Association of Particulate Matter Exposure With Lung Function and Mortality Among Patients With Fibrotic Interstitial Lung Disease

Gillian C Goobie; Christopher Carlsten; Kerri A Johannson; Nasreen Khalil; Veronica Marcoux; Deborah Assayag; Hélène Manganas; Jolene H Fisher; Martin R J Kolb; Kathleen O Lindell; James P Fabisiak; Xiaoping Chen; Kevin F Gibson; Yingze Zhang; Daniel J Kass; Christopher J Ryerson; S Mehdi Nouraie

doi:10.1001/jamainternmed.2022.4696

. 2022 Oct 17;182(12):1248–1259. doi: 10.1001/jamainternmed.2022.4696

Association of Particulate Matter Exposure With Lung Function and Mortality Among Patients With Fibrotic Interstitial Lung Disease

Gillian C Goobie ^1,^2,^3,^✉, Christopher Carlsten ^4,⁵, Kerri A Johannson ⁶, Nasreen Khalil ⁴, Veronica Marcoux ⁷, Deborah Assayag ⁸, Hélène Manganas ⁹, Jolene H Fisher ¹⁰, Martin R J Kolb ¹¹, Kathleen O Lindell ^2,¹², James P Fabisiak ¹³, Xiaoping Chen ^2,¹⁴, Kevin F Gibson ^2,¹⁴, Yingze Zhang ¹⁴, Daniel J Kass ^2,¹⁴, Christopher J Ryerson ^4,¹⁵, S Mehdi Nouraie ¹⁴

¹Department of Human Genetics, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania

²Simmons Center for Interstitial Lung Disease, Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania

³Clinician Investigator Program, Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada

⁴Division of Respiratory Medicine, Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada

⁵Air Pollution Exposure Laboratory, Vancouver Coastal Health Research Institute, Vancouver, British Columbia, Canada

⁶Division of Respiratory Medicine, Department of Medicine, University of Calgary, Calgary, Alberta, Canada

⁷Division of Respirology, Critical Care, and Sleep Medicine, College of Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

⁸Division of Respiratory Medicine, Department of Medicine, McGill University, Montreal, Quebec, Canada

⁹Département de Médecine, Centre Hospitalier de l’Université de Montréal, Montreal, Quebec, Canada

¹⁰Division of Respirology, Department of Medicine, University of Toronto, Toronto, Ontario, Canada

¹¹Department of Medicine, Firestone Institute for Respiratory Health, The Research Institute of St Joe’s Hamilton, St Joseph’s Healthcare, McMaster University, Hamilton, Ontario, Canada

¹²College of Nursing, Medical University of South Carolina, Charleston

¹³Department of Environmental and Occupational Health, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania

¹⁴Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania

¹⁵Centre for Heart Lung Innovation, St Paul’s Hospital, Vancouver, British Columbia, Canada

Accepted for Publication: August 23, 2022.

Published Online: October 17, 2022. doi:10.1001/jamainternmed.2022.4696

Correction: This article was corrected on December 5, 2022, to fix an error in Figure 2.

^✉

Corresponding Author: Gillian C. Goobie, MD, Department of Human Genetics, Graduate School of Public Health, University of Pittsburgh, 130 DeSoto St, Pittsburgh, PA 15261 (goobiegc@upmc.edu).

Author Contributions: Drs Goobie and Nouraie had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Ryerson and Nouraie served as co–senior authors.

Concept and design: Goobie, Carlsten, Johannson, Kolb, Fabisiak, Gibson, Zhang, Kass, Ryerson, Nouraie.

Acquisition, analysis, or interpretation of data: Goobie, Johannson, Khalil, Marcoux, Assayag, Manganas, Fisher, Lindell, Fabisiak, Chen, Gibson, Zhang, Kass, Ryerson, Nouraie.

Drafting of the manuscript: Goobie, Carlsten, Johannson, Lindell, Gibson, Kass, Nouraie.

Critical revision of the manuscript for important intellectual content: Goobie, Carlsten, Johannson, Khalil, Marcoux, Assayag, Manganas, Fisher, Kolb, Fabisiak, Chen, Zhang, Kass, Ryerson, Nouraie.

Statistical analysis: Goobie, Nouraie.

Obtained funding: Goobie, Kolb, Fabisiak, Ryerson, Nouraie.

Administrative, technical, or material support: Goobie, Carlsten, Khalil, Marcoux, Manganas, Fisher, Kolb, Lindell, Fabisiak, Chen, Gibson, Zhang, Kass, Ryerson.

Supervision: Carlsten, Johannson, Fabisiak, Gibson, Zhang, Kass, Ryerson, Nouraie.

Conflict of Interest Disclosures: Dr Goobie reported receiving grants from Boehringer Ingelheim and the Pulmonary Fibrosis Foundation (PFF) during the conduct of the study. Dr Johannson reported receiving nonfinancial support and personal fees from Boehringer Ingelheim; personal fees from Hoffman-La Roche, Pliant Therapeutics, the Three Lakes Foundation, and Thyron SAB; grants from the University of Calgary Cumming School of Medicine, the Three Lakes Foundation, and the Pulmonary Fibrosis Society of Calgary; and grants from the University Hospital Foundation outside the submitted work. Dr Marcoux reported receiving grants from Boehringer Ingelheim Canada outside the submitted work and during the conduct of the study; grants and personal fees from Hoffman-La Roche and AstraZeneca; personal fees from Boehringer Ingelheim Canada and Roche; and grants from the University of Saskatchewan, the Royal University Hospital Foundation, and Roche. Dr Assayag reported receiving personal fees and grants from Boehringer Ingelheim and Novartis and personal fees from Hoffman-La Roche outside the submitted work. Dr Manganas reported receiving grants from Hoffman-La Roche, Galapagos Pharma, and BMS Pharma outside the submitted work; grants from the University of British Columbia during the conduct of the study; grants from Boehringer Ingelheim Pharma; honoraria for educational events from the Boehringer Ingelheim Advisory Board; and personal fees from Boehringer Ingelheim Pharma. Dr Fisher reported receiving grants from the Canadian PFF and Boehringer Ingelheim during the conduct of the study (Boehringer Ingelheim provides funding to the University of British Columbia for the Canadian Registry for Pulmonary Fibrosis [CARE-PF]; the University of British Columbia subsequently provides funding to each CARE-PF site) and personal fees from Boehringer Ingelheim and AstraZeneca outside of the submitted work. Dr Kolb reported receiving grants from Boehringer Ingelheim during the conduct of the study; personal fees from Boehringer Ingelheim, Algernon, Hoffman-La Roche, Horizon, CSL Behring, LabCorp, Nitto Denko, BMS, United Therapeutics, and Bellerophon Therapeutics outside the submitted work; and personal fees and grants from GSK, Gilead, Actelion, RespiVert, Genoa, Alkermes, Pharmaxis, Prometric, Indalo, and Third Pole outside the submitted work. Dr Fabisiak reported receiving funding from the Heinz Endowments. Dr Zhang reported receiving a grant from the National Institutes of Health (NIH) and grants from Boehringer Ingelheim and PFF during the conduct of the study. Dr Kass reported receiving grants from the NIH, grants from Boehringer Ingelheim and PFF during the conduct of the study, and collaborative research funding from Regeneron Pharmaceuticals outside the submitted work. Dr Ryerson reported receiving personal fees and grants from Boehringer Ingelheim and Hoffman-La Roche outside the submitted work. Dr Nouraie reported receiving a grant from Boehringer Ingelheim during the conduct of the study. No other disclosures were reported.

Funding/Support: This was an independent, investigator-initiated study supported by Boehringer Ingelheim Pharmaceuticals, Inc (BIPI). Dr Goobie receives research funding and support through the PFF Scholars Award Program and the University of British Columbia Clinician Investigator Program. Dr Carlsten is supported by the Canada Research Chair program.

Role of the Funder/Sponsor: BIPI had no role in the design, analysis, or interpretation of the results in this study. BIPI was given the opportunity to review the manuscript for medical and scientific accuracy as it relates to BIPI substances, as well as intellectual property considerations.

Additional Contributions: We would like to thank the patients from the Dorothy P. and Richard P. Simmons Center for Interstitial Lung Disease and each of the CARE-PF registry sites for their time and commitment to clinical research; all patients who participated in the PFF Registry; the investigators and other staff at participating PFF care centers for providing clinical data; the PFF, which established and has maintained the PFF Registry since 2016; and the many generous donors of the PFF Registry.

Additional Information: Because of the restrictions of the consent forms signed by participants in the Simmons Center for Interstitial Lung Disease Registry, PFF, and CARE-PF, individual participant data will not be made available. The code repository for this manuscript, containing code for exposure matching, statistical analysis, and figure creation, is available from https://github.com/gcgoobie/PM2.5_ClinicalOutcomes. Further information regarding registry protocols, consent forms, or other specific data can be made available on request by contacting the corresponding author, Dr Goobie (goobiegc@upmc.edu or gcgoobie@alumni.ubc.ca).

^✉

Corresponding author.

PMCID: PMC9577882 PMID: 36251286

Key Points

Question

Is exposure to particulate matter 2.5 μm or less in diameter (PM_2.5) and its constituents associated with clinical outcomes of patients with fibrotic interstitial lung disease (fILD)?

Findings

In this cohort study of 6683 patients with fILD, exposure to PM_2.5 varied greatly across North America, with patients in western Pennsylvania exposed to the most PM_2.5 and its human-derived constituents. Increasing exposure to PM_2.5 constituent mixtures, particularly sulfate, nitrate, and ammonium, was consistently associated with worse baseline lung function, more rapid disease progression, and increased mortality.

Meaning

The findings of this study suggest that exposure to PM_2.5 was associated with adverse outcomes among patients with fILD with regard to baseline severity, disease progression, and mortality.

Abstract

Importance

Particulate matter 2.5 μm or less in diameter (PM_2.5) is associated with adverse outcomes for patients with idiopathic pulmonary fibrosis, but its association with other fibrotic interstitial lung diseases (fILDs) and the association of PM_2.5 composition with adverse outcomes remain unclear.

Objective

To investigate the association of PM_2.5 exposure with mortality and lung function among patients with fILD.

