How low can you go? Air pollution affects mortality at very low levels.

The World Health Organization (WHO) recently released new guidelines for outdoor fine particulate air pollution (PM2.5) recommending an annual average concentration of 5 µg/m3. Yet, our understanding of the concentration-response relationship between outdoor PM2.5 and mortality in this range of near-background concentrations remains incomplete. To address this uncertainty, we conducted a population-based cohort study of 7.1 million adults in one of the world's lowest exposure environments. Our findings reveal a supralinear concentration-response relationship between outdoor PM2.5 and mortality at very low (<5 µg/m3) concentrations. Our updated global concentration-response function incorporating this new information suggests an additional 1.5 million deaths globally attributable to outdoor PM2.5 annually compared to previous estimates. The global health benefits of meeting the new WHO guideline for outdoor PM2.5 are greater than previously assumed and indicate a need for continued reductions in outdoor air pollution around the world.

A multiphase CMAQ version 5.0 adjoint.

We present the development of a multiphase adjoint for the Community Multiscale Air Quality (CMAQ) model, a widely used chemical transport model. The adjoint model provides location- and time-specific gradients that can be used in various applications such as backward sensitivity analysis, source attribution, optimal pollution control, data assimilation, and inverse modeling. The science processes of the CMAQ model include gas-phase chemistry, aerosol dynamics and thermodynamics, cloud chemistry and dynamics, diffusion, and advection. Discrete adjoints are implemented for all the science processes, with an additional continuous adjoint for advection. The development of discrete adjoints is assisted with algorithmic differentiation (AD) tools. Particularly, the Kinetic PreProcessor (KPP) is implemented for gas-phase and aqueous chemistry, and two different automatic differentiation tools are used for other processes such as clouds, aerosols, diffusion, and advection. The continuous adjoint of advection is developed manually. For adjoint validation, the brute-force or finite-difference method (FDM) is implemented process by process with box- or column-model simulations. Due to the inherent limitations of the FDM caused by numerical round-off errors, the complex variable method (CVM) is adopted where necessary. The adjoint model often shows better agreement with the CVM than with the FDM. The adjoints of all science processes compare favorably with the FDM and CVM. In an example application of the full multiphase adjoint model, we provide the first estimates of how emissions of particulate matter (PM2.5) affect public health across the US.

Disease assimilation: The mortality impacts of fine particulate matter on immigrants to Canada.

BACKGROUND: Immigrants make up 20% of the Canadian population; however, little is known about the mortality impacts of fine particulate matter (PM2.5) air pollution on immigrants compared with non-immigrants, or about how impacts may change with duration in Canada. DATA AND METHODS: This study used the 2001 Canadian Census Health and Environment Cohort, a longitudinal cohort of 3.5 million individuals, of which 764,000 were classified as immigrants (foreign-born). Postal codes from annual income tax files were used to account for mobility among respondents and to assign annual PM2.5 concentrations from 1998 to 2016. Exposures were estimated as a three-year moving average prior to the follow-up year. Cox survival models were used to determine hazard ratios (HRs) for cause-specific mortality, comparing the Canadian and foreign-born populations, with further stratification by year of immigration grouped into 10-year cohorts. RESULTS: Differences in urban-rural settlement patterns resulted in greater exposure to PM2.5 for immigrants compared with non-immigrants (mean = 9.3 vs. 7.5 µg/m3), with higher exposures among more recent immigrants. In fully adjusted models, immigrants had higher HRs per 10 µg/m3 increase in PM2.5 concentration compared with Canadian-born individuals for cardiovascular mortality (HR [95% confidence interval] = 1.22 [1.12 to 1.34] vs. 1.12 [1.07 to 1.18]) and cerebrovascular mortality (HR = 1.25 [1.03 to 1.52] vs. 1.03 [0.93 to 1.15]), respectively. However, tests for differences between the two groups were not significant when Cochran's Q test was used. No significant associations were found for respiratory outcomes, except for lung cancer in non-immigrants (HR = 1.10 [1.02 to 1.18]). When stratified by year of immigration, differences in HRs across varied by cause of death. DISCUSSION: In Canada, PM2.5 is an equal-opportunity risk factor, with immigrants experiencing similar if not higher mortality risks compared with non-immigrants for cardiovascular-related causes of death. Some notable differences also existed with cerebrovascular and lung cancer deaths. Continued reductions in air pollution, particularly in urban areas, will improve the health of the Canadian population as a whole.

