Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Dec 20.
Published in final edited form as: Sci Total Environ. 2025 Sep 29;1002:180490. doi: 10.1016/j.scitotenv.2025.180490

Ambient air pollution exposure and lung cancer risk in a large prospective U.S. cohort

Jared A Fisher 1, Linda M Liao 2, Barry I Graubard 3, Joel D Kaufman 4, Debra T Silverman 1, Rena R Jones 1
PMCID: PMC12716236  NIHMSID: NIHMS2112463  PMID: 41027348

Abstract

Background and Aim:

Outdoor air pollution, including fine particulate matter (PM2.5), is an established cause of lung cancer; however, few studies have evaluated associations by histologic subtype.

Methods:

We estimated outdoor PM2.5 and nitrogen dioxide (NO2) concentrations at the residential enrollment (1995–1996) address for 486,101 participants of the NIH-AARP Diet and Health Study, a cohort located in 6 states and 2 metropolitan areas. We used annual estimates from a national spatiotemporal model to assess concentrations for a 5-year pre-enrollment period (1990–1994). We used Cox regression to estimate hazard ratios and 95% confidence intervals (HR[CI]) for associations with incident lung cancer overall and by histologic subtype per IQR increase in PM2.5 and NO2. Models were adjusted for demographic characteristics, smoking status and intensity, and census tract-level median household income; we separately evaluated the joint effects for both pollutants. We evaluated effect modification by smoking status, sex, and study catchment area, and evaluated statistical interaction via Wald tests.

Results:

22,453 lung cancers were diagnosed in follow-up through 2018. For PM2.5, we observed a small increased risk of lung cancer overall (HR=1.01[0.99–1.03] per 4.4μg/m3). Associations were stronger for adenocarcinoma (HR=1.05[1.01–1.09]) and squamous cell carcinoma (HR=1.05[1.00–1.10]). In models of NO2, a 10.4 ppb increase was associated with lung cancer overall (HR=1.02[1.00–1.04]) and adenocarcinoma (HR=1.06[1.03–1.09]). When both pollutants were considered simultaneously, stronger associations were noted between squamous cell carcinoma and PM2.5 and between adenocarcinoma and NO2. No clear differences in risk were noted by smoking status or sex (p-interaction all >0.05). Associations varied widely between the 8 cohort catchment areas (p-interaction <0.01).

Conclusions:

Our findings from this large U.S. cohort indicate that risk of squamous cell carcinoma increases with PM2.5 and adenocarcinoma with NO2. Observed heterogeneity in associations by region suggesting that PM2.5 constituency may influence differences in lung cancer risk.

Keywords: ambient air pollution, particulate matter, lung cancer, adenocarcinoma, squamous cell carcinoma

Graphical Abstract

graphic file with name nihms-2112463-f0001.jpg

Introduction

Traffic-related air pollution, a complex mixture of particulate matter (PM), nitrogen dioxide (NO2), and various other pollutants has been established as a major contributor to the burden of non-smoking related lung cancer worldwide (Cohen et al., 2017). The International Agency for Research on Cancer classified outdoor air pollution, and specifically particulate matter, as carcinogenic to humans based on sufficient evidence of a relationship with the development of lung cancer in epidemiologic studies and experimental animals, with strong mechanistic evidence in exposed humans (IARC, 2015). Epidemiologic studies from Europe (Beelen et al., 2008; Fischer et al., 2015; Heinrich et al., 2013; Raaschou-Nielsen et al., 2013), North America (Cheng et al., 2022; Laden et al., 2006; Pope et al., 2011; Puett et al., 2014), and Asia (Cao et al., 2011; Katanoda et al., 2011; Wong et al., 2016) have consistently found associations between lung cancer and both fine PM (PM2.5, diameter <2.5μm) and nitrogen dioxide.

While many studies have examined the overall relationship between air pollution and lung cancer, few have evaluated associations by histologic subtype. A small number of studies have shown an association between PM2.5 (Hystad et al., 2013; Puett et al., 2014) or PM10 (Moon et al., 2020) and increased risk of adenocarcinoma, the most common subtype of lung cancer and the most common among non-smokers. Associations for less common subtypes, including squamous cell and small cell carcinomas, have rarely been reported (Wang et al., 2023) as studies are often limited by the relatively smaller numbers of these cancers. Whether certain demographic or lifestyle characteristics modify associations between air pollution and lung cancer is also of interest. While several studies have found stronger associations with lung cancer overall among men (Hystad et al., 2013; McDonnell et al., 2000) and among never or former smokers (Hystad et al., 2013; Puett et al., 2014), no studies have examined these associations jointly by histologic subtype. Analyses by subtype offer more homogeneous groups to study, affording the discernment of potential effects between groups. Additionally, observation of etiologic heterogeneity between subtype groups can potentially inform carcinogenic mechanisms.

Despite declines in smoking prevalence over time (Meza et al., 2023), lung cancer remains the leading cause of cancer-related death in the U.S (National Cancer Institute, 2025). Given the known role of outdoor air pollution in lung cancer development and the widespread nature of these exposures additional research to improve understanding of this etiologic relationship may have considerable public health impact. Our goal was to evaluate associations with PM2.5 and NO2 and incident lung cancer, overall and by histologic subtype, while adjusting for smoking status and intensity and other potential confounders.

