115 to 0.076 for NO_2. The CV in the R²_adj also decreased as N_vis increased from 6 to 16; from 0.090 to 0.014 for UFP. As N_loc and N_vis increase, the variability in the coefficient sizes across the different model realizations were also seen to decrease.

Statistical modeling of the spatial variability of environmental noise levels in Montreal, Canada, using noise measurements and land use characteristics

The availability of noise maps to assess exposure to noise is often limited, especially in North American cities. We developed land use regression (LUR) models for LA_eq24h, L_night, and L_den to assess the long-term spatial variability of environmental noise levels in Montreal, Canada, considering various transportation noise sources (road, rail, and air). To explore the effects of sampling duration, we compared our LA_eq24h levels that were computed over at least five complete contiguous days of measurements to shorter sampling periods (20 min and 24 h). LUR models were built with General Additive Models using continuous 2-min noise measurements from 204 sites. Model performance (adjusted R²) was 0.68, 0.59, and 0.69 for LA_eq24h, L_night, and L_den, respectively. Main predictors of measured noise levels were road-traffic and vegetation variables. Twenty-minute non-rush hour measurements corresponded well with LA_eq24h levels computed over 5 days at road-traffic sites (bias: -0.7 dB(A)), but not at rail (-2.1 dB(A)) nor at air (-2.2 dB(A)) sites. Our study provides important insights into the spatial variation of environmental noise levels in a Canadian city. To assess long-term noise levels, sampling strategies should be stratified by noise sources and preferably should include 1 week of measurements at locations exposed to rail and aircraft noise.

Near roadway air pollution across a spatially extensive road and cycling network

This study investigates the variability in near-road concentrations of ultra-fine particles (UFP). Our results are based on a mobile data collection campaign conducted in 2012 in Montreal, Canada using instrumented bicycles and covering approximately 475 km of unique roadways. The spatial extent of the data collected included a diverse array of roads and land use patterns. Average concentrations of UFP per roadway segment varied greatly across the study area (1411-192,340 particles/cm(3)) as well as across the different visits to the same segment. Mixed effects linear regression models were estimated for UFP (R(2) = 43.80%), incorporating a wide range of predictors including land-use, built environment, road characteristics, and meteorology. Temperature and wind speed had a large negative effect on near-road concentrations of UFP. Both the day of the week and time of day had a significant effect with Tuesdays and afternoon periods positively associated with UFP. Since UFP are largely associated with traffic emissions and considering the wide spatial extent of our data collection campaign, it was impossible to collect traffic volume data. For this purpose, we used simulated data for traffic volumes and speeds across the region and observed a positive effect for volumes and negative effect for speed. Finally, proximity to truck routes was also associated with higher UFP concentrations.

An examination of population exposure to traffic related air pollution: Comparing spatially and temporally resolved estimates against long-term average exposures at the home location

Air pollution in metropolitan areas is mainly caused by traffic emissions. This study presents the development of a model chain consisting of a transportation model, an emissions model, and atmospheric dispersion model, applied to dynamically evaluate individuals' exposure to air pollution by intersecting daily trajectories of individuals and hourly spatial variations of air pollution across the study domain. This dynamic approach is implemented in Montreal, Canada to highlight the advantages of the method for exposure analysis. The results for nitrogen dioxide (NO2), a marker of traffic related air pollution, reveal significant differences when relying on spatially and temporally resolved concentrations combined with individuals' daily trajectories compared to a long-term average NO2 concentration at the home location. We observe that NO2 exposures based on trips and activity locations visited throughout the day were often more elevated than daily NO2 concentrations at the home location. The percentage of all individuals with a lower 24-hour daily average at home compared to their 24-hour mobility exposure is 89.6%, of which 31% of individuals increase their exposure by more than 10% by leaving the home. On average, individuals increased their exposure by 23-44% while commuting and conducting activities out of home (compared to the daily concentration at home), regardless of air quality at their home location. We conclude that our proposed dynamic modelling approach significantly improves the results of traditional methods that rely on a long-term average concentration at the home location and we shed light on the importance of using individual daily trajectories to understand exposure.