Evaluating the Sensitivity of PM2.5-Mortality Associations to the Spatial and Temporal Scale of Exposure Assessment.

BACKGROUND: The temporal and spatial scales of exposure assessment may influence observed associations between fine particulate air pollution (PM2.5) and mortality, but few studies have systematically examined this question. METHODS: We followed 2.4 million adults in the 2001 Canadian Census Health and Environment Cohort for nonaccidental and cause-specific mortality between 2001 and 2011. We assigned PM2.5 exposures to residential locations using satellite-based estimates and compared three different temporal moving averages (1, 3, and 8 years) and three spatial scales (1, 5, and 10 km) of exposure assignment. In addition, we examined different spatial scales based on age, employment status, and urban/rural location, and adjustment for O3, NO2, or their combined oxidant capacity (Ox). RESULTS: In general, longer moving averages resulted in stronger associations between PM2.5 and mortality. For nonaccidental mortality, we observed a hazard ratio of 1.11 (95% CI = 1.08, 1.13) for the 1-year moving average compared with 1.23 (95% CI = 1.20, 1.27) for the 8-year moving average. Respiratory and lung cancer mortality were most sensitive to the spatial scale of exposure assessment with stronger associations observed at smaller spatial scales. Adjustment for oxidant gases attenuated associations between PM2.5 and cardiovascular mortality and strengthened associations with lung cancer. Despite these variations, PM2.5 was associated with increased mortality in nearly all of the models examined. CONCLUSIONS: These findings support a relationship between outdoor PM2.5 and mortality at low concentrations and highlight the importance of longer-exposure windows, more spatially resolved exposure metrics, and adjustment for oxidant gases in characterizing this relationship.

Examining the Shape of the Association between Low Levels of Fine Particulate Matter and Mortality across Three Cycles of the Canadian Census Health and Environment Cohort.

BACKGROUND: Ambient fine particulate air pollution with aerodynamic diameter ≤2.5 µm (PM2.5) is an important contributor to the global burden of disease. Information on the shape of the concentration-response relationship at low concentrations is critical for estimating this burden, setting air quality standards, and in benefits assessments. OBJECTIVES: We examined the concentration-response relationship between PM2.5 and nonaccidental mortality in three Canadian Census Health and Environment Cohorts (CanCHECs) based on the 1991, 1996, and 2001 census cycles linked to mobility and mortality data. METHODS: Census respondents were linked with death records through 2016, resulting in 8.5 million adults, 150 million years of follow-up, and 1.5 million deaths. Using annual mailing address, we assigned time-varying contextual variables and 3-y moving-average ambient PM2.5 at a 1×1 km spatial resolution from 1988 to 2015. We ran Cox proportional hazards models for PM2.5 adjusted for eight subject-level indicators of socioeconomic status, seven contextual covariates, ozone, nitrogen dioxide, and combined oxidative potential. We used three statistical methods to examine the shape of the concentration-response relationship between PM2.5 and nonaccidental mortality. RESULTS: The mean 3-y annual average estimate of PM2.5 exposure ranged from 6.7 to 8.0 µg/m3 over the three cohorts. We estimated a hazard ratio (HR) of 1.053 [95% confidence interval (CI): 1.041, 1.065] per 10-µg/m3 change in PM2.5 after pooling the three cohort-specific hazard ratios, with some variation between cohorts (1.041 for the 1991 and 1996 cohorts and 1.084 for the 2001 cohort). We observed a supralinear association in all three cohorts. The lower bound of the 95% CIs exceeded unity for all concentrations in the 1991 cohort, for concentrations above 2 µg/m3 in the 1996 cohort, and above 5 µg/m3 in the 2001 cohort. DISCUSSION: In a very large population-based cohort with up to 25 y of follow-up, PM2.5 was associated with nonaccidental mortality at concentrations as low as 5 µg/m3. https://doi.org/10.1289/EHP5204.

Low concentrations of fine particle air pollution and mortality in the Canadian Community Health Survey cohort.