Methods

Study population and cancer ascertainment

The NIH-AARP Diet and Health Study is a prospective cohort that was recruited in 1995–1996 from the membership of the AARP (formerly the American Association of Retired Persons) in six states (California, Florida, Louisiana, New Jersey, North Carolina, Pennsylvania) and two metropolitan areas (Atlanta, Georgia, and Detroit, Michigan). Study details have been described previously (Schatzkin et al., 2001). Briefly, participants were 50 to 71 years old at the time of enrollment and completed a self-administered mailed questionnaire that included demographic characteristics, behavioral and dietary factors, and their health status and medical history. The study was approved by the National Institutes of Health Institutional Review Board. Of 566,398 participants at enrollment, we excluded those with a pre-enrollment cancer diagnosis except non-melanoma skin cancer (N=51,132), whose death was attributed to cancer but not found in registries (N=12,899) and proxy questionnaire respondents (N=15,760). Among the remaining cancer-free participants, we excluded those with missing air pollution data (N=252) or for whom person-time could not be estimated (N=254). After exclusions, 486,101 study participants were included in the analytic population.

Incident lung cancer cases were identified through linkage to cancer registries in the catchment states and in additional states to which participants most frequently moved (Arizona, Nevada, and Texas). The validity of the cancer ascertainment was previously demonstrated, with >90% sensitivity and 99.5% specificity (Michaud et al., 2005). Vital status was determined by linkage with the Social Security Administration Death Master File and the National Death Index. Participants were censored at the earliest date of either cancer diagnosis, relocation out of the registry areas as determined through routine tracing efforts, death, or the end of follow-up (December 31, 2018). Over a median of 22 years of follow-up, 22,453 primary lung cancers were diagnosed. Major histologic subtypes defined according to the International Classification of Diseases for Oncology-3 (codes C340-C349) included adenocarcinoma (41%), small cell carcinoma (12%), and squamous cell carcinoma (20%). The remaining 23% of mixed or unknown subtypes were excluded from histologic-specific analyses.

Exposure assessment

Yearly exposure estimates for PM2.5 and NO2 at each participant’s geocoded enrollment address were obtained from previously-validated spatiotemporal prediction models (Keller et al., 2015; Kim et al., 2017; Kirwa et al., 2021). PM2.5 and NO2 models were developed independently but used the same overall framework. Briefly, PM2.5 measurements from the U.S. Environmental Protection Agency’s Federal Reference Method (FRM) network (1999–2010) and the Interagency Monitoring of Protected Visual Environment (IMPROVE) network (1999–2012) were used in the development of the historical PM2.5 model, which incorporated approximately 300 geographic predictors in universal kriging. To estimate PM2.5 concentrations from before monitoring data were widely available, temporal trend estimation for 1980 to 2010 was determined using the FRM and IMPROVE data, Clean Air Status and Trends Network annual average PM2.5 sulfate concentrations (1987–2010), and Weather Bureau Army Navy network visual ranges (1980–2010). In the validation effort, the historic PM2.5 model generally performed well, though there was limited availability of pre-1999 PM2.5 measurements (Kim et al., 2017). Annual averages were estimated at each participant’s enrollment address for each year back to 1980 for PM2.5 and 1990 for NO2.

Statistical analysis

We used Cox proportional hazards models with age as the time scale to estimate hazard ratios (HR) and 95% confidence intervals (CI) of the relationship between PM2.5 and NO2 exposures and risk of lung cancer. We evaluated the proportional hazards assumption for the main exposures by including interaction terms with follow-up time and using the Wald procedure to test whether coefficients equaled zero. We primarily focused on exposures during the 5-year period directly before enrollment (1990–1994), evaluating associations in quartiles and as a continuous variable (per IQR increase). We also estimated relationships in two other pre-enrollment time periods for exposure to PM2.5, 1980–1984 and 1985–1989. We evaluated the linearity of associations with continuous exposures using restricted cubic splines. Log-likelihood tests were used to assess deviations from linearity. For tests of trend in categorical analyses, we treated the median value of each exposure category as a continuous variable and evaluated the linearity of relationships with a Wald test. As a sensitivity analysis, we conducted models using time-varying exposures over the entire study period using 5-year moving averages. As PM2.5 estimates were available starting in 1980 (as opposed to 1990 for NO2), 5-year averages for PM2.5 were lagged 10 years to account for cancer latency.

We selected covariates at the time of enrollment based on a review of the literature for potential risk factors for lung cancer and possible correlates of exposure. In addition to inherent adjustment for age (used for time-scale), all models were adjusted for sex, race and ethnicity (Non-Hispanic White, Non-Hispanic Black, Hispanic, Asian/Pacific Islander/American Indian/Native Alaskan), smoking status and intensity (never, former/<1 pack/day, former/1–2 packs/day, former/2+ packs/day, current/just quit/<1 pack/day, current/just quit/1–2 packs/day, current/just quit/2+ packs/day), catchment area, and census tract median household income (continuous). We considered additional covariates for inclusion using a step-wise change in effect estimate (>10%) criterion, but no variables were selected. Considered covariates included education (less than high school or high school equivalent, post high-school or some college, college graduate, postgraduate, unknown), body mass index (kg/m2), alcohol use, and physical activity. We also ran separate models for PM2.5 and NO2 that mutually adjusted for continuous exposure to the other pollutant.

We evaluated effect modification by sex (male or female), smoking status (never, former, or current smoker), and catchment state in stratified models. P-values for heterogeneity were estimated using a Wald chi-squared test comparing models with and without cross-product terms between the continuous exposure and non-reference levels of the modifier. Given the correlation between the two pollutants (Spearman’s rho=0.73), we also conducted analyses of PM2.5 and NO2 with a common referent group of low exposure (1st tertile) to both pollutants. We conducted sensitivity analyses restricted to participants with high quality geocoded addresses (street level or better, N=443,949, 91.3%). Given the older age of the cohort, we examined death as a competing risk using Fine-Gray subdistribution hazard models of all-cause mortality (other than death due to lung cancer) as an additional sensitivity analysis. We also stratified models by follow-up period, examining associations independently from baseline through 2006 (approximately 10 years from enrollment) and from 2007 through 2018.