Operational and environmental determinants of in-vehicle CO and PM2.5 exposure

This study presents a modeling framework to quantify the complex roles that traffic, seasonality, vehicle characteristics, ventilation, meteorology, and ambient air quality play in dictating in-vehicle commuter exposure to CO and PM2.5. For this purpose, a comprehensive one-year monitoring program of 25 different variables was coupled with a multivariate regression analysis to develop models to predict in-vehicle CO and PM2.5 exposure using a database of 119 mobile tests and 120 fume leakage tests. The study aims to improve the understanding of in-cabin exposure, as well as interior-exterior pollutant exchange. Model results highlighted the strong correlation between out-vehicle and in-vehicle concentrations, with the effect of ventilation type only discerned for PM2.5 levels. Car type, road conditions, as well as meteorological conditions all played a significant role in modulating in-vehicle exposure. The CO and PM2.5 exposure models were able to explain 72 and 92% of the variability in measured concentrations, respectively. Both models exhibited robustness and no-evidence of over-fitting.

A land use regression model for ambient ultrafine particles in Montreal, Canada: A comparison of linear regression and a machine learning approach

Existing evidence suggests that ambient ultrafine particles (UFPs) (<0.1µm) may contribute to acute cardiorespiratory morbidity. However, few studies have examined the long-term health effects of these pollutants owing in part to a need for exposure surfaces that can be applied in large population-based studies. To address this need, we developed a land use regression model for UFPs in Montreal, Canada using mobile monitoring data collected from 414 road segments during the summer and winter months between 2011 and 2012. Two different approaches were examined for model development including standard multivariable linear regression and a machine learning approach (kernel-based regularized least squares (KRLS)) that learns the functional form of covariate impacts on ambient UFP concentrations from the data. The final models included parameters for population density, ambient temperature and wind speed, land use parameters (park space and open space), length of local roads and rail, and estimated annual average NOx emissions from traffic. The final multivariable linear regression model explained 62% of the spatial variation in ambient UFP concentrations whereas the KRLS model explained 79% of the variance. The KRLS model performed slightly better than the linear regression model when evaluated using an external dataset (R(2)=0.58 vs. 0.55) or a cross-validation procedure (R(2)=0.67 vs. 0.60). In general, our findings suggest that the KRLS approach may offer modest improvements in predictive performance compared to standard multivariable linear regression models used to estimate spatial variations in ambient UFPs. However, differences in predictive performance were not statistically significant when evaluated using the cross-validation procedure.

Investigating the Use Of Portable Air Pollution Sensors to Capture the Spatial Variability Of Traffic-Related Air Pollution

Advances in microsensor technologies for air pollution monitoring encourage a growing use of portable sensors. This study aims at testing their performance in the development of exposure surfaces for nitrogen dioxide (NO2) and ozone (O3). In Montreal, Canada, a data-collection campaign was conducted across three seasons in 2014 for 76 sites spanning the range of land uses and built environments of the city; each site was visited from 6 to 12 times, for 20 min, using NO2 and O3 sensors manufactured by Aeroqual. Land-use regression models were developed, achieving R(2) values of 0.86 for NO2 and 0.92 for O3 when adjusted for regional meteorology to control for the fact that all of the locations were not monitored at the same time. A total of two exposure surfaces were then developed for NO2 and O3 as averages over spring, summer, and fall. Validation against the fixed-station data and previous campaigns suggests that Aeroqual sensors tend to overestimate the highest NO2 and O3 concentrations, thus increasing the range of values across the city. However, the sensors suggest a good performance with respect to capturing the spatial variability in NO2 and O3 and are very convenient to use, having great potential for capturing temporal variability.