Design, Setting, and Participants

In this multicenter, international, prospective cohort study, patients were enrolled in the Simmons Center for Interstitial Lung Disease Registry at the University of Pittsburgh in Pittsburgh, Pennsylvania; 42 sites of the Pulmonary Fibrosis Foundation Registry; and 8 sites of the Canadian Registry for Pulmonary Fibrosis. A total of 6683 patients with fILD were included (Simmons, 1424; Pulmonary Fibrosis Foundation, 1870; and Canadian Registry for Pulmonary Fibrosis, 3389). Data were analyzed from June 1, 2021, to August 2, 2022.

Exposures

Exposure to PM_2.5 and its constituents was estimated with hybrid models, combining satellite-derived aerosol optical depth with chemical transport models and ground-based PM_2.5 measurements.

Main Outcomes and Measures

Multivariable linear regression was used to test associations of exposures 5 years before enrollment with baseline forced vital capacity and diffusion capacity for carbon monoxide. Multivariable Cox models were used to test associations of exposure in the 5 years before censoring with mortality, and linear mixed models were used to test associations of exposure with a decrease in lung function. Multiconstituent analyses were performed with quantile-based g-computation. Cohort effect estimates were meta-analyzed. Models were adjusted for age, sex, smoking history, race, a socioeconomic variable, and site (only for Pulmonary Fibrosis Foundation and Canadian Registry for Pulmonary Fibrosis cohorts).

Results

Median follow-up across the 3 cohorts was 2.9 years (IQR, 1.5-4.5 years), with death for 28% of patients and lung transplant for 10% of patients. Of the 6683 patients in the cohort, 3653 were men (55%), 205 were Black (3.1%), and 5609 were White (84.0%). Median (IQR) age at enrollment across all cohorts was 66 (58-73) years. A PM_2.5 exposure of 8 μg/m³ or more was associated with a hazard ratio for mortality of 4.40 (95% CI, 3.51-5.51) in the Simmons cohort, 1.71 (95% CI, 1.32-2.21) in the Pulmonary Fibrosis Foundation cohort, and 1.45 (95% CI, 1.18-1.79) in the Canadian Registry for Pulmonary Fibrosis cohort. Increasing exposure to sulfate, nitrate, and ammonium PM_2.5 constituents was associated with increased mortality across all cohorts, and multiconstituent models demonstrated that these constituents tended to be associated with the most adverse outcomes with regard to mortality and baseline lung function. Meta-analyses revealed consistent associations of exposure to sulfate and ammonium with mortality and with the rate of decrease in forced vital capacity and diffusion capacity of carbon monoxide and an association of increasing levels of PM_2.5 multiconstituent mixture with all outcomes.

Conclusions and Relevance

This cohort study found that exposure to PM_2.5 was associated with baseline severity, disease progression, and mortality among patients with fILD and that sulfate, ammonium, and nitrate constituents were associated with the most harm, highlighting the need for reductions in human-derived sources of pollution.

This cohort study assesses whether exposure to particulate matter 2.5 μm or less in diameter and its constituents was associated with increased mortality, worse baseline lung function, and more rapid decrease in lung function among patients with fibrotic interstitial lung disease.

Introduction

Fibrotic interstitial lung diseases (fILDs) are a group of pulmonary conditions characterized by dyspnea, radiographic pulmonary fibrosis, and high morbidity and mortality.^1,2 Idiopathic pulmonary fibrosis (IPF) is the most common and severe form of fILD, whose etiology remains incompletely understood.¹ Air pollution is associated with IPF development and progression^3,4,5,6; however, the association of particulate matter 2.5 μm or less in diameter (PM_2.5) with outcomes among patients with diverse forms of fILD remains unclear.⁷ Furthermore, to our knowledge, the association between specific PM_2.5 constituents and these outcomes has never been explored. Given the severity and complex etiology of fILDs, there exists an urgent need to understand how environmental factors are associated with these diseases.

Made up of a miasma of fine airborne particles that exist in the atmosphere alongside gaseous pollutants, PM_2.5 is estimated to be responsible for 4.2 to 8.9 million premature deaths annually.^8,9 Satellite-derived hybrid models can estimate ambient PM_2.5 levels across the globe, with recent approaches enabling the speciation of PM_2.5 into constituent components, including sulfate, nitrate, ammonium, black carbon (BC), organic matter (OM), sea salt, and soil.^10,11 Constituents primarily derived from anthropogenic (ie, human-derived) sources include sulfate, nitrate, and ammonium. Complex reactions of gaseous emissions from fossil fuel combustion, industrial activities (eg, steel production), and agriculture result in the formation of ammonium sulfate, ammonium nitrate, and acidic sulfate and nitrate particles,^12,13,14 with acidic particles associated with some of the most adverse human health outcomes.^15,16 Black carbon is derived from multiple sources, including fossil fuel and wood combustion from anthropogenic and natural sources.¹⁷ Sea salt, soil, and OM constituents are components of normal atmospheric composition.¹⁸

We sought to evaluate the association of exposure to PM_2.5 and its constituents with the health outcomes among patients with fILD, using a multinational cohort of patients with fILD from across the United States and Canada. We hypothesized that higher levels of PM_2.5 and its constituents from anthropogenic sources (sulfate, nitrate, ammonium, and BC) would be associated with lower baseline lung function, more rapid decrease in lung function, and increased mortality. To our knowledge, this work reflects the largest, most geographically diverse evaluation of the associations between air pollution and fILD and is the first study to evaluate the association between PM_2.5 constituents and health outcomes in this population.

Methods

Study Populations

Adult patients with fILD whose diagnoses were made by specialist ILD physicians according to current clinical practice guidelines and the best available evidence were eligible for inclusion in this cohort study.^1,19 Nonfibrotic ILDs were excluded. The first US cohort (“Simmons”) included patients with fILD who were prospectively enrolled in the University of Pittsburgh Dorothy P. and Richard P. Simmons Center for Interstitial Lung Disease Registry in Pittsburgh, Pennsylvania, between 2000 and 2021. The second US cohort (“PFF”) included patients with fILD who were prospectively enrolled in 1 of 42 Pulmonary Fibrosis Foundation registry sites between 2016 and 2021.²⁰ Overlapping patients in the Simmons and PFF cohorts were excluded from the PFF cohort. The Canadian cohort (“CARE-PF”) included patients with fILD who were prospectively enrolled in 1 of 8 Canadian Registry for Pulmonary Fibrosis sites between 2015 and 2021²¹ and patients previously enrolled in single-center registries at CARE-PF sites before 2015. Ethics approval was obtained from the University of Pittsburgh and the University of British Columbia. Oral and written consent was provided by patients enrolled in all 3 registry studies. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Demographic Characteristics, Residential Data, and Clinical Outcomes

Patient demographic characteristics (age at enrollment, sex, smoking history, and self-reported race [ethnicity, which is collected differently from race and refers primarily in these studies to Hispanic vs non-Hispanic, was not adjusted for because of a high degree of missing data in all 3 cohorts]), most recent residential address (or 5-digit zip code for the PFF cohort), specific fILD diagnosis, and lung function (height- and weight-adjusted percentage estimated forced vital capacity [FVC] and diffusion capacity of the lung for carbon monoxide [D_LCO]) were discerned from electronic health records. Simmons data were in part extracted using the University of Pittsburgh Health Record Research Request Service.²² Baseline FVC and D_LCO were defined as the first tests performed within 6 months of enrollment. All FVC and D_LCO measurements obtained throughout follow-up were collected. Date of death and lung transplant were confirmed through periodic extraction of these data from electronic health records by each site’s registry managers. Patients were considered lost to follow-up if they had not died, had received a lung transplant, or had their registry record updated within 1 year of the censorship date (January 27, 2021, for Simmons; March 23, 2021, for CARE-PF). Details on losses to follow-up were not made available by the PFF registry (censor date July 15, 2021).

The most recent residential address was geocoded into latitude and longitude coordinates with ArcGIS Pro version 2.5 (Esri Inc). Patients from the PFF cohort were assigned the centroid coordinates of their 5-digit zip code. Residential location was used to determine a socioeconomic variable, calculating the area deprivation index for the Simmons cohort, the Canadian Index of Multiple Deprivation for the CARE-PF cohort,^23,24,25 and the percentage of individuals in the 5-digit zip code below the poverty level for the PFF cohort (based on US Census data).²⁶

Determination of Exposure to Particulate Matter and Its Constituents

Estimates of monthly mean total PM_2.5 and constituent (sulfate, nitrate, ammonium, BC, OM, sea salt, and soil) mass (micrograms per meter cubed) were acquired from the Atmospheric Composition Analysis Group online repository from 2000 to 2018 for PM_2.5 and from 2000 to 2017 for its constituents.^10,11 These data provide estimates for PM_2.5 and constituents at 0.01° by 0.01° (approximately 1.1 km²) across North America based on satellite-derived aerosol optical depth measurements combined with chemical transport models and ground-based measurements. Total PM_2.5 mass and the mass of the constituents sulfate, nitrate, ammonium, BC, OM, sea salt, and soil were 10-fold cross-validated with ground-based measurements, demonstrating R² (and root-mean-square error) values of 0.70 (1.6) for total PM_2.5, 0.96 (0.3) for sulfate, 0.90 (0.3) for nitrate, 0.86 (0.2) for ammonium, 0.59 (0.1) for BC, 0.57 (0.8) for OM, 0.80 (0.1) for sea salt, and 0.60 (0.2) for soil.¹⁰ Residential coordinates were matched to the nearest coordinates of pollutant data with the ncdf4 package in R, version 4.0.2 (R Project for Statistical Computing). Mean exposures for each patient were determined 5 years before censoring (with censoring defined as death, lung transplant, or cessation of follow-up) for analyses of mortality and decrease in lung function. Mean exposures were determined 5 years before enrollment for analyses of baseline lung function and as a sensitivity analysis for decrease in lung function.

Exposure to PM_2.5 was evaluated continuously and was dichotomized as low vs high exposure (<8 μg/m³ or ≥8 μg/m³) according to American Thoracic Society recommendations for yearly mean PM_2.5 exposures.²⁷ Constituents were dichotomized according to their median value across the 3 cohorts.