All statistical analyses were performed in SAS v. 9.4 and R (v4.2.2, Boston, MA). We used a two-sided alpha level of 0.05 without adjustment for multiple comparisons for determining statistical significance.

Results

Most participants in the analysis were non-Hispanic White (90.9%), followed by non-Hispanic Black (4.0%), Hispanic (2.0%), and Asian, Pacific Islander, American Indian/Alaskan (grouped as other, 1.7%). The relative proportions by race and ethnicity varied across exposure status; 1.5% of those in the lowest exposure quartile were non-Hispanic Black compared to 8% of those in the top quartile (Table 1). Participants in the first quartile of exposure were also less likely to have never smoked compared to those in the fourth quartile (31.8% vs 36.5%; Table 1). No differences in exposure levels were observed by age, sex, or education. The 5th to 95th percentile intervals for five-year estimated exposures ranged from 10.7 – 21.5 μg/m3 for PM2.5 and 4.8 – 33.0 ppb for NO2. Exposure levels also varied by catchment state (Supplemental Table 1). Mean and median PM2.5 values were highest in Georgia, Pennsylvania, and Michigan, and NO2 levels were highest in and California, Pennsylvania and New Jersey. Florida, Louisiana, and North Carolina had the lowest levels for both pollutants. NO2 and PM2.5 showed positive correlation among all participants (0.73), but correlations varied across catchment states (Supplemental Table 2).

Table 1.

Selected demographic and other baseline characteristics of the AARP study population across quartiles of residential PM2.5 concentrationsa

Characteristic Q1 Q2 Q3 Q4
(3.5–13.1 μg/m3) (13.1–15.5 μg/m3) (15.5–17.5 μg/m3) (17.5–26.9 μg/m3)
Age (mean ± sd) 62.5 ± 5.3 61.8 ± 5.4 61.8 ± 5.4 62.0 ± 5.4
Sex (%)
 Male 60.5 60.3 60.9 56.4
 Female 39.5 39.7 39.1 43.6
Race/ethnicity (%)
 Non-Hispanic White 94.5 92.6 92.0 85.1
 Non-Hispanic Black 1.5 2.5 3.8 8.0
 Hispanic 1.9 2.0 1.3 2.6
 Other 0.9 1.6 1.7 2.5
 Unknown 1.3 1.3 1.2 1.8
Smoking status (%)
 Never 31.8 35.8 36.9 36.5
 Former 50.2 47.3 46.5 45.1
 Current 14.3 13.1 12.8 14.3
 Unknown 3.7 3.8 3.8 4.1
Median Household Income ($; mean ± sd) 46,582 ± 18,774 56,761 ± 25,789 60,501 ± 25,512 51,073 ± 21,319
Highest schooling achieved (%)
 Less than high school 6.3 5.9 5.6 6.4
 Completed high school 19.3 19.5 19.4 20.3
 Some college 35.8 32.7 30.8 32.5
 College and post-graduate 35.6 39.0 41.5 37.7
 Unknown 3.1 2.9 2.7 3.2
State (Metropolitan area) (%)
 California 27.0 28.0 24.2 44.4
 Florida 59.1 22.8 2.9 0.0
 Georgia (Atlanta) 0.0 0.5 4.0 6.8
 Louisiana 3.1 6.5 5.4 0.3
 Michigan (Detroit) 0.5 3.7 7.9 8.0
 North Carolina 4.3 11.3 13.5 3.8
 New Jersey 4.2 17.5 20.5 9.0
 Pennsylvania 1.8 9.7 21.7 27.8
NO2 (ppb; mean ± sd)a 8.5 ± 3.6 12.1 ± 5.2 16.0 ± 5.8 24.7 ± 10.0
Open in a new tab
a

Concentrations of PM2.5 and NO2 calculated as 5-year average (1990–1994)

In models of continuous exposure, we observed a small increased risk in lung cancer overall (HR=1.01; 95%CI=0.99–1.03) and a 5% increased risk of adenocarcinoma (HR=1.05; 95%CI=1.01–1.09) and squamous cell carcinoma (HR=1.05; 95%CI=1.00–1.10) per 4.4 μg/m3 increase in PM2.5 from 1990 through 1994 (Table 2). These associations were consistent in the categorical analyses, with exposure-response trends apparent over increasing quartiles for lung cancer overall (HRQ4vQ1=1.06; 95%CI=1.01–1.11; p-trend=0.03) and stronger association for adenocarcinoma (HRQ4vQ1=1.15; 95%CI=1.06–1.24; p-trend<0.01) and squamous cell carcinoma (HRQ4vQ1=1.14; 95%CI=1.02–1.28; p-trend=0.02). Results for PM2.5 were nearly identical for the other two exposure time periods examined (1980–1984 and 1985–1989). NO2 exposure was also associated with a small increased risk of lung cancer overall (HR=1.02; 95%CI=1.00–1.04 per 10.4 ppb) and adenocarcinoma (HR=1.06; 95%CI=1.03–1.09 per 10.4 ppb), although the association with squamous cell carcinoma was weaker (HR=1.03; 95%CI=0.98–1.07). Categorical analyses also yielded exposure-response trends; participants in the highest exposure quartile had an increased risk of lung cancer overall (HRQ4vsQ1=1.10; 95%CI=1.05–1.15; p-trend<0.01) and adenocarcinoma (HRQ4vsQ1=1.22; 95%CI=1.14–1.31; p-trend<0.01), with a weaker association observed for squamous cell carcinoma (HRQ4vsQ1=1.08; 95%CI=0.98–1.07; p-trend=0.07). We observed no increased risk in small cell carcinoma associated with exposure to either pollutant.