BACKGROUND: Approximately 2.9 million deaths are attributed to ambient fine particle air pollution around the world each year (PM2.5). In general, cohort studies of mortality and outdoor PM2.5 concentrations have limited information on individuals exposed to low levels of PM2.5 as well as covariates such as smoking behaviours, alcohol consumption, and diet which may confound relationships with mortality. This study provides an updated and extended analysis of the Canadian Community Health Survey-Mortality cohort: a population-based cohort with detailed PM2.5 exposure data and information on a number of important individual-level behavioural risk factors. We also used this rich dataset to provide insight into the shape of the concentration-response curve for mortality at low levels of PM2.5. METHODS: Respondents to the Canadian Community Health Survey from 2000 to 2012 were linked by postal code history from 1981 to 2016 to high resolution PM2.5 exposure estimates, and mortality incidence to 2016. Cox proportional hazard models were used to estimate the relationship between non-accidental mortality and ambient PM2.5 concentrations (measured as a three-year average with a one-year lag) adjusted for socio-economic, behavioural, and time-varying contextual covariates. RESULTS: In total, 50,700 deaths from non-accidental causes occurred in the cohort over the follow-up period. Annual average ambient PM2.5 concentrations were low (i.e. 5.9 µg/m3, s.d. 2.0) and each 10 µg/m3 increase in exposure was associated with an increase in non-accidental mortality (HR = 1.11; 95% CI 1.04-1.18). Adjustment for behavioural covariates did not materially change this relationship. We estimated a supra-linear concentration-response curve extending to concentrations below 2 µg/m3 using a shape constrained health impact function. Mortality risks associated with exposure to PM2.5 were increased for males, those under age 65, and non-immigrants. Hazard ratios for PM2.5 and mortality were attenuated when gaseous pollutants were included in models. CONCLUSIONS: Outdoor PM2.5 concentrations were associated with non-accidental mortality and adjusting for individual-level behavioural covariates did not materially change this relationship. The concentration-response curve was supra-linear with increased mortality risks extending to low outdoor PM2.5 concentrations.

Evaluation of a method to indirectly adjust for unmeasured covariates in the association between fine particulate matter and mortality.

BACKGROUND: Indirect adjustment via partitioned regression is a promising technique to control for unmeasured confounding in large epidemiological studies. The method uses a representative ancillary dataset to estimate the association between variables missing in a primary dataset with the complete set of variables of the ancillary dataset to produce an adjusted risk estimate for the variable in question. The objective of this paper is threefold: 1) evaluate the method for non-linear survival models, 2) formalize an empirical process to evaluate the suitability of the required ancillary matching dataset, and 3) test modifications to the method to incorporate time-varying exposure data, and proportional weighting of datasets. METHODS: We used the association between fine particle air pollution (PM2.5) with mortality in the 2001 Canadian Census Health and Environment Cohort (CanCHEC, N = 2.4 million, 10-years follow-up) as our primary dataset, and the 2001 cycle of the Canadian Community Health Survey (CCHS, N = 80,630) as the ancillary matching dataset that contained confounding risk factor information not available in CanCHEC (e.g., smoking). The main evaluation process used a gold-standard approach wherein two variables (education and income) available in both datasets were excluded, indirectly adjusted for, and compared to true models with education and income included to assess the amount of bias correction. An internal validation for objective 1 used only CanCHEC data, whereas an external validation for objective 2 replaced CanCHEC with the CCHS. The two proposed modifications were applied as part of the validation tests, as well as in a final indirect adjustment of four missing risk factor variables (smoking, alcohol use, diet, and exercise) in which adjustment direction and magnitude was compared to models using an equivalent longitudinal cohort with direct adjustment for the same variables. RESULTS: At baseline (2001) both cohorts had very similar PM2.5 distributions across population characteristics, although levels for CCHS participants were consistently 1.8-2.0 µg/m3 lower. Applying sample-weighting largely corrected for this discrepancy. The internal validation tests showed minimal downward bias in PM2.5 mortality hazard ratios of 0.4-0.6% using a static exposure, and 1.7-3% when a time-varying exposure was used. The external validation of the CCHS as the ancillary dataset showed slight upward bias of -0.7 to -1.1% and downward bias of 1.3-2.3% using the static and time-varying approaches respectively. CONCLUSIONS: The CCHS was found to be fairly well representative of CanCHEC and its use in Canada for indirect adjustment is warranted. Indirect adjustment methods can be used with survival models to correct hazard ratio point estimates and standard errors in models missing key covariates when a representative matching dataset is available. The results of this formal evaluation should encourage other cohorts to assess the suitability of ancillary datasets for the application of the indirect adjustment methodology to address potential residual confounding.