Statistical Analysis

Survival analyses were performed with Cox proportional hazards regressions, considering death and lung transplant as a composite outcome. Assumptions were checked with Schoenfeld residuals. Spline models were constructed to evaluate the hazard ratio (HR) for mortality across different ranges of PM_2.5 and its constituents across each cohort. Associations of PM_2.5 and its constituents with baseline FVC and D_LCO were evaluated with multivariable linear regression. Linear mixed-effects models with random intercepts and slopes for each patient were used to evaluate associations with rate of change in FVC or D_LCO.

Adjusted models included covariates of age at enrollment, sex, smoking history, race, area deprivation index for the Simmons cohort, percentage of individuals below the poverty level for the PFF cohort, Canadian Index of Multiple Deprivation for the CARE-PF cohort, and site (PFF and CARE-PF cohorts only). Adjustments for age, sex, baseline lung function, and neighborhood-level disadvantage were made because these factors are associated with mortality among patients with fILDs^23,28; adjustment for race was made given its association with pulmonary function and mortality²⁹; adjustment for smoking was made given its association with lung function and potentially with mortality³⁰; and adjustment for site was made to control for site-specific associations.

Cohort-specific attributable risk fractions were calculated to determine the proportion of mortality attributable to PM_2.5 or anthropogenic constituent exposure, using the following formula: cohort attributable risk fraction = P_e(HR − 1)/[1 + P_e(HR − 1)], where P_e is the prevalence of high pollutant exposure in the cohort and HR is the HR for high exposure in the fully adjusted models. Multiconstituent analyses of mortality and baseline lung function outcomes were performed with quantile-based g-computation with a linear additive approach for the addition of each PM_2.5 constituent, as has been previously used with this exposure-matching approach.³¹ A random-effects meta-analysis of effect estimates across the 3 cohorts was performed for all primary outcomes, with I² values for heterogeneity reported.

Subgroup and sensitivity analyses are provided in the Supplement (eTables 7-9), which includes subgroup results for patients with IPF, a sensitivity analysis of mortality before and after 2015 (in which 2015 is the median year of enrollment across the 3 cohorts) to account for time-varying confounding, and a sensitivity analysis of the association between PM_2.5 exposure and mortality in 5-year precensoring mean values of warm-month (April to September) vs cold-month (October to March) exposures to account for seasonality. Data were analyzed from June 1, 2021, to August 2, 2022. Analyses were performed with R, version 4.0.2.³²

Results

Baseline Patient Characteristics and Pollutant Exposures

Eligibility was met by 1424 patients with fILD in the Simmons cohort, 1870 in the PFF cohort, and 3389 in the CARE-PF cohort, who were followed up for a median of 3.1, 2.5, and 3.2 years, respectively. The median follow-up across the 3 cohorts was 2.9 years (IQR, 1.5-4.5 years), with death for 28% of patients and lung transplant for 10% of patients. Baseline characteristics are shown in Table 1, with exposure breakdowns by site in eTable 1 in the Supplement. Of the 6683 patients in the cohort, 3653 were men (55%), 3030 were women (45%), 435 were Asian (6.5%), 205 were Black (3.1%), 93 were Indigenous (1.4%; includes Native American, American Indian, Alaskan First Nations, and other Indigenous persons in the United States, and First Nations, Métis, Inuit, and other Indigenous persons in Canada), 29 were Pacific Islander (0.4%), 5609 were White (84.0%), and 312 were of unknown race (4.7%). Median (IQR) age at enrollment across all cohorts was 66 (58-73) years. The most common diagnosis in the Simmons and PFF cohorts was IPF vs connective tissue disease with ILD in the CARE-PF cohort. Patient characteristics by low vs high PM_2.5 exposures (eTable 2 in the Supplement) demonstrated that higher proportions of non-White patients lived in high-exposure areas compared with low-exposure ones. In the Simmons cohort, 133 of 1030 patients (13%) in the group with high PM_2.5 exposure were non-White compared with 33 of 394 patients (8%) in the group with low exposure. In the PFF cohort, 122 of 900 patients (14%) in the high-exposure group were non-White compared with 79 of 958 (8%) in the low-exposure group. In the CARE-PF cohort, 48 of 406 patients (12%) in the high-exposure group were non-White compared with 653 of 2960 (22%) in the low-exposure group.

Table 1. Demographic Characteristics of Patients by Cohort.

Characteristic	Patients, No. (%)
Characteristic	Simmons cohort (n = 1424)	PFF cohort (n = 1870)	CARE-PF cohort (n = 3389)
5-y Preenrollment PM_2.5, median (IQR), μg/m³	11.4 (9.8 to 13.6)	9.1 (7.9 to 10.2)	6.2 (5.2 to 8.1)
5-y Precensoring PM_2.5, median (IQR), μg/m³	9.4 (7.8 to 11.4)	7.9 (7.0 to 8.8)	6.2 (5.3 to 7.3)
Age at enrollment, median (IQR), y	66 (58 to 73)	68 (61 to 73)	66 (57 to 73)
Sex
Male	795 (56)	1186 (63)	1672 (49)
Female	629 (44)	684 (37)	1717 (51)
Self-reported race
Asian	5 (0.4)	48 (2.6)	382 (11.3)
Black	56 (3.9)	96 (5.1)	53 (1.6)
Indigenous^a	2 (0.1)	3 (0.2)	88 (2.6)
Pacific Islander	0	3 (0.2)	26 (0.8)
White	1258 (88.3)	1669 (89.3)	2682 (79.1)
Unknown	103 (7.2)	51 (2.7)	158 (4.7)
Smoking history
Never	413 (29.0)	779 (41.7)	1279 (37.7)
Former	664 (46.6)	Ever 1091 (58.3)	1922 (56.7)
Current	38 (2.7)	Ever 1091 (58.3)	176 (5.2)
Unknown	309 (21.7)	0	0
fILD diagnostic group
Idiopathic pulmonary fibrosis	716 (50.3)	1202 (64.3)	924 (27.3)
Connective tissue disease with ILD	300 (21.1)	310 (16.6)	1298 (38.3)
Fibrotic hypersensitivity pneumonitis	55 (3.9)	152 (8.1)	259 (7.6)
Pneumoconiosis	26 (1.8)	0	28 (0.8)
Non-IPF idiopathic interstitial pneumonia	68 (4.8)	144 (7.7)	109 (3.2)
Other fILD^b	50 (3.5)	0	121 (3.6)
Unclassifiable or not yet diagnosed	209 (14.7)	62 (3.3)	650 (19.2)
Urbanicity
Metropolitan (>50 000 people)	1078 (75.7)	1611 (86.1)	2333 (68.8)
Micropolitan (10 000-50 000 people)	223 (15.7)	138 (7.4)	597 (17.6)
Rural (<10 000 people)	122 (8.6)	120 (6.4)	459 (13.5)
Neighborhood disadvantage (ADI for Simmons or CIMD for CARE-PF), median score (IQR)	62 (44 to 78)	NA	−0.02 (−0.35 to 0.38)
Percentage of individuals in 5-digit zip code below poverty level, median (IQR)	NA	9 (6 to 14)	NA
Baseline FVC % predicted, median (IQR)^c	66 (53 to 81)	67 (55 to 80)	75 (62 to 89)
Baseline D_LCO % predicted, median (IQR)^d	49 (37 to 63)	40 (31 to 51)	57 (44 to 71)
Follow-up duration, median (IQR), y	3.1 (1.2 to 6.3)	2.5 (1.4 to 3.5)	3.2 (1.8 to 5.2)
Cause of censoring
Death	707 (49.6)	429 (22.9)	765 (22.6)
Lung transplant	201 (14.1)	258 (14.0)	176 (5.2)
Lost to follow-up (no registry update for >1 y)	181 (12.7)	NA	23 (0.7)
Censored	335 (23.5)	1183 (63.3)	2425 (71.6)

Open in a new tab

Abbreviations: ADI, area deprivation index; CARE-PF, Canadian Registry for Pulmonary Fibrosis; CIMD, Canadian Index of Multiple Deprivation; D_LCO, diffusion capacity of the lung for carbon monoxide; fILD, fibrotic interstitial lung disease; FVC, forced vital capacity; IPF, idiopathic pulmonary fibrosis; NA, not applicable; PFF, Pulmonary Fibrosis Foundation; PM_2.5, particulate matter 2.5 μm or less in diameter; Simmons, Simmons Center for Interstitial Lung Disease Registry.

^{^a}

Includes Native American, American Indian, Alaskan First Nations, and other Indigenous persons in the US, as well as First Nations, Métis, Inuit, and other Indigenous persons in Canada.

^{^b}

Includes drug-, radiation-, aspiration-, or acute lung injury–induced fILD.

^{^c}

We evaluated the rate of decrease in FVC for 1054 of 1424 patients (74%) in the Simmons cohort, 1159 of 1870 patients (62%) in the PFF cohort, and 2948 of 3389 patients (87%) in the CARE-PF cohort.

^{^d}

We evaluated the rate of decrease in D_LCO for 1011 of 1424 patients (71%) in the Simmons cohort, 1085 of 1870 patients (58%) in the PFF cohort, and 2779 of 3389 patients (82%) in the CARE-PF cohort.

Figure 1 shows the geographic distribution of PM_2.5 across North America, site locations, and the median breakdown of constituents across the 3 cohorts. eFigure 1 in the Supplement shows cohort-specific correlations of PM_2.5 constituents in the 5 years before censoring. Exposures to total PM_2.5 mass, sulfate, nitrate, ammonium, and BC were highest in patients from the Simmons cohort, followed by the PFF cohort and then the CARE-PF cohort. For the Simmons, PFF, and CARE-PF cohorts, respectively, median (IQR) PM_2.5 level was 9.4 (7.8-11.4) μg/m³, 7.9 (7.0-8.8) μg/m³, and 6.3 (5.3-7.3) μg/m³; median (IQR) sulfate level was 2.1 (1.6-3.7) μg/m³, 1.2 (0.9-1.5) μg/m³, and 0.5 (0.4-1.0) μg/m³; median (IQR) nitrate level was 0.9 (0.8-1.1) μg/m³, 0.8 (0.5-1.1) μg/m³, and 0.5 (0.4-0.8) μg/m³; median (IQR) ammonium level was 0.8 (0.5-1.5) μg/m³, 0.3 (0.2-0.4) μg/m³, and 0.2 (0.1-0.3) μg/m³; and median (IQR) BC level was 0.8 (0.7-1.0) μg/m³, 0.6 (0.5-0.7) μg/m³, and 0.5 (0.3-0.6) μg/m³. Exposures to PM_2.5 in the 5 years before enrollment and the 5 years before censoring were highly correlated (combined cohorts, r = 0.84).