Table 2.

Associationa between PM2.5 and NO2 exposures and risk of lung cancer, overall and by histologic type

All lung cancers Adenocarcinoma Small cell carcinoma Squamous cell carcinoma
Exposure N cases HR (95% CI) N cases HR (95% CI) N cases HR (95% CI) N cases HR (95% CI)
PM2.5 1980–1984
 Q1 5,980 1.0 (Ref) 2,332 1.0 (Ref) 724 1.0 (Ref) 1,162 1.0 (Ref)
 Q2 5,467 1.03 (0.99–1.07) 2,176 1.04 (0.97–1.11) 635 0.96 (0.85–1.07) 1,079 1.08 (0.99–1.18)
 Q3 5,413 1.06 (1.02–1.11) 2,248 1.12 (1.04–1.20) 622 0.94 (0.83–1.07) 1,052 1.12 (1.02–1.25)
 Q4 5,593 1.05 (1.01–1.10) 2,337 1.14 (1.06–1.23) 649 0.91 (0.80–1.04) 1,099 1.14 (1.03–1.27)
p-trend 0.02 <0.01 0.18 0.01
Continuousb 22,453 1.02 (0.99–1.04) 9,093 1.06 (1.02–1.09) 2,630 0.97 (0.91–1.04) 4,392 1.05 (1.00–1.10)
PM2.5 1985–1989
 Q1 5,989 1.0 (Ref) 2326 1.0 (Ref) 724 1.0 (Ref) 1167 1.0 (Ref)
 Q2 5,463 1.04 (1.00–1.08) 2190 1.06 (0.99–1.13) 639 0.97 (0.87–1.10) 1077 1.09 (0.99–1.19)
 Q3 5,439 1.08 (1.03–1.13) 2244 1.13 (1.05–1.22) 628 0.95 (0.83–1.09) 1052 1.13 (1.02–1.25)
 Q4 5,562 1.05 (1.01–1.10) 2333 1.15 (1.07–1.24) 639 0.90 (0.78–1.03) 1096 1.15 (1.03–1.28)
p-trend 0.03 <0.01 0.11 0.01
 Continuousb 22,453 1.01 (0.99–1.04) 9,093 1.05 (1.02–1.09) 2,630 0.97 (0.91–1.03) 4,392 1.05 (1.00–1.10)
PM2.5 1990–1994
 Q1 6003 1.0 (Ref) 2346 1.0 (Ref) 712 1.0 (Ref) 1170 1.0 (Ref)
 Q2 5455 1.04 (1.00–1.09) 2188 1.05 (0.98–1.12) 642 1.02 (0.90–1.15) 1078 1.10 (1.00–1.21)
 Q3 5446 1.08 (1.03–1.13) 2222 1.11 (1.03–1.19) 635 0.99 (0.87–1.14) 1064 1.14 (1.03–1.27)
 Q4 5549 1.06 (1.01–1.11) 2337 1.15 (1.06–1.24) 641 0.94 (0.82–1.08) 1080 1.14 (1.02–1.28)
p-trend 0.03 <0.01 0.31 0.02
 Continuousb 22,453 1.01 (0.99–1.03) 9,093 1.05 (1.01–1.09) 2,630 0.96 (0.90–1.03) 4,392 1.05 (1.00–1.10)
NO2 1990–1994
 Q1 5637 1.0 (Ref) 2063 1.0 (Ref) 710 1.0 (Ref) 1194 1.0 (Ref)
 Q2 5558 1.03 (0.99–1.07) 2251 1.10 (1.03–1.17) 643 0.96 (0.86–1.07) 1069 0.97 (0.89–1.05)
 Q3 5669 1.10 (1.06–1.15) 2416 1.23 (1.15–1.31) 668 1.03 (0.92–1.16) 1055 1.02 (0.93–1.12)
 Q4 5589 1.10 (1.05–1.15) 2363 1.22 (1.14–1.31) 609 0.97 (0.85–1.11) 1074 1.08 (0.98–1.20)
p-trend <0.01 <0.01 0.83 0.07
 Continuousb 22,453 1.02 (1.00–1.04) 9,093 1.06 (1.03–1.09) 2,630 0.97 (0.92–1.03) 4,392 1.03 (0.98–1.07)
Open in a new tab
a

Models adjusted for: age, state, sex, race and ethnicity, census tract median household income and smoking status and intensity.

b

Association per IQR 5.1 μg/m3 (PM2.5 1980–1984), 4.8 μg/m3 (PM2.5 1985–1989), 4.4 μg/m3 (PM2.5 1990–1994), 10.4 ppb(NO2 1990–1994).

In stratified models, we saw no evidence of statistical interaction between smoking status and either PM2.5 (Table 3) or NO2 (Table 4) on lung cancer associations (p-interaction>0.05). Risks were higher for males than for females overall and for each histologic subtype for both pollutants, although tests of interaction were not statistically significant (p-interaction>0.05). Risk of lung cancer overall for both pollutants also varied by catchment state (p-interaction<0.01). PM2.5 associations with lung cancer overall were highest in Pennsylvania (HR=1.17; 95%CI=1.08–1.27 per IQR increase) and New Jersey (HR=1.13; 95%CI=1.04–1.24), and the patterns were similar for squamous cell carcinoma (Table 3). For NO2, the associated risk of adenocarcinoma was highest in Pennsylvania (HR=1.27; 95%CI=1.14–1.40) and Florida (HR=1.24;95%CI-1.09–1.42); Table 4.