Investigating the Use Of Portable Air Pollution Sensors to Capture the Spatial Variability Of Traffic-Related Air Pollution

Advances in microsensor technologies for air pollution monitoring encourage a growing use of portable sensors. This study aims at testing their performance in the development of exposure surfaces for nitrogen dioxide (NO2) and ozone (O3). In Montreal, Canada, a data-collection campaign was conducted across three seasons in 2014 for 76 sites spanning the range of land uses and built environments of the city; each site was visited from 6 to 12 times, for 20 min, using NO2 and O3 sensors manufactured by Aeroqual. Land-use regression models were developed, achieving R(2) values of 0.86 for NO2 and 0.92 for O3 when adjusted for regional meteorology to control for the fact that all of the locations were not monitored at the same time. A total of two exposure surfaces were then developed for NO2 and O3 as averages over spring, summer, and fall. Validation against the fixed-station data and previous campaigns suggests that Aeroqual sensors tend to overestimate the highest NO2 and O3 concentrations, thus increasing the range of values across the city. However, the sensors suggest a good performance with respect to capturing the spatial variability in NO2 and O3 and are very convenient to use, having great potential for capturing temporal variability.

Characterizing the spatial distribution of ambient ultrafine particles in Toronto, Canada: A land use regression model

Exposure models are needed to evaluate the chronic health effects of ambient ultrafine particles (<0.1 μm) (UFPs). We developed a land use regression model for ambient UFPs in Toronto, Canada using mobile monitoring data collected during summer/winter 2010-2011. In total, 405 road segments were included in the analysis. The final model explained 67% of the spatial variation in mean UFPs and included terms for the logarithm of distances to highways, major roads, the central business district, Pearson airport, and bus routes as well as variables for the number of on-street trees, parks, open space, and the length of bus routes within a 100 m buffer. There was no systematic difference between measured and predicted values when the model was evaluated in an external dataset, although the R(2) value decreased (R(2) = 50%). This model will be used to evaluate the chronic health effects of UFPs using population-based cohorts in the Toronto area.

Investigating the role of transportation models in epidemiologic studies of traffic related air pollution and health effects

In two earlier case-control studies conducted in Montreal, nitrogen dioxide (NO2), a marker for traffic-related air pollution was found to be associated with the incidence of postmenopausal breast cancer and prostate cancer. These studies relied on a land use regression model (LUR) for NO2 that is commonly used in epidemiologic studies for deriving estimates of traffic-related air pollution. Here, we investigate the use of a transportation model developed during the summer season to generate a measure of traffic emissions as an alternative to the LUR model. Our traffic model provides estimates of emissions of nitrogen oxides (NOx) at the level of individual roads, as does the LUR model. Our main objective was to compare the distribution of the spatial estimates of NOx computed from our transportation model to the distribution obtained from the LUR model. A secondary objective was to compare estimates of risk using these two exposure estimates. We observed that the correlation (spearman) between our two measures of exposure (NO2 and NOx) ranged from less than 0.3 to more than 0.9 across Montreal neighborhoods. The most important factor affecting the "agreement" between the two measures in a specific area was found to be the length of roads. Areas affected by a high level of traffic-related air pollution had a far better agreement between the two exposure measures. A comparison of odds ratios (ORs) obtained from NO2 and NOx used in two case-control studies of breast and prostate cancer, showed that the differences between the ORs associated with NO2 exposure vs NOx exposure differed by 5.2-8.8%.

Socioeconomic status and environmental noise exposure in Montreal, Canada

Background

This study’s objective was to determine whether socioeconomically deprived populations are exposed to greater levels of environmental noise.

Methods

Indicators of socioeconomic status were correlated with LAeq24h noise levels estimated with a land-use regression model at a small geographic scale.

Results

We found that noise exposure was associated with all socioeconomic indicators, with the strongest correlations found for median household income, proportion of people who spend over 30% of their income on housing, proportion of people below the low income boundary and with a social deprivation index combining several socio-economic variables.

Conclusion

Our results were inconsistent with a number of studies performed elsewhere, indicating that locally conducted studies are imperative to assessing whether this double burden of noise exposure and low socioeconomic status exists in other contexts. The primary implication of our study is that noise exposure represents an environmental injustice in Montreal, which is an issue that merits both investigation and concern.