Association of PM_2.5 and Constituent Components With Survival

In all cohorts, 5-year precensoring PM_2.5 exposures of 8 μg/m³ or higher were associated with increased mortality, with the highest effect size in the Simmons cohort (HR, 4.40 [95% CI, 3.51-5.51]; P < .001; PFF cohort HR, 1.71 [95% CI, 1.32-2.21] P < .001; CARE-PF cohort HR, 1.45 [95% CI, 1.18-1.79]; P < .001) (Figure 2; eTable 3 in the Supplement). Continuous models demonstrated similar findings, although the association was not significant for the CARE-PF cohort, with spline models indicating a potentially nonlinear association between total PM_2.5 mass and mortality in that cohort (eFigure 2 in the Supplement). Meta-analysis indicated high heterogeneity between the cohorts (I² = 98%) but noted that an increasing PM_2.5 level was associated with mortality (HR, 1.18 [95% CI, 1.02-1.37]; P = .03) (eTable 4 in the Supplement).

Figure 2. — Kaplan-Meier survival curves for associations of exposures to PM_2.5 total mass in the 5 years before censoring, in which death and transplant are considered composite outcomes. Hazard ratios (HRs) reported for dichotomized and continuous models are adjusted for age at enrollment, sex, race, smoking history, a socioeconomic variable, and site (Pulmonary Fibrosis Foundation Registry and Canadian Registry for Pulmonary Fibrosis only). A, Simmons Center for Interstitial Lung Disease Registry: continuous PM_2.5 exposure (HR, 1.33 [95% CI, 1.29-1.36]; P < .001) and PM_2.5 level of 8 μg/m³ or higher (HR, 4.40 [95% CI, 3.51-5.51]; P < .001). B, Pulmonary Fibrosis Foundation Registry: continuous PM_2.5 exposure (HR, 1.20 [95% CI, 1.10-1.31]; P < .001) and PM_2.5 level of 8 μg/m³ or higher (HR, 1.71 [95% CI, 1.32-2.21]; P < .001). C, Canadian Registry for Pulmonary Fibrosis: continuous PM_2.5 exposure (HR, 1.00 [95% CI, 0.96-1.05]; P = .89) and PM_2.5 level of 8 μg/m³ or higher (HR, 1.45 [95% CI, 1.18-1.79]; P < .001).

Analyses broken down by PM_2.5 constituents (Table 2; eTable 3 in the Supplement) demonstrated strong associations between higher sulfate, nitrate, and ammonium levels in the 5 years before censoring and mortality across all cohorts (eFigure 3 in the Supplement). Associations between other constituents (BC, OM, sea salt, and soil) and mortality were less consistent (eTable 3 in the Supplement). Spline models of continuous HRs for total PM_2.5 mass and each constituent are shown in eFigure 2 in the Supplement. Multiconstituent models demonstrated consistent associations between increasing PM_2.5 constituent mixture and mortality (meta-analysis HR, 2.30 per 1-quantile increase in mixture [95% CI, 2.11-2.50]; P < .001; I² = 25%), with sulfate and ammonium being the most harmful in all cohorts (Figure 3; eTables 4 and 5 in the Supplement).

Table 2. Results From Adjusted Models for Primary Outcomes of Mortality, Baseline FVC and D_LCO, and Rate of Decrease in FVC and D_LCO^a.

Outcome	Simmons			PFF			CARE-PF			Meta-analysis
Outcome	HR or β (95% CI)	P value	No.	HR or β (95% CI)	P value	No.	HR or β (95% CI)	P value	No.	HR or β (95% CI)	P value	I², %	No.
HR for mortality
Total PM_2.5	1.33 (1.29 to 1.36)	<.001^b	1372	1.20 (1.10 to 1.31)	<.001^b	1832	1.00 (0.96 to 1.05)	.89	3353	1.18 (1.02 to 1.37)	.03^b	98	6557
SO₄²⁻	1.79 (1.70 to 1.89)	<.001^b		132.19 (78.12 to 223.70)	<.001^b		2.26 (2.05 to 2.48)	<.001^b		8.02 (0.52 to 122.63)	.13	99
NO₃⁻	3.59 (3.10 to 4.16)	<.001^b		2.48 (1.74 to 3.53)	<.001^b		6.26 (4.16 to 9.42)	<.001^b		3.78 (2.30 to 6.20)	<.001^b	82
NH₄⁺	4.31 (3.83 to 4.86)	<.001^b		903.17 (408.40 to 1998.00)	<.001^b		36.22 (27.32 to 48.03)	<.001^b		50.99 (2.46 to 1056.64)	.01^b	99
MC	2.19 (1.93 to 2.48)	<.001^b		2.76 (2.15 to 3.54)	<.001^b		2.30 (2.02 to 2.61)	<.001^b		2.30 (2.11 to 2.50)	<.001^b	25
β for Baseline FVC
Total PM_2.5	−0.98 (−1.45 to −0.50)	<.001^b	1048	0.20 (−0.40 to 0.79)	.52	1672	−0.07 (−0.59 to 0.46)	.80	2958	−0.30 (−1.00 to 0.41)	.41	82	5678
SO₄^2-	−1.85 (−2.73 to −0.98)	<.001^b		−1.17 (−2.72 to 0.38)	.14		1.38 (−1.17 to 3.93)	.29		−0.90 (−2.52 to 0.73)	.28	65
NO₃⁻	−4.13 (−7.61 to −0.65)	.02^b		1.11 (−1.21 to 3.42)	.35		1.77 (−1.47 to 5.01)	.28		−0.29 (−3.78 to 3.20)	.87	73
NH₄⁺	−4.80 (−7.09 to −2.52)	<.001^b		−1.83 (−4.74 to 1.08)	.22		4.87 (0.49 to 9.25)	.03^b		−0.85 (−6.28 to 4.58)	.76	87
MC	−4.44 (−6.20 to −2.69)	<.001^b		−1.61 (−3.66 to 0.44)	.12		−3.75 (−5.13 to −2.37)	<.001^b		−3.38 (−4.88 to −1.87)	<.001^b	56
β for Baseline D_LCO
Total PM_2.5	−0.13 (−0.63 to 0.36)	.60	978	−0.86 (−1.42 to −0.31)	.002^b	1547	0.006 (−0.54 to 0.56)	.98	2383	−0.32 (−0.84 to 0.19)	.22	64	4908
SO₄²⁻	−0.23 (−1.14 to 0.69)	.62		−3.76 (−5.19 to −2.32)	<.001^b		−0.03 (−2.83 to 2.78)	.98		−1.43 (−3.87 to 1.02)	.25	88
NO₃⁻	0.49 (−3.12 to 4.08)	.79		−0.51 (−2.66 to 1.63)	.64		0.23 (−3.16 to 3.62)	.89		−0.14 (−1.76 to 1.47)	.86	0
NH₄⁺	−0.25 (−2.63 to 2.13)	.83		−6.54 (−9.23 to −3.85)	<.001^b		2.84 (−1.89 to 7.57)	.24		−1.54 (−6.88 to 3.80)	.57	88
MC	−4.14 (−5.96 to −2.33)	<.001^b		−2.40 (−4.31 to −0.48)	.01^b		−4.02 (−5.47 to −2.57)	<.001^b		−3.64 (−4.61 to −2.66)	<.001^b	9
β for FVC decrease
Total PM_2.5	−0.40 (−0.53 to −0.27)	<.001^b	1055	0.007 (−0.28 to 0.30)	.96	1153	−0.01 (−0.13 to 0.11)	.86	2959	−0.15 (−0.42 to 0.12)	.29	90	5167
SO₄²⁻	−0.88 (−1.13 to −0.64)	<.001^b		−3.39 (−5.37 to −1.40)	<.001^b		−3.73 (−4.95 to −2.52)	<.001^b		−2.53 (−4.45 to −0.62)	.01^b	92
NO₃⁻	−2.82 (−3.90 to −1.74)	<.001^b		−0.89 (−2.21 to 0.42)	.18		−1.35 (−2.40 to −0.29)	.01^b		−1.72 (−2.86 to −0.58)	.003^b	67
NH₄⁺	−2.16 (−2.73 to −1.58)	<.001^b		−7.04 (−10.41 to −3.68)	<.001^b		−9.05 (−11.19 to −6.91)	<.001^b		−5.93 (−10.18 to −1.69)	.006^b	95
β for D_LCO decrease
Total PM_2.5	−0.28 (−0.42 to −0.15)	<.001^b	1013	0.09 (−0.24 to 0.43)	.58	1090	0.08 (−0.04 to 0.21)	.19	2775	−0.05 (−0.31 to 0.21)	.70	88	4878
SO₄²⁻	−0.67 (−0.93 to −0.41)	<.001^b		−2.93 (−5.28 to −0.58)	.02^b		−3.29 (−4.64 to −1.95)	<.001^b		−2.12 (−3.93 to −0.30)	.02^b	88
NO₃⁻	−2.61 (−3.80 to −1.43)	<.001^b		−0.23 (−1.77 to 1.34)	.79		−0.66 (−1.79 to 0.46)	.25		−1.21 (−2.66 to 0.24)	.10	75
NH₄⁺	−1.74 (−2.35 to −1.12)	<.001^b		−4.04 (−8.15 to 0.07)	.05		−8.42 (−10.78 to −6.06)	<.001^b		−4.66 (−8.77 to −0.54)	.03^b	93

Open in a new tab

Abbreviations: CARE-PF, Canadian Registry for Pulmonary Fibrosis; D_LCO, diffusion capacity of the lung for carbon monoxide; FVC, forced vital capacity; HR, hazard ratio; MC, multiconstituent; NH₄⁺, ammonium; NO₃⁻, nitrate; PFF, Pulmonary Fibrosis Foundation; PM_2.5, particulate matter 2.5 μm or less in diameter; Simmons, Simmons Center for Interstitial Lung Disease Registry; SO₄²⁻, sulfate.