Table 3.

Associationa between PM2.5 (1990–1994) and risk of lung cancer, stratified by demographic and lifestyle characteristics and state

All lung cancers Adenocarcinoma Small cell carcinoma Squamous cell carcinoma
Group N cases HR (95% CI) N cases HR (95% CI) N cases HR (95% CI) N cases HR (95% CI)
PM2.5 (1990–1994)
Never Smokers 1,553 1.00 (0.92–1.09) 914 1.02 (0.92–1.14) 60 0.94 (0.59–1.48) 101 1.14 (0.80–1.61)
Former Smokers 10,902 1.02 (0.99–1.06) 4,930 1.05 (1.00–1.10) 972 1.01 (0.90–1.13) 1,977 1.08 (1.00–1.17)
Current Smokers 9,186 1.00 (0.97–1.04) 2,940 1.07 (1.00–1.14) 1,497 0.94 (0.86–1.03) 2,143 1.02 (0.95–1.09)
p-interaction 0.51 0.88 0.86 0.81
Male 13,876 1.03 (1.00–1.06) 5,371 1.06 (1.01–1.11) 1,545 1.02 (0.93–1.11) 3,063 1.08 (1.01–1.15)
Female 8,577 0.98 (0.95–1.02) 3,722 1.04 (0.99–1.10) 1,085 0.90 (0.81–0.99) 1,329 1.00 (0.91–1.09)
p-interaction 0.16 0.38 0.18 0.64
California 6,395 1.00 (0.97–1.02) 2,749 1.04 (1.00–1.09) 660 0.95 (0.87–1.03) 1,098 1.01 (0.95–1.08)
Florida 5,398 1.02 (0.95–1.11) 2,118 1.04 (0.92–1.17) 633 1.04 (0.83–1.30) 1,097 1.00 (0.84–1.19)
Georgia 582 0.89 (0.68–1.15) 242 0.74 (0.50–1.11) 66 1.30 (0.59–2.85) 109 0.71 (0.39–1.29)
Louisiana 850 0.90 (0.75–1.07) 294 1.08 (0.80–1.47) 102 0.51 (0.31–0.84) 221 1.02 (0.71–1.45)
Michigan 1,000 0.99 (0.83–1.18) 385 0.94 (0.71–1.25) 137 0.76 (0.48–1.20) 211 0.89 (0.61–1.30)
North Carolina 1,722 0.93 (0.84–1.03) 626 0.97 (0.82–1.16) 232 1.08 (0.81–1.43) 344 0.91 (0.72–1.14)
New Jersey 2,959 1.13 (1.04–1.24) 1,265 1.11 (0.97–1.27) 330 0.96 (0.75–1.24) 614 1.35 (1.12–1.64)
Pennsylvania 3,547 1.17 (1.08–1.27) 1,414 1.25 (1.10–1.44) 470 1.08 (0.87–1.36) 698 1.39 (1.15–1.68)
p-interaction <0.01 0.13 0.19 <0.01
Open in a new tab
a

Models adjusted for: age, state, sex, race and ethnicity, census-tract median household income and smoking status and intensity (when not stratified on that factor). Associations per IQR; 4.4 μg/m3 (PM2.5 1990–1994).

Table 4.

Associationa between NO2 (1990–1994) and risk of lung cancer, stratified by demographic and lifestyle characteristics and state

All lung cancers Adenocarcinoma Small cell carcinoma Squamous cell carcinoma
Group N cases HR (95% CI) N cases HR (95% CI) N cases HR (95% CI) N cases HR (95% CI)
NO2 (1990–1994)
Never Smokers 1,553 1.01 (0.95–1.08) 914 1.05 (0.97–1.15) 60 0.90 (0.61–1.32) 101 1.03 (0.78–1.37)
Former Smokers 10,902 1.03 (1.00–1.06) 4,930 1.06 (1.02–1.10) 972 1.01 (0.92–1.11) 1,977 1.03 (0.96–1.10)
Current Smokers 9,186 1.01 (0.98–1.04) 2,940 1.05 (1.00–1.11) 1,497 0.96 (0.89–1.03) 2,143 1.02 (0.96–1.09)
p-interaction 0.38 0.88 0.46 0.38
Male 13,876 1.04 (1.02–1.07) 5,371 1.07 (1.03–1.12) 1,545 1.04 (0.96–1.12) 3,063 1.03 (0.97–1.08)
Female 8,577 1.00 (0.97–1.02) 3,722 1.04 (0.99–1.08) 1,085 0.89 (0.82–0.97) 1,329 1.02 (0.95–1.10)
p-interaction 0.69 0.91 0.07 0.45
California 6,395 1.00 (0.98–1.02) 2,749 1.02 (0.98–1.05) 660 0.96 (0.89–1.03) 1,098 1.01 (0.96–1.07)
Florida 5,398 1.04 (0.96–1.13) 2,118 1.24 (1.09–1.42) 633 0.86 (0.68–1.09) 1,097 0.78 (0.65–0.94)
Georgia 582 0.89 (0.72–1.11) 242 0.82 (0.58–1.15) 66 0.92 (0.49–1.75) 109 0.80 (0.49–1.32)
Louisiana 850 0.89 (0.77–1.04) 294 1.12 (0.87–1.44) 102 0.78 (0.49–1.22) 221 0.85 (0.63–1.16)
Michigan 1,000 0.97 (0.82–1.15) 385 0.93 (0.70–1.21) 137 0.71 (0.45–1.10) 211 0.85 (0.59–1.23)
North Carolina 1,722 0.96 (0.85–1.08) 626 1.01 (0.83–1.22) 232 1.18 (0.87–1.60) 344 0.83 (0.64–1.09)
New Jersey 2,959 1.09 (1.03–1.15) 1,265 1.11 (1.02–1.22) 330 1.02 (0.86–1.20) 614 1.23 (1.09–1.38)
Pennsylvania 3,547 1.16 (1.09–1.24) 1,414 1.27 (1.14–1.40) 470 1.08 (0.91–1.28) 698 1.21 (1.05–1.40)
p-interaction <0.01 <0.01 0.45 <0.01
Open in a new tab
a