Diminishing Returns or Compounding Benefits of Air Pollution Control? The Case of NOx and Ozone.

UNLABELLED: A common measure used in air quality benefit-cost assessment is marginal benefit (MB), or the monetized societal benefit of reducing 1 ton of emissions. Traditional depictions of MB for criteria air pollutants are such that each additional ton of emission reduction incurs less benefit than the previous ton. Using adjoint sensitivity analysis in a state-of-the-art air quality model, we estimate MBs for NOx emitted from mobile and point sources, characterized based on the estimated ozone-related premature mortality in the U.S. POPULATION: Our findings indicate that nation-wide emission reductions in the U.S. significantly increase NOx MBs for all sources, without exception. We estimate that MBs for NOx emitted from mobile sources increase by 1.5 and 2.5 times, on average, for 40% and 80% reductions in anthropogenic emissions across the U.S. Our results indicate a strictly concave damage function and compounding benefits of progressively lower levels of NOx emissions, providing economic incentive for higher levels of abatement than were previously advisible. These findings suggest that the traditional perception of a convex damage function and decreasing MB with abatement may not hold true for secondary pollutants such as O3.

Attainment vs exposure: ozone metric responses to source-specific NOx controls using adjoint sensitivity analysis.

We establish linkages between sources of NOx emissions and two types of national ozone metrics in Canada and the U.S. using the adjoint of an air quality model. We define an attainment-based metric using probabilistic design values (PDVs) exceeding 65 ppb to represent polluted regions and define an exposure-based metric as the premature mortality count related to short-term ozone exposure, both in Canada and the U.S. Our results reveal differences in both temporally averaged and day-specific influences of NOx emission controls across source locations. We find NOx emission reductions in California and the eastern U.S. to be most effective for reducing attainment- and exposure-based metrics, amounting to a total reduction of 6500 ppb in PDVs and 613 deaths/season nationally from a 10% reduction in NOx emissions from those source locations. While source controls in the remainder of the western U.S. are beneficial at reducing nonattainment, these reductions are less influential on ozone mortality. We also find that while exposure-based metrics are sensitive to daily emission reductions, much of the reduction in PDVs arises from controlling emissions on only a fraction of simulation days. We further illustrate the dependency of adjoint estimates of emission influences on the choice of averaging period as a follow-up to previous work.

Source attribution of health benefits from air pollution abatement in Canada and the United States: an adjoint sensitivity analysis.

BACKGROUND: Decision making regarding air pollution can be better informed if air quality impacts are traced back to individual emission sources. Adjoint or backward sensitivity analysis is a modeling tool that can achieve this goal by allowing for quantification of how emissions from sources in different locations influence human health metrics. OBJECTIVES: We attributed short-term mortality (valuated as an overall "health benefit") in Canada and the United States to anthropogenic nitrogen oxides (NO(x)) and volatile organic compound (VOC) emissions across North America. METHODS: We integrated epidemiological data derived from Canadian and U.S. time-series studies with the adjoint of an air quality model and also estimated influences of anthropogenic emissions at each location on nationwide health benefits. RESULTS: We found significant spatiotemporal variability in estimated health benefit influences of NO(x) and VOC emission reductions on Canada and U.S. mortality. The largest estimated influences on Canada (up to $250,000/day) were from emissions originating in the Quebec City-Windsor Corridor, where population centers are concentrated. Estimated influences on the United States tend to be widespread and more substantial owing to both larger emissions and larger populations. The health benefit influences calculated using 24-hr average ozone (O(3)) concentrations are lower in magnitude than estimates calculated using daily 1-hr maximum O(3) concentrations. CONCLUSIONS: Source specificity of the adjoint approach provides valuable information for guiding air quality decision making. Adjoint results suggest that the health benefits of reducing NO(x) and VOC emissions are substantial and highly variable across North America.