^{^a}

Death and transplant were considered composite outcomes. Single pollutant effect estimates are per each increase in PM_2.5 or constituent of 1 μg/m³. Multiconstituent effect estimates are per 1-quantile increase in the PM_2.5 mixture of all constituents (SO₄²⁻, NO₃⁻, NH₄⁺, black carbon, organic matter, sea salt, and soil). Analyses are adjusted for age at enrollment, sex, smoking history, race, a socioeconomic variable (area deprivation index for the Simmons cohort, percentage of individuals in the 5-digit zip code below the poverty level in the PFF cohort, and Canadian Index of Multiple Deprivation for the CARE-PF cohort), and site (in the PFF and CARE-PF cohorts). Exposure period for models of mortality and decrease in lung function is the mean of monthly exposures in the 5 years before censoring, whereas baseline lung function exposure periods are for the mean of monthly exposures in the 5 years before enrollment. Unadjusted models and models for other PM_2.5 constituents are reported in eTables 3 through 5 and 10 through 13 in the Supplement.

^{^b}

Significant associations.

Figure 3. — Results are reported from adjusted quantile-based g-computation Cox proportional hazards survival models in which 5-year precensoring estimates for SO₄²⁻, NO₃⁻, NH₄⁺, BC, OM, SS, and soil were included. All models were adjusted for age at enrollment, sex, race, smoking history, a socioeconomic variable, and site (for the Pulmonary Fibrosis Foundation Registry and Canadian Registry for Pulmonary Fibrosis). Bars going in the harmful direction are displayed in light brown, and bars going in the beneficial direction are displayed in light blue. The sum of all positive weights equals 1, and that of all negative weights equals −1 (ie, effect size cannot be directly compared between positive and negative weights). Hazard ratios for 1-quantile increase in overall mixture are 2.19 (95% CI, 1.93-2.48) for the Simmons Center for Interstitial Lung Disease Registry (A), 2.76 (95% CI, 2.15-3.54) for the Pulmonary Fibrosis Foundation Registry (B), and 2.29 (95% CI, 2.02-2.61) for the Canadian Registry for Pulmonary Fibrosis (C) (all P < .001). BC indicates black carbon; NH₄⁺, ammonium; NO₃⁻, nitrate; OM, organic matter; SO₄²⁻, sulfate; and SS, sea salt.

Attributable risk fractions were calculated for dichotomized total PM_2.5 mass and sulfate, nitrate, and ammonium. The attributable risk fraction for PM_2.5 of 8 μg/m³ or higher was 0.71 in the Simmons cohort, 0.26 in the PFF cohort, and 0.05 in the CARE-PF cohort, indicating that if high exposures to PM_2.5 were removed, 71% of the premature mortality in the Simmons cohort could be avoided compared with only 5% in the CARE-PF cohort. The attributable fractions for anthropogenic constituents were greatest in the Simmons cohort (sulfate, 0.93; nitrate, 0.57; and ammonium, 0.84), followed by the PFF cohort (sulfate, 0.75; nitrate, 0.25; and ammonium, 0.70) and the CARE-PF cohort (sulfate, 0.44; nitrate, 0.17; and ammonium, 0.52) (eTable 6 in the Supplement). Sulfate and ammonium carried the highest risk burdens.

Subgroup analyses of patients with IPF showed generally consistent associations (eTable 7 in the Supplement). Effect sizes varied before and after enrollment in 2015 for subgroups, indicating some time variability, but directionality remained consistent (eTable 8 in the Supplement). In the Simmons and CARE-PF cohorts, the HR associated with increasing PM_2.5 was higher for cold months compared with warm ones (Simmons cohort HR, 1.41 vs 1.25; CARE-PF cohort HR, 1.13 vs 0.95), whereas the opposite was observed for the PFF cohort (HR, 1.09 vs 1.35), indicating regional variability in seasonal associations (eTable 9 in the Supplement).

Association of PM_2.5 and Constituent Components With Baseline Lung Function

In adjusted Simmons cohort models, an increase of 1 μg/m³ in the 5-year preenrollment PM_2.5 level was associated with a 0.98% lower estimated percentage baseline FVC (95% CI, −1.45 to −0.50; P < .001) but was not significant in the PFF or CARE-PF cohort (Table 2; eTable 10 in the Supplement). Multiconstituent analyses in the Simmons and CARE-PF cohorts demonstrated that increased exposure to the PM_2.5 constituent mixture was associated with lower baseline FVC (eFigure 4 in the Supplement), with sulfate and ammonium consistently demonstrating an association with adverse outcomes. The β value for a 1-quantile increase in PM_2.5 mixture in the Simmons cohort was −4.44 (95% CI, −6.20 to −2.69; P <. 001); for the PFF cohort, −1.61 (95% CI, −3.66 to 0.44; P = .12); and for the CARE-PF cohort, −3.75 (95% CI, −5.13 to −2.37; P < .001), with sulfate and ammonium having consistent negative associations in all 3 cohorts. Meta-analyses indicated that across the 3 cohorts, a 1-quantile increase in the constituent mixture was associated with a 3.38% lower estimated percentage baseline FVC (95% CI, −4.88 to −1.87; P < .001) (Table 2; eTables 4 and 5 in the Supplement).

Total PM_2.5 level in the 5 years before enrollment was associated with lower baseline D_LCO only in the PFF cohort (β = −0.86 [95% CI, −1.42 to −0.31]; P = .002), as were sulfate (β = −3.76 [95% CI, −5.19 to −2.32]) and ammonium (β = −6.54 [95% CI, −9.23 to −3.85]) (Table 2; eTable 11 in the Supplement). Multiconstituent models indicated consistent negative associations between increasing PM_2.5 mixture and baseline D_LCO (Table 2; eFigure 5 and eTable 11 in the Supplement), again with sulfate and ammonium consistently demonstrating an association with adverse outcomes. Meta-analysis indicated that each 1-quantile increase in constituent mixture was associated with a 3.64% lower estimated baseline percentage D_LCO (95% CI, −4.61 to −2.66; P < .001; I² = 9%).

Association of PM_2.5 and Its Constituents With Decrease in Lung Function

We evaluated the rate of decrease in FVC for 1054 of 1424 patients (74%) in the Simmons cohort, 1159 of 1870 patients (62%) in the PFF cohort, and 2948 of 3389 patients (87%) in the CARE-PF cohort. Each increase of 1 μg/m³ in 5-year precensoring PM_2.5 exposure in the Simmons cohort was associated with an additional 0.4% decrease in FVC percentage estimated per year (95% CI, −0.53 to −0.27; P < .001), but this association was not observed in the PFF or CARE-PF cohorts (Table 2; eTable 12 in the Supplement). Higher exposures to sulfate, nitrate, and ammonium were associated with a more rapid decrease in FVC percentage estimated per year in meta-analysis (Table 2; eTable 4 in the Supplement). The β value for sulfate was −2.53 (95% CI, −4.45 to −0.62; P = .01); for nitrate, −1.72 (95% CI, −2.86 to −0.58; P = .003); and for ammonium, −5.93 (95% CI, −10.18 to −1.69; P = .006).

We evaluated the rate of decrease in D_LCO for 1011 of 1424 patients (71%) in the Simmons cohort, 1085 of 1870 patients (58%) in the PFF cohort, and 2779 of 3389 patients (82%) in the CARE-PF cohort. Each increase of 1 μg/m³ in 5-year precensoring PM_2.5 exposure in the Simmons cohort was associated with an additional 0.28% decrease in D_LCO percentage estimated per year (95% CI, −0.42% to −0.15%; P < .001), but this association was not observed in the PFF or CARE-PF cohort (Table 2; eTable 13 in the Supplement). Higher exposures to sulfate and ammonium were associated with a more rapid decrease in D_LCO percentage estimated per year in meta-analysis (sulfate: β = −2.12 [95% CI, −3.93 to −0.30]; P = .02; ammonium: β = −4.66 [95% CI, −8.77 to −0.54]; P = .03) (Table 2; eTable 4 in the Supplement).

Discussion

This study of 6683 patients with fILD from across North America demonstrates that PM_2.5 and its constituents, primarily sulfate, nitrate, and ammonium, were associated with adverse health outcomes (ie, increased mortality, worse baseline lung function, and more rapid disease progression) among patients with fILD. Differences in mortality and lung function between the 3 cohorts demonstrated how PM_2.5 constituents from industry and human activities are significantly associated with the adverse outcomes of patients with fILD.

In this geographically and diagnostically diverse cohort, we demonstrate an association between high PM_2.5 exposure and mortality. This association was most pronounced in the Simmons cohort, which had the highest burden of heavy industry–associated PM_2.5 constituents: sulfate, nitrate, and ammonium. We demonstrated consistency in the association of mortality with exposure to sulfate, nitrate, and ammonium across all 3 cohorts, highlighting how these constituents of PM_2.5 may be the primary constituents associated with mortality among patients with fILDs. Our findings are consistent with recent literature that demonstrates increased all-cause mortality associated with the PM_2.5 constituents sulfate and nitrate compared with constituents such as soil or OM.^33,34 Recent work also indicates that a higher ammonium level may be associated with increased incidence of ILD among patients with rheumatoid arthritis, also indicating that this constituent may have pathophysiologic relevance to both the development and progression of fILDs.³¹

Attributable risk fractions illustrate how the association between PM_2.5 and fILD mortality varies substantially, depending on the mass and constituent makeup of PM_2.5 in a region. An attributable risk fraction can exceed 100% because of complex interactions between social, environmental, and biologic risk factors, indicating the need to interpret these findings with caution,³⁵ but the Simmons cohort’s attributable risk fraction of 0.71 for patients exposed to high PM_2.5 levels (≥8 μg/m³) implies that 71% of this group’s premature mortality could have been averted if the exposure had not occurred. This metric is most useful for weighing relative burdens across cohorts, implying that the mortality burden attributable to high PM_2.5 levels in the Simmons cohort was approximately 2.7 times greater than that of the PFF cohort and approximately 14 times greater than that of the CARE-PF cohort. Policy interventions that reduce PM_2.5 total mass to below American Thoracic Society standards, with specific targeting of anthropogenic sources of emissions, may result in the greatest reduction of mortality in this vulnerable group.