Models adjusted for: age, state, sex, race and ethnicity, census tract median household income and smoking status and intensity (when not stratified on that factor). Associations per IQR (10.4 ppb).

When we examined joint effects of PM2.5 and NO2, statistical interaction was noted between both pollutants for models of lung cancer overall and adenocarcinoma (Table 5). Using a common referent approach to examine risk within categorized exposures, the highest tertile (T3) for both PM2.5 and NO2 was associated with an 8% increased risk of lung cancer (HR=1.08; 95%CI=1.03–1.14) compared to the common referent (T1 PM2.5 and T1 NO2). The association for the category of high (T3) PM2.5 and low (T1) NO2 (HR=1.09; 95%CI=0.96–1.23) was elevated as was the association in the high (T3) NO2 and low (T1) PM2.5 category (HR=1.10; 95%CI=0.96–1.25) suggesting both pollutants increased lung cancer risk. For adenocarcinoma, participants who were jointly exposed to the highest levels of both pollutants (i.e., in T3 PM2.5 and T3 NO2) had a 22% increased risk (HR=1.22; 95% CI=1.13–1.31). However, adenocarcinoma was primarily associated with NO2 exposure, as the risk was strongest in association with combined exposure to the highest NO2 tertile and the lowest PM2.5 tertile (HR=1.28; 95%CI=1.05–1.56), nearly 20% higher than the risk in the highest PM2.5 tertile and lowest NO2 tertile (HR=1.12; 95%CI=0.92–1.37). In contrast, associations with squamous cell carcinoma were strongest in the top tertile of PM2.5 across all tertiles of NO2.

Table 5.

Associationa between lung cancer and tertilesb of PM2.5 and NO2 using a common referent approach

All lung cancers
T1 NO2 T2 NO2 T3 NO2
PM2.5 N cases HR (95% CI) N cases HR (95% CI) N cases HR (95% CI) Interactionc
T1 5262 1.0 (Ref) 2392 1.04 (0.99–1.09) 238 1.10 (0.96–1.25)
T2 2014 1.00 (0.95–1.07) 2982 1.07 (1.02–1.13) 2157 1.09 (1.03–1.16)
T3 296 1.09 (0.96–1.23) 2016 1.01 (0.95–1.08) 5096 1.08 (1.03–1.14) p-int<0.01
Adenocarcinoma
T1 NO2 T2 NO2 T3 NO2
PM2.5 N cases HR (95% CI) N cases HR (95% CI) N cases HR (95% CI) Interactionc
T1 1967 1.0 (Ref) 1005 1.13 (1.05–1.22) 108 1.28 (1.05–1.56)
T2 753 1.04 (0.94–1.14) 1288 1.19 (1.10–1.29) 902 1.14 (1.04–1.24)
T3 114 1.12 (0.92–1.37) 805 1.08 (0.98–1.20) 2151 1.22 (1.13–1.31) p-int<0.01
Small cell carcinoma
T1 NO2 T2 NO2 T3 NO2
PM2.5 N cases HR (95% CI) N cases HR (95% CI) N cases HR (95% CI) Interactionc
T1 649 1.0 (Ref) 276 1.01 (0.87–1.16) 19 0.75 (0.47–1.18)
T2 235 0.84 (0.71–0.99) 345 0.99 (0.86–1.15) 248 1.08 (0.92–1.28)
T3 40 1.06 (0.75–1.48) 248 0.88 (0.74–1.05) 570 0.92 (0.80–1.06) p-int=0.92
Squamous cell carcinoma
T1 NO2 T2 NO2 T3 NO2
PM2.5 N cases HR (95% CI) N cases HR (95% CI) N cases HR (95% CI) Interactionc
T1 1098 1.0 (Ref) 405 0.88 (0.78–0.98) 40 0.97 (0.70–1.33)
T2 434 1.08 (0.95–1.22) 553 1.02 (0.91–1.15) 393 1.08 (0.95–1.23)
T3 57 1.14 (0.86–1.52) 412 1.10 (0.96–1.27) 1000 1.12 (1.01–1.25) p-int=0.46
Open in a new tab
a

Models adjusted for: age, state, sex, race and ethnicity, census tract median household income and smoking status and intensity.

b

1990–1994 tertile ranges for PM2.5 T1: 3.5–13.9 μg/m3; T2: 13.9–16.8 μg/m3; T3: 16.8–26.9 μg/m3. Tertile ranges for NO2 T1:0.8–10.4 ppb; T2: 10.4–17.1 ppb; T3: 17.1–87.5 ppb.

c

Interaction tests are between continuous variables for PM2.5 and NO2.