Electronic supplementary material

The online version of this article (doi:10.1186/s12889-015-1571-2) contains supplementary material, which is available to authorized users.

Smog and socioeconomics: an evaluation of equity in traffic-related air pollution generation and exposure

How traffic-related air pollution generation and exposure is distributed among different population groups is an important environmental justice concern. From a social equity perspective, many questions arise at the metropolitan scale. Do socially disadvantaged communities have higher exposure levels to traffic-related air pollution? Do discrepancies exist wherein neighborhoods are not exposed to levels of pollution similar to those they themselves generate? And, is there a relationship between this discrepancy and social disadvantage? These questions are examined for the Montreal Metropolitan Region through the development of an integrated transport and emissions model. Two measures of traffic-related air pollution are estimated at the traffic analysis zone level: (1) generation (average emissions per household), and (2) exposure (average residential zone concentration). A social disadvantage index is also calculated that incorporates elements of social and material deprivation. Three levels of inequity exist regarding emissions, exposure, and socioeconomics. Social disadvantage was found to have a positive relationship with exposure, meaning that the most socially disadvantaged communities tend to experience the highest levels of traffic-related air pollution. Spatial discrepancies in emission generation versus emission exposure are also present for most of the metropolitan region. Furthermore, the communities that face a double burden of greater disadvantage and higher exposure also tend to create the lowest quantities of pollution.

Exposure to traffic-related air pollution during physical activity and acute changes in blood pressure, autonomic and micro-vascular function in women: a cross-over study

Background

Traffic-related air pollution may contribute to cardiovascular morbidity. In urban areas, exposures during physical activity are of interest owing to increased breathing rates and close proximity to vehicle emissions.

Methods

We conducted a cross-over study among 53 healthy non-smoking women in Montreal, Canada during the summer of 2013. Women were exposed to traffic pollutants for 2-hours on three separate occasions during cycling on high and low-traffic routes as well as indoors. Personal air pollution exposures (PM_2.5, ultrafine particles (UFP), black carbon, NO₂, and O₃) were evaluated along each route and linear mixed-effects models with random subject intercepts were used to estimate the impact of air pollutants on acute changes in blood pressure, heart rate variability, and micro-vascular function in the hours immediately following exposure. Single and multi-pollutant models were examined and potential effect modification by mean regional air pollution concentrations (PM_2.5, NO₂, and O₃) was explored for the 24-hour and 5-day periods preceding exposure.

Results

In total, 143 exposure routes were completed. Each interquartile increase (10,850/cm³) in UFP exposure was associated with a 4.91% (95% CI: -9.31, -0.512) decrease in reactive hyperemia index (a measure of micro-vascular function) and each 24 ppb increase in O₃ exposure corresponded to a 2.49% (95% CI: 0.141, 4.84) increase in systolic blood pressure and a 3.26% (95% CI: 0.0117, 6.51) increase in diastolic blood pressure 3-hours after exposure. Personal exposure to PM_2.5 was associated with decreases in HRV measures reflecting parasympathetic modulation of the heart and regional PM_2.5 concentrations modified these relationships (p < 0.05). In particular, stronger inverse associations were observed when regional PM_2.5 was higher on the days prior to the study period. Regional PM_2.5 also modified the impact of personal O₃ on the standard deviation of normal to normal intervals (SDNN) (p < 0.05): a significant inverse relationship was observed when regional PM_2.5 was low prior to study periods and a significant positive relationship was observed when regional PM_2.5 was high.

Conclusion

Exposure to traffic pollution may contribute to acute changes in blood pressure, autonomic and micro-vascular function in women. Regional air pollution concentrations may modify the impact of these exposures on autonomic function.

Electronic supplementary material

The online version of this article (doi:10.1186/s12989-014-0070-4) contains supplementary material, which is available to authorized users.