Although there was consistency in the association of exposure to high PM_2.5 levels (≥8 μg/m³) with mortality and baseline FVC across all 3 cohorts, there were inconsistencies in some constituent analyses, baseline D_LCO models, and analyses of the decrease in lung function. Inconsistencies in the analyses of the decrease in lung function may be associated with the higher proportions of patients with IPF in the Simmons and PFF cohorts, whereas the CARE-PF cohort had a larger proportion of patients with connective tissue disease with ILD, wherein patients may experience a more indolent disease course.³⁶ The distribution of PM_2.5 constituents also varied greatly, with the Simmons cohort demonstrating the highest proportion of deleterious constituents from industrial activities, including sulfate and ammonium, whereas the CARE-PF cohort had higher proportions of OM and other unmeasured constituents, which are frequently derived from biomass burning and nonanthropogenic sources.¹⁸ Similarly, patients in the CARE-PF cohort experienced lower total exposures to PM_2.5 and constituent components compared with those in the Simmons or PFF cohort. In the Simmons cohort, this was likely associated with the earlier establishment of the ILD registry alongside higher historical pollution exposures in western Pennsylvania, where steel, coal, and other metal industries predominate.³⁷ The relatively minimal exposure of patients in the CARE-PF cohort to sulfate and ammonium may have blunted the significance of the associations between total PM_2.5 mass and mortality.

Strengths and Limitations

Although this study was strengthened by its large geographically and diagnostically diverse multinational cohort of patients with fILD, it is not without limitations. Pollution exposures were estimated at a patient’s most recent residential address, but this approach does not account for mobility or changes in address (for which data were unavailable), leading to some risk of exposure misclassification. This study also evaluated pollution exposures during a prespecified period of 5 years before censoring (or 5 years before enrollment for baseline lung function analyses), and future work is needed to determine the best at-risk period for these patients. Given that this was not a population-based study, analyses of associations between exposures and clinical outcomes may have been affected by selection bias because not all patients may have access to subspecialist ILD care. In addition, patients lost to follow-up represented more than 10% of the Simmons cohort and were not available for the PFF cohort, which may have affected mortality estimates if these losses were not truly random as was assumed. Further work is also needed to understand the biologic mechanisms underpinning the association of PM_2.5 and its constituent components with adverse clinical outcomes in this population.⁷ Last, extensions of this work should use strategies to source-apportion the PM_2.5 constituents sulfate, nitrate, and ammonium that patients are exposed to, thereby enabling more stringent regulatory oversight of the human sources of emissions from which these constituents are derived.

Conclusions

To date, this study represents the largest and most geographically diverse evaluation of the associations between PM_2.5 pollution and disease in patients with diverse forms of fILD, and to our knowledge, it is the first to evaluate the association between specific PM_2.5 constituents and health outcomes in this population. We found that exposures to total PM_2.5 mass of 8 μg/m³ or greater were consistently associated with increased mortality and worse baseline lung function. Multiconstituent models demonstrated that sulfate, nitrate, and ammonium, which are primary by-products of industrial and transportation activities, are the main PM_2.5 constituents associated with adverse outcomes among patients with fILD. This work unveils new paradigms in our understanding of the importance of PM_2.5 composition in health outcomes, indicating a strong need to evaluate constituent-specific associations with other diseases. These findings are of critical importance to the development of policies aimed at protecting vulnerable populations, such as patients with fILD, because they demonstrate how anthropogenic sources of pollution may be associated more significantly with disease morbidity and mortality.

Supplement.

eTable 1. PM_2.5 and Constituent Component Breakdown by Study Site

eTable 2. Demographic Characteristics for Cohorts Split by Low vs High PM_2.5 Exposures (Less Than 8 or at Least 8 μg/m³) in the 5 Years Precensoring

eTable 3. Full Cohort Mortality Models

eTable 4. Meta-analysis Results

eTable 5. Multiconstituent Model Results

eTable 6. Cohort Attributable Risk Fractions for PM_2.5, Sulfate, Nitrate, and Ammonium

eTable 7. IPF Subgroup Analyses for Primary Outcomes

eTable 8. Pre-2015 and Post-2015 Year of Enrollment Subgroup Analyses for Mortality Outcome

eTable 9. Sensitivity Analysis for Mortality in Warm vs Cold Months

eTable 10. Full Cohort Baseline FVC Models

eTable 11. Full Cohort Baseline DLCO Models

eTable 12. Full Cohort FVC Decline Models

eTable 13. Full Cohort DLCO Decline Models

eFigure 1. Scatterplot Matrix of 5-Year Precensoring Exposures to PM_2.5 and Constituent Components

eFigure 2. PM_2.5 and Constituent Spline Models

eFigure 3. Survival by Low vs High Anthropogenic Constituent Exposures in the 5 Years Precensoring

eFigure 4. Tornado Plots of PM_2.5 Constituent Associations With Baseline Forced Vital Capacity (FVC) in Multipollutant Models

eFigure 5. Tornado Plots of PM_2.5 Constituent Associations With Baseline Diffusion Capacity of the Lung for Carbon Monoxide (DLCO) in Multipollutant Models

Click here for additional data file.^{(3.2MB, pdf)}

References

1.Raghu G, Remy-Jardin M, Myers JL, et al. ; American Thoracic Society, European Respiratory Society, Japanese Respiratory Society, and Latin American Thoracic Society . Diagnosis of idiopathic pulmonary fibrosis: an official ATS/ERS/JRS/ALAT clinical practice guideline. Am J Respir Crit Care Med. 2018;198(5):e44-e68. doi: 10.1164/rccm.201807-1255ST [DOI] [PubMed] [Google Scholar]
2.Cottin V, Hirani NA, Hotchkin DL, et al. Presentation, diagnosis and clinical course of the spectrum of progressive-fibrosing interstitial lung diseases. Eur Respir Rev. 2018;27(150):180076. doi: 10.1183/16000617.0076-2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Sesé L, Nunes H, Cottin V, et al. Role of atmospheric pollution on the natural history of idiopathic pulmonary fibrosis. Thorax. 2018;73(2):145-150. doi: 10.1136/thoraxjnl-2017-209967 [DOI] [PubMed] [Google Scholar]
4.Winterbottom CJ, Shah RJ, Patterson KC, et al. Exposure to ambient particulate matter is associated with accelerated functional decline in idiopathic pulmonary fibrosis. Chest. 2018;153(5):1221-1228. doi: 10.1016/j.chest.2017.07.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Conti S, Harari S, Caminati A, et al. The association between air pollution and the incidence of idiopathic pulmonary fibrosis in Northern Italy. Eur Respir J. 2018;51(1):1700397. doi: 10.1183/13993003.00397-2017 [DOI] [PubMed] [Google Scholar]
6.Johannson KA, Vittinghoff E, Morisset J, et al. Air pollution exposure is associated with lower lung function, but not changes in lung function, in patients with idiopathic pulmonary fibrosis. Chest. 2018;154(1):119-125. doi: 10.1016/j.chest.2018.01.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Goobie GC, Nouraie M, Zhang Y, et al. Air pollution and interstitial lung diseases: defining epigenomic effects. Am J Respir Crit Care Med. 2020;202(9):1217-1224. doi: 10.1164/rccm.202003-0836PP [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Cohen AJ, Brauer M, Burnett R, et al. Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution: an analysis of data from the Global Burden of Diseases Study 2015. Lancet. 2017;389(10082):1907-1918. doi: 10.1016/S0140-6736(17)30505-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Burnett R, Chen H, Szyszkowicz M, et al. Global estimates of mortality associated with long-term exposure to outdoor fine particulate matter. Proc Natl Acad Sci U S A. 2018;115(38):9592-9597. doi: 10.1073/pnas.1803222115 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.van Donkelaar A, Martin RV, Li C, Burnett RT. Regional estimates of chemical composition of fine particulate matter using a combined geoscience-statistical method with information from satellites, models, and monitors. Environ Sci Technol. 2019;53(5):2595-2611. doi: 10.1021/acs.est.8b06392 [DOI] [PubMed] [Google Scholar]
11.Hammer MS, van Donkelaar A, Li C, et al. Global estimates and long-term trends of fine particulate matter concentrations (1998-2018). Environ Sci Technol. 2020;54(13):7879-7890. doi: 10.1021/acs.est.0c01764 [DOI] [PubMed] [Google Scholar]
12.Zhang H, Hu J, Kleeman M, Ying Q. Source apportionment of sulfate and nitrate particulate matter in the eastern United States and effectiveness of emission control programs. Sci Total Environ. 2014;490:171-181. doi: 10.1016/j.scitotenv.2014.04.064 [DOI] [PubMed] [Google Scholar]
13.Plautz J. Ammonia, a poorly understood smog ingredient, could be key to limiting deadly pollution. Science. Published September 13, 2018. Accessed March 2, 2022. https://www.science.org/content/article/ammonia-poorly-understood-smog-ingredient-could-be-key-limiting-deadly-pollution
14.Hewitt CN. The atmospheric chemistry of sulphur and nitrogen in power station plumes. Atmos Environ. 2002;35:1155-1170. doi: 10.1016/S1352-2310(00)00463-5 [DOI] [Google Scholar]
15.Spengler JD, Koutrakis P, Dockery DW, Raizenne M, Speizer FE. Health effects of acid aerosols on North American children: air pollution exposures. Environ Health Perspect. 1996;104(5):492-499. doi: 10.1289/ehp.96104492 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Gwynn RC, Burnett RT, Thurston GD. A time-series analysis of acidic particulate matter and daily mortality and morbidity in the Buffalo, New York, region. Environ Health Perspect. 2000;108(2):125-133. doi: 10.1289/ehp.00108125 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Forbes MS, Raison RJ, Skjemstad JO. Formation, transformation and transport of black carbon (charcoal) in terrestrial and aquatic ecosystems. Sci Total Environ. 2006;370(1):190-206. doi: 10.1016/j.scitotenv.2006.06.007 [DOI] [PubMed] [Google Scholar]
18.Philip S, Martin RV, van Donkelaar A, et al. Global chemical composition of ambient fine particulate matter for exposure assessment. Environ Sci Technol. 2014;48(22):13060-13068. doi: 10.1021/es502965b [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Raghu G, Remy-Jardin M, Ryerson CJ, et al. Diagnosis of hypersensitivity pneumonitis in adults: an official ATS/JRS/ALAT clinical practice guideline. Am J Respir Crit Care Med. 2020;202(3):e36-e69. doi: 10.1164/rccm.202005-2032ST [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Wang BR, Edwards R, Freiheit EA, et al. The Pulmonary Fibrosis Foundation Patient Registry: rationale, design, and methods. Ann Am Thorac Soc. 2020;17(12):1620-1628. doi: 10.1513/AnnalsATS.202001-035SD [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Ryerson CJ, Tan B, Fell CD, et al. The Canadian Registry for Pulmonary Fibrosis: design and rationale of a national pulmonary fibrosis registry. Can Respir J. 2016;2016:3562923. doi: 10.1155/2016/3562923 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Visweswaran S, McLay B, Cappella N, et al. An atomic approach to the design and implementation of a research data warehouse. J Am Med Inform Assoc. 2022;29(4):601-608. doi: 10.1093/jamia/ocab204 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Goobie GC, Ryerson CJ, Johannson KA, et al. Neighborhood-level disadvantage impacts on patients with fibrotic interstitial lung disease. Am J Respir Crit Care Med. 2022;205(4):459-467. doi: 10.1164/rccm.202109-2065OC [DOI] [PubMed] [Google Scholar]
24.The Canadian Index of Multiple Deprivation : user guide. Published June 12, 2019. Accessed September 28, 2019. https://www150.statcan.gc.ca/n1/en/pub/45-20-0001/452000012019002-eng.pdf?st=XEILlJq4
25.Kind AJH, Buckingham WR. Making neighborhood-disadvantage metrics accessible—the Neighborhood Atlas. N Engl J Med. 2018;378(26):2456-2458. doi: 10.1056/NEJMp1802313 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.US Census Bureau . S1701: poverty status in the past 12 months: 2020: ACS 5-year estimates subject tables. Accessed August 8, 2022. https://data.census.gov/cedsci/table?q=poverty&g=0100000US%248600000&tid=ACSST5Y2020.S1701
27.Cromar KR, Gladson LA, Hicks EA, Marsh B, Ewart G. Excess morbidity and mortality associated with air pollution above American Thoracic Society recommended standards, 2017-2019. Ann Am Thorac Soc. 2022;19(4):603-613. doi: 10.1513/AnnalsATS.202107-860OC [DOI] [PubMed] [Google Scholar]
28.Ley B, Ryerson CJ, Vittinghoff E, et al. A multidimensional index and staging system for idiopathic pulmonary fibrosis. Ann Intern Med. 2012;156(10):684-691. doi: 10.7326/0003-4819-156-10-201205150-00004 [DOI] [PubMed] [Google Scholar]
29.Adegunsoye A, Oldham JM, Bellam SK, et al. African-American race and mortality in interstitial lung disease: a multicentre propensity-matched analysis. Eur Respir J. 2018;51(6):1800255. doi: 10.1183/13993003.00255-2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Margaritopoulos GA, Vasarmidi E, Jacob J, Wells AU, Antoniou KM. Smoking and interstitial lung diseases. Eur Respir Rev. 2015;24(137):428-435. doi: 10.1183/16000617.0050-2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Zhao N, Al-Aly Z, Zheng B, et al. Fine particulate matter components and interstitial lung disease in rheumatoid arthritis. Eur Respir J. 2022;60(1):2102149. doi: 10.1183/13993003.02149-2021 [DOI] [PubMed] [Google Scholar]
32.R Project for Statistical Computing. Accessed August 30, 2022. https://www.r-project.org/
33.Wang C, Cardenas A, Hutchinson JN, et al. Short- and intermediate-term exposure to ambient fine particulate elements and leukocyte epigenome-wide DNA methylation in older men: the Normative Aging Study. Environ Int. 2022;158:106955. doi: 10.1016/j.envint.2021.106955 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Kazemiparkouhi F, Honda T, Eum KD, Wang B, Manjourides J, Suh HH. The impact of long-term PM_2.5 constituents and their sources on specific causes of death in a US Medicare cohort. Environ Int. 2022;159:106988. doi: 10.1016/j.envint.2021.106988 [DOI] [PubMed] [Google Scholar]
35.Levine B. What does the population attributable fraction mean? Prev Chronic Dis. 2007;4(1):A14. [PMC free article] [PubMed] [Google Scholar]
36.Ryerson CJ, Vittinghoff E, Ley B, et al. Predicting survival across chronic interstitial lung disease: the ILD-GAP model. Chest. 2014;145(4):723-728. doi: 10.1378/chest.13-1474 [DOI] [PubMed] [Google Scholar]
37.Dutzik T, Barber Z. Cutting through the smoke: why the Allegheny County Health Department must turn the corner on decades of weak clean air enforcement. PennEnvironment Research & Policy Center. Published August 2019. Accessed August 30, 2022. https://environmentamerica.org/pennsylvania/wp-content/uploads/2019/08/PA_cuttingsmoke_final_scrn-revd-1.pdf