The results from multipollutant models (Supplemental Table 3) were consistent with the joint effects analysis. PM2.5 models of squamous cell carcinoma did not change with adjustment for NO2, but models of PM2.5 were attenuated, though still elevated in categorical analysis, for lung cancer overall and for adenocarcinoma. For NO2 models, PM2.5 adjustment widened confidence intervals but had no change on the effect estimate with adenocarcinoma (HRTable2=1.06; 95%CI=1.03–1.09; HRadj=1.06; 95%CI=1.01–1.10), however associations with squamous cell carcinoma were attenuated to the null after adjustment (HR=0.99; 95%CI=0.93–1.06). State-specific models of PM2.5 showed attenuated associations overall and for adenocarcinoma with mutual adjustment for NO2; most notably, the positive association with adenocarcinoma in PA was attenuated to the null. Positive associations with squamous cell carcinoma and New Jersey and PA remained (Supplemental Table 4). In contrast, mutual adjustment of the NO2 models for PM2.5 did not materially change associations with adenocarcinoma, but associations with squamous cell carcinoma were weakened. Spline analyses showed general consistency with the linear and categorical models (Supplemental Figures 1 and 2). For both PM2.5 and NO2, hazard ratios for lung cancer overall, adenocarcinoma, and squamous cell carcinoma increased approximately linearly over most of the exposure distribution with hazard ratios plateauing at the highest exposure levels. In sensitivity analysis, results were similar when restricted to high quality geocodes (Supplemental Table 5) and in models accounting for death as a competing risk (Supplemental Table 6). Effect estimates from models using time-varying exposures were similar in magnitude to those using fixed time-windows for lung cancer overall, adenocarcinoma, and squamous cell carcinoma for both PM2.5 and NO2 (Supplemental Table 7). Additionally, our results were consistent in analyses stratified by follow-up period (Supplemental Table 8).

Discussion

In the largest cohort investigation of this relationship in the U.S. to-date, we observed a positive association with residential levels of outdoor PM2.5 and squamous cell carcinoma of the lung in analyses accounting for concomitant NO2 exposure. NO2 exposure was more consistently associated with lung cancer overall and with adenocarcinoma, the predominant histologic subtype. We also observed some evidence of heterogeneity in associations by region but no clear interactions by sex or smoking status.

Prior studies have demonstrated a positive relationship between PM2.5 and both lung cancer incidence and mortality (Cheng et al., 2022; Hamra et al., 2014; Pope et al., 2019), and independent meta-analyses of these and other studies that account for smoking and other potential confounders have reported that a 10 μg/m3 increase in exposure to PM2.5 and NO2 was associated with a 9–16% and 4% increase in lung cancer risk, respectively (Ciabattini et al., 2021; Hamra et al., 2015, 2014). We observed a very modest association with PM2.5 and lung cancer overall of similar magnitude (2% increase per ~5 μg/m3) with these prior studies. Our analyses yielded positive associations for PM2.5 and squamous cell carcinoma subtype and a weaker association with adenocarcinoma. This pattern was consistent across stratified analyses as well. We observed no clear relationship with small cell carcinoma of the lung. Few studies of incident lung cancer evaluated associations by histologic subtype to enable comparison with our findings. In the Multiethnic Cohort Study, a positive overall association with lung cancer risk was observed, but patterns were similar across major histologic cell types, including adenocarcinoma, squamous and small cell carcinomas (Cheng et al., 2022). In the Nurses’ Health Study, the association with outdoor PM2.5 and lung cancer was evident and stronger in analyses restricted to adenocarcinoma; numbers of cases of other cell types were too few for analysis (Puett et al., 2014).

Prior studies investigating PM2.5 exposures have frequently also observed an association with NO2 (Cheng et al., 2022; Hamra et al., 2015). However, because NO2 only imparts weak mutagenic effects (Koehler et al., 2013; Victorin, 1994), many studies have concluded that this association is being driven by other carcinogenic components of the outdoor air pollution mixture arising from traffic, for which NO2 serves as a proxy. We observed a consistent relationship between NO2 exposure and adenocarcinoma of the lung that was robust to adjustment for PM2.5 in the main and across most of the stratified analyses. Additionally, the PM2.5-adenocarcinoma association was somewhat weaker after accounting for NO2. Because of the positive correlation between PM2.5 and NO2 exposures, we additionally evaluated risk cross-classified by PM2.5 and NO2 in an attempt to disentangle the effects of these highly correlated pollutants. Our results corroborate the findings of our co-pollutant models, in that both pollutants are contributing to an increased risk of lung cancer in the cohort, although these analyses indicate a stronger PM2.5 association with squamous cell carcinoma and NO2 more so with adenocarcinoma. Whether NO2 is exerting a carcinogenic effect or serves as a proxy for other traffic constituents remains poorly understood (Sigsgaard and Hoffmann, 2024). An analysis in a subset of the NIH-AARP cohort in Southern California yielded an adenocarcinoma association with ultrafine particulate matter, another common traffic-related pollutant and common correlate of NO2, a finding that was robust to adjustment for both PM2.5 and NO2 (Jones et al., 2024).

The mechanisms underlying the overall air pollution-lung cancer association include inflammation, oxidative stress, DNA damage and epigenetic alterations, among others (Wang et al., 2023). How these mechanisms might contribute to differential associations by histologic type is not understood. Subtypes arise from different types of cells in the lung that potentially have varying susceptibility to pollutants and pathways of carcinogenesis (Spitz et al., 2006). It is possible that there are relevant molecular and cellular changes following air pollution exposure that may vary depending on the subtype. This is supported by analyses of cigarette smoking, which suggest different biological responses depending on the exposure and extent of tissue damage and location of cells with lower differentiation and higher potency for regeneration that are more prevalent in anatomically distinct areas of the lung (Pesch et al., 2012). Air pollutants may also influence different components of the tumor microenvironment, influencing tumor initiation for specific subtypes differently. Genetic susceptibility could also interact with air pollutants to influence differential associations by subtype (Wang et al., 2023).