Characterizing the impact of traffic and the built environment on near-road ultrafine particle and black carbon concentrations

BACKGROUND

Increasing evidence suggests that ultrafine particles (UFPs) may contribute to cardiorespiratory morbidity. We examined the relationship between near road UFPs and several traffic and built environment factors to identify predictors that may be used to estimate exposures in population-based studies. Black carbon (BC) was also examined.

METHODS

Data were collected on up to 6 occasions at 73 sites in Montreal, Canada over 6-week period during summer, 2012. After excluding highly correlated variables, road width, truck ratio (trucks/total traffic), building height, land zoning parameters, and meteorological factors were evaluated. Random-effect models were used to estimate percent changes in UFP and BC concentrations with interquartile changes in each candidate predictor adjusted for meteorological factors.

RESULTS

Mean pollutant concentrations varied substantially across sites (UFP range: 1977-94, 798 particles/cm(3); BC range: 29-9460 ng/m(3)). After adjusting for meteorology, interquartile increases in road width (14%, 95% CI: 0, 30), building height (13%, 95% CI: 5, 22), and truck ratio (13%, 95% CI: 3, 23) were the most important predictors of mean UFP concentrations. Road width (28%, 95% CI: 9, 51) and industrial zoning (18%, 95% CI: 2, 37) were the strongest predictors of maximum UFP concentrations. Industrial zoning (35%, 95% CI: 9, 67) was the strongest predictor of BC.

CONCLUSIONS

A number of traffic and built environmental factors were identified as important predictors of near road UFP and BC concentrations. Exposure models incorporating these factors may be useful in evaluating the health effects of traffic related air pollution.

A web-based route planning tool to reduce cyclists' exposures to traffic pollution: a case study in Montreal, Canada

We developed a web-based route planning tool for cyclists in Montreal, Canada, using spatial monitoring data for ambient nitrogen dioxide (NO2). With this tool, we estimated exposures to NO2 along shortest routes and lower exposure alternatives using origin-destination survey data. On average, exposures were estimated to be lower by 0.76 ppb (95% CI: 0.72, 0.80) relative to the shortest route, with decreases of up to 6.1 ppb for a single trip. Cumulative exposure levels (ppb km) decreased by approximately 4%. In general, the benefits of decreased exposure could be achieved with little increase (less than 1 km) in the overall route length.

The impact of traffic volume, composition, and road geometry on personal air pollution exposures among cyclists in Montreal, Canada

Cyclists may experience increased exposure to traffic-related air pollution owing to increased minute ventilation and close proximity to vehicle emissions. The aims of this study were to characterize personal exposures to air pollution among urban cyclists and to identify potential determinants of exposure including the type of cycling lane (separated vs on-road), traffic counts, and meteorological factors. In total, personal air pollution exposure data were collected over 64 cycling routes during morning and evening commutes in Montreal, Canada, over 32 days during the summer of 2011. Measured pollutants included ultrafine particles (UFPs), fine particles (PM(2.5)), black carbon (BC), and carbon monoxide (CO). Counts of diesel vehicles were important predictors of personal exposures to BC, with each 10 vehicle/h increase associated with a 15.0% (95% confidence interval (CI): 5.7%, 24.0%) increase in exposure. Use of separated cycling lanes had less impact on personal exposures with a 12% (95% CI: -43%, 14%) decrease observed for BC and smaller decreases observed for UFPs (mean: -1.3%, 95% CI: -20%, 17%) and CO (mean: -5.6%, 95% CI: -17%, 4%) after adjusting for meteorological factors and traffic counts. On average, PM(2.5) exposure increased 7.8% (95% CI: -17%, 35%) with separate cycling lane use, but this estimate was imprecise and not statistically significant. In general, our findings suggest that diesel vehicle traffic is an important contributor to personal BC exposures and that separate cycling lanes may have a modest impact on personal exposure to some air pollutants. Further evaluation is required, however, as the impact of separate cycling lanes and/or traffic counts on personal exposures may vary between regions.