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement.

eTable 1. PM_2.5 and Constituent Component Breakdown by Study Site

eTable 2. Demographic Characteristics for Cohorts Split by Low vs High PM_2.5 Exposures (Less Than 8 or at Least 8 μg/m³) in the 5 Years Precensoring

eTable 3. Full Cohort Mortality Models

eTable 4. Meta-analysis Results

eTable 5. Multiconstituent Model Results

eTable 6. Cohort Attributable Risk Fractions for PM_2.5, Sulfate, Nitrate, and Ammonium

eTable 7. IPF Subgroup Analyses for Primary Outcomes

eTable 8. Pre-2015 and Post-2015 Year of Enrollment Subgroup Analyses for Mortality Outcome

eTable 9. Sensitivity Analysis for Mortality in Warm vs Cold Months

eTable 10. Full Cohort Baseline FVC Models

eTable 11. Full Cohort Baseline DLCO Models

eTable 12. Full Cohort FVC Decline Models

eTable 13. Full Cohort DLCO Decline Models

eFigure 1. Scatterplot Matrix of 5-Year Precensoring Exposures to PM_2.5 and Constituent Components

eFigure 2. PM_2.5 and Constituent Spline Models

eFigure 3. Survival by Low vs High Anthropogenic Constituent Exposures in the 5 Years Precensoring

eFigure 4. Tornado Plots of PM_2.5 Constituent Associations With Baseline Forced Vital Capacity (FVC) in Multipollutant Models

eFigure 5. Tornado Plots of PM_2.5 Constituent Associations With Baseline Diffusion Capacity of the Lung for Carbon Monoxide (DLCO) in Multipollutant Models

Click here for additional data file.^{(3.2MB, pdf)}

[ioi220062r1] 1.Raghu G, Remy-Jardin M, Myers JL, et al. ; American Thoracic Society, European Respiratory Society, Japanese Respiratory Society, and Latin American Thoracic Society . Diagnosis of idiopathic pulmonary fibrosis: an official ATS/ERS/JRS/ALAT clinical practice guideline. Am J Respir Crit Care Med. 2018;198(5):e44-e68. doi: 10.1164/rccm.201807-1255ST [DOI] [PubMed] [Google Scholar]

[ioi220062r2] 2.Cottin V, Hirani NA, Hotchkin DL, et al. Presentation, diagnosis and clinical course of the spectrum of progressive-fibrosing interstitial lung diseases. Eur Respir Rev. 2018;27(150):180076. doi: 10.1183/16000617.0076-2018 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r3] 3.Sesé L, Nunes H, Cottin V, et al. Role of atmospheric pollution on the natural history of idiopathic pulmonary fibrosis. Thorax. 2018;73(2):145-150. doi: 10.1136/thoraxjnl-2017-209967 [DOI] [PubMed] [Google Scholar]

[ioi220062r4] 4.Winterbottom CJ, Shah RJ, Patterson KC, et al. Exposure to ambient particulate matter is associated with accelerated functional decline in idiopathic pulmonary fibrosis. Chest. 2018;153(5):1221-1228. doi: 10.1016/j.chest.2017.07.034 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r5] 5.Conti S, Harari S, Caminati A, et al. The association between air pollution and the incidence of idiopathic pulmonary fibrosis in Northern Italy. Eur Respir J. 2018;51(1):1700397. doi: 10.1183/13993003.00397-2017 [DOI] [PubMed] [Google Scholar]

[ioi220062r6] 6.Johannson KA, Vittinghoff E, Morisset J, et al. Air pollution exposure is associated with lower lung function, but not changes in lung function, in patients with idiopathic pulmonary fibrosis. Chest. 2018;154(1):119-125. doi: 10.1016/j.chest.2018.01.015 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r7] 7.Goobie GC, Nouraie M, Zhang Y, et al. Air pollution and interstitial lung diseases: defining epigenomic effects. Am J Respir Crit Care Med. 2020;202(9):1217-1224. doi: 10.1164/rccm.202003-0836PP [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r8] 8.Cohen AJ, Brauer M, Burnett R, et al. Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution: an analysis of data from the Global Burden of Diseases Study 2015. Lancet. 2017;389(10082):1907-1918. doi: 10.1016/S0140-6736(17)30505-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r9] 9.Burnett R, Chen H, Szyszkowicz M, et al. Global estimates of mortality associated with long-term exposure to outdoor fine particulate matter. Proc Natl Acad Sci U S A. 2018;115(38):9592-9597. doi: 10.1073/pnas.1803222115 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r10] 10.van Donkelaar A, Martin RV, Li C, Burnett RT. Regional estimates of chemical composition of fine particulate matter using a combined geoscience-statistical method with information from satellites, models, and monitors. Environ Sci Technol. 2019;53(5):2595-2611. doi: 10.1021/acs.est.8b06392 [DOI] [PubMed] [Google Scholar]

[ioi220062r11] 11.Hammer MS, van Donkelaar A, Li C, et al. Global estimates and long-term trends of fine particulate matter concentrations (1998-2018). Environ Sci Technol. 2020;54(13):7879-7890. doi: 10.1021/acs.est.0c01764 [DOI] [PubMed] [Google Scholar]

[ioi220062r12] 12.Zhang H, Hu J, Kleeman M, Ying Q. Source apportionment of sulfate and nitrate particulate matter in the eastern United States and effectiveness of emission control programs. Sci Total Environ. 2014;490:171-181. doi: 10.1016/j.scitotenv.2014.04.064 [DOI] [PubMed] [Google Scholar]

[ioi220062r13] 13.Plautz J. Ammonia, a poorly understood smog ingredient, could be key to limiting deadly pollution. Science. Published September 13, 2018. Accessed March 2, 2022. https://www.science.org/content/article/ammonia-poorly-understood-smog-ingredient-could-be-key-limiting-deadly-pollution

[ioi220062r14] 14.Hewitt CN. The atmospheric chemistry of sulphur and nitrogen in power station plumes. Atmos Environ. 2002;35:1155-1170. doi: 10.1016/S1352-2310(00)00463-5 [DOI] [Google Scholar]

[ioi220062r15] 15.Spengler JD, Koutrakis P, Dockery DW, Raizenne M, Speizer FE. Health effects of acid aerosols on North American children: air pollution exposures. Environ Health Perspect. 1996;104(5):492-499. doi: 10.1289/ehp.96104492 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r16] 16.Gwynn RC, Burnett RT, Thurston GD. A time-series analysis of acidic particulate matter and daily mortality and morbidity in the Buffalo, New York, region. Environ Health Perspect. 2000;108(2):125-133. doi: 10.1289/ehp.00108125 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r17] 17.Forbes MS, Raison RJ, Skjemstad JO. Formation, transformation and transport of black carbon (charcoal) in terrestrial and aquatic ecosystems. Sci Total Environ. 2006;370(1):190-206. doi: 10.1016/j.scitotenv.2006.06.007 [DOI] [PubMed] [Google Scholar]

[ioi220062r18] 18.Philip S, Martin RV, van Donkelaar A, et al. Global chemical composition of ambient fine particulate matter for exposure assessment. Environ Sci Technol. 2014;48(22):13060-13068. doi: 10.1021/es502965b [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r19] 19.Raghu G, Remy-Jardin M, Ryerson CJ, et al. Diagnosis of hypersensitivity pneumonitis in adults: an official ATS/JRS/ALAT clinical practice guideline. Am J Respir Crit Care Med. 2020;202(3):e36-e69. doi: 10.1164/rccm.202005-2032ST [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r20] 20.Wang BR, Edwards R, Freiheit EA, et al. The Pulmonary Fibrosis Foundation Patient Registry: rationale, design, and methods. Ann Am Thorac Soc. 2020;17(12):1620-1628. doi: 10.1513/AnnalsATS.202001-035SD [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r21] 21.Ryerson CJ, Tan B, Fell CD, et al. The Canadian Registry for Pulmonary Fibrosis: design and rationale of a national pulmonary fibrosis registry. Can Respir J. 2016;2016:3562923. doi: 10.1155/2016/3562923 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r22] 22.Visweswaran S, McLay B, Cappella N, et al. An atomic approach to the design and implementation of a research data warehouse. J Am Med Inform Assoc. 2022;29(4):601-608. doi: 10.1093/jamia/ocab204 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r23] 23.Goobie GC, Ryerson CJ, Johannson KA, et al. Neighborhood-level disadvantage impacts on patients with fibrotic interstitial lung disease. Am J Respir Crit Care Med. 2022;205(4):459-467. doi: 10.1164/rccm.202109-2065OC [DOI] [PubMed] [Google Scholar]

[ioi220062r24] 24.The Canadian Index of Multiple Deprivation : user guide. Published June 12, 2019. Accessed September 28, 2019. https://www150.statcan.gc.ca/n1/en/pub/45-20-0001/452000012019002-eng.pdf?st=XEILlJq4

[ioi220062r25] 25.Kind AJH, Buckingham WR. Making neighborhood-disadvantage metrics accessible—the Neighborhood Atlas. N Engl J Med. 2018;378(26):2456-2458. doi: 10.1056/NEJMp1802313 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r26] 26.US Census Bureau . S1701: poverty status in the past 12 months: 2020: ACS 5-year estimates subject tables. Accessed August 8, 2022. https://data.census.gov/cedsci/table?q=poverty&g=0100000US%248600000&tid=ACSST5Y2020.S1701

[ioi220062r27] 27.Cromar KR, Gladson LA, Hicks EA, Marsh B, Ewart G. Excess morbidity and mortality associated with air pollution above American Thoracic Society recommended standards, 2017-2019. Ann Am Thorac Soc. 2022;19(4):603-613. doi: 10.1513/AnnalsATS.202107-860OC [DOI] [PubMed] [Google Scholar]

[ioi220062r28] 28.Ley B, Ryerson CJ, Vittinghoff E, et al. A multidimensional index and staging system for idiopathic pulmonary fibrosis. Ann Intern Med. 2012;156(10):684-691. doi: 10.7326/0003-4819-156-10-201205150-00004 [DOI] [PubMed] [Google Scholar]

[ioi220062r29] 29.Adegunsoye A, Oldham JM, Bellam SK, et al. African-American race and mortality in interstitial lung disease: a multicentre propensity-matched analysis. Eur Respir J. 2018;51(6):1800255. doi: 10.1183/13993003.00255-2018 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r30] 30.Margaritopoulos GA, Vasarmidi E, Jacob J, Wells AU, Antoniou KM. Smoking and interstitial lung diseases. Eur Respir Rev. 2015;24(137):428-435. doi: 10.1183/16000617.0050-2015 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r31] 31.Zhao N, Al-Aly Z, Zheng B, et al. Fine particulate matter components and interstitial lung disease in rheumatoid arthritis. Eur Respir J. 2022;60(1):2102149. doi: 10.1183/13993003.02149-2021 [DOI] [PubMed] [Google Scholar]

[ioi220062r32] 32.R Project for Statistical Computing. Accessed August 30, 2022. https://www.r-project.org/

[ioi220062r33] 33.Wang C, Cardenas A, Hutchinson JN, et al. Short- and intermediate-term exposure to ambient fine particulate elements and leukocyte epigenome-wide DNA methylation in older men: the Normative Aging Study. Environ Int. 2022;158:106955. doi: 10.1016/j.envint.2021.106955 [DOI] [PMC free article] [PubMed] [Google Scholar]

[ioi220062r34] 34.Kazemiparkouhi F, Honda T, Eum KD, Wang B, Manjourides J, Suh HH. The impact of long-term PM_2.5 constituents and their sources on specific causes of death in a US Medicare cohort. Environ Int. 2022;159:106988. doi: 10.1016/j.envint.2021.106988 [DOI] [PubMed] [Google Scholar]

[ioi220062r35] 35.Levine B. What does the population attributable fraction mean? Prev Chronic Dis. 2007;4(1):A14. [PMC free article] [PubMed] [Google Scholar]

[ioi220062r36] 36.Ryerson CJ, Vittinghoff E, Ley B, et al. Predicting survival across chronic interstitial lung disease: the ILD-GAP model. Chest. 2014;145(4):723-728. doi: 10.1378/chest.13-1474 [DOI] [PubMed] [Google Scholar]

[ioi220062r37] 37.Dutzik T, Barber Z. Cutting through the smoke: why the Allegheny County Health Department must turn the corner on decades of weak clean air enforcement. PennEnvironment Research & Policy Center. Published August 2019. Accessed August 30, 2022. https://environmentamerica.org/pennsylvania/wp-content/uploads/2019/08/PA_cuttingsmoke_final_scrn-revd-1.pdf

PERMALINK

Association of Particulate Matter Exposure With Lung Function and Mortality Among Patients With Fibrotic Interstitial Lung Disease

Gillian C Goobie, MD

Christopher Carlsten, MD, MPH

Kerri A Johannson, MD, MPH

Nasreen Khalil, MD

Veronica Marcoux, MD

Deborah Assayag, MD

Hélène Manganas, MD

Jolene H Fisher, MD

Martin R J Kolb, MD, PhD

Kathleen O Lindell, RN, PhD

James P Fabisiak, PhD

Xiaoping Chen, MS

Kevin F Gibson, MD

Yingze Zhang, PhD

Daniel J Kass, MD

Christopher J Ryerson, MD, MAS

S Mehdi Nouraie, MD, PhD

Key Points

Question

Findings

Meaning

Abstract

Importance

Objective

Design, Setting, and Participants

Exposures

Main Outcomes and Measures

Results

Conclusions and Relevance

Introduction

Methods

Study Populations

Demographic Characteristics, Residential Data, and Clinical Outcomes

Determination of Exposure to Particulate Matter and Its Constituents

Statistical Analysis

Results

Baseline Patient Characteristics and Pollutant Exposures

Table 1. Demographic Characteristics of Patients by Cohort.

Figure 1. Distribution and Constituent Composition of Particulate Matter 2.5 μm or Less in Diameter (PM2.5) Across 3 Cohorts.

Association of PM2.5 and Constituent Components With Survival

Figure 2. Survival by Exposures to Low (<8 μg/m3) vs High (≥8 μg/m3) Levels of Particulate Matter 2.5 μm or Less in Diameter (PM2.5) 5 Years Before Censoring.

Table 2. Results From Adjusted Models for Primary Outcomes of Mortality, Baseline FVC and DLCO, and Rate of Decrease in FVC and DLCOa.

Figure 3. Tornado Plots of Association of Constituents of Particulate Matter 2.5 μm or Less in Diameter With Mortality in Multipollutant Models.

Association of PM2.5 and Constituent Components With Baseline Lung Function

Association of PM2.5 and Its Constituents With Decrease in Lung Function

Discussion

Strengths and Limitations

Conclusions

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Figure 1. Distribution and Constituent Composition of Particulate Matter 2.5 μm or Less in Diameter (PM_2.5) Across 3 Cohorts.

Association of PM_2.5 and Constituent Components With Survival

Figure 2. Survival by Exposures to Low (<8 μg/m³) vs High (≥8 μg/m³) Levels of Particulate Matter 2.5 μm or Less in Diameter (PM_2.5) 5 Years Before Censoring.

Table 2. Results From Adjusted Models for Primary Outcomes of Mortality, Baseline FVC and D_LCO, and Rate of Decrease in FVC and D_LCO^a.

Association of PM_2.5 and Constituent Components With Baseline Lung Function

Association of PM_2.5 and Its Constituents With Decrease in Lung Function