Our models stratified by sex yielded limited evidence of statistical heterogeneity by these factors, although the PM2.5 association for lung cancer overall and for squamous cell carcinoma was apparent only among men. The reason for the sex differences is unclear, although several other studies have similarly observed a PM2.5 association with lung cancer only in men (Hystad et al., 2013; McDonnell et al., 2000). Occupational information and time spent at the residence were not available for the cohort, both of which could influence the extent of exposure misclassification by sex.

A number of other studies have reported elevated air pollution associations with lung cancer that are stronger among never or former smokers, mitigating concerns about confounding by smoking (Hystad et al., 2013; Puett et al., 2014). In our study, associations between PM2.5 and squamous cell carcinoma were stronger for never and former smokers, but the number of never-smoking cases was relatively small. In contrast, we observed no differences by smoking status for the other subtypes or for the association between NO2 and adenocarcinoma. While smoking is the predominant risk factor for lung cancer, the strength of the smoking association differs across lung cancer subtypes (Thun et al., 2017). Lung cancer histology has also shifted over time, with increases since the 1980s in the proportion of adenocarcinoma cases observed and simultaneous reductions in the proportions of more strongly smoking-related squamous and small cell histotypes (Devesa et al., 2005). These shifts have been attributed to a decline in active cigarette smoking. To our knowledge, ours is the first study to examine PM2.5 and NO2 associations with risk of squamous cell carcinoma and small cell carcinoma by smoking status. Our main analyses were well-controlled for smoking, and our ability to jointly stratify on smoking status and histology offer new clues about air pollution as a possibly also contributing to the observed secular trends in incidence.

We observed some heterogeneity in the PM2.5 -squamous cell carcinoma association by geographic region, particularly elevated risk in Pennsylvania and New Jersey, despite similar distributions of this pollutant in these and other areas, including California and Detroit, where risk was not increased. These findings suggest that differences are potentially due to the known regional variation in the chemical composition of this pollutant (Bell et al., 2007). Major components of PM vary due to its diverse sources (i.e., other than traffic) and include ions (sulfate and nitrate), reactive gases, organic compounds, elemental carbon and other carbonaceous material, and transition metals and minerals. PM2.5 cytotoxicity has been demonstrated via multiple pathways (Valavanidis et al., 2008), suggesting the potential importance of specific components to cancer development. Several cohort studies in the US have found positive associations between lung cancer mortality and long-term exposure to sulfate (Krewski, D. et al., 2001; Thurston et al., 2013), a PM constituent that varies considerably in concentration between eastern and western U.S. (Bell et al., 2007). These findings suggest that future studies evaluate PM constituency to further our understanding of these etiologic relationships. Regional variability in the overall lung cancer and adenocarcinoma associations, albeit more limited, was also observed for NO2, with positive associations observed in Pennsylvania (median=15.4 ppb), Florida (median=9.2 ppb), and New Jersey (median=16.3 ppb) but no other states. The reason for the heterogeneity in these associations is not clear, especially given the differences in NO2 distributions in these areas. We postulate that one explanation may be differences in the correlations between NO2 and other pollutants across these states. Our findings of varying correlations between NO2 and PM2.5 by state may be due to statewide differences in participant residential proximity to combustion sources (e.g., roadways) or differences in the relative proportion of natural versus anthropogenic sources (e.g. sea salt/dust versus traffic).

Strengths and Limitations

Our study was conducted within a large, geographically diverse population with a considerable range in PM2.5 exposures, high-quality cancer registration with histologic subtype, and detailed information on participants’ potential confounders from questionnaire data. Unlike many previous studies, we assessed exposures prior to study start and were able to implement lagged analyses to allow for the long latency of lung cancer.

The potential for exposure misclassification is the primary weakness of the study. Despite our use of a validated historical exposure model, we did not have residence history information for all participants, and residential mobility prior to study enrollment may have led to errors in exposure classification. We note, however, that residential histories estimated for some participants in California, approximately 10% of the total cohort, indicated long duration at the enrollment address prior to the study start (Medgyesi et al., 2021). Given that participants were cancer free at enrollment, we do not expect such errors would be differential by disease status and in general would expect any resulting bias in the observed HRs to be toward the null. Additionally, occupational sources of air pollution exposure may be important to lung cancer risk, and information on workplace exposure was not collected in the cohort questionnaires. We had few measures of individual socioeconomic status apart from education to consider as potential confounders, but note that the cohort is generally of moderate to high SES. Though it is not clear that occupation would be related to one’s residential exposure to PM2.5 and NO2, we cannot exclude the possibility of confounding by occupation in our analysis. Lastly, while our cohort was geographically spread, it was not population-based by design, which may limit the generalizability of our findings.

Conclusions

Independent associations between PM2.5 and risk of squamous cell carcinoma of the lung and between NO2 and adenocarcinoma in this large prospective cohort study offers new insights about these ambient pollutants as risk factors for this common malignancy. The ubiquity of these exposures underscores the potential public health impact of even some of the more modest associations we observed. Future studies exploiting the varying chemical constituency of PM2.5 and evaluating exposure to other near-roadway air pollutants would add valuable new data to further inform our understanding of these relationships.

Supplementary Material

1

Highlights.

  • Conducted within a large prospective cohort of over 500,000 participants in the U.S.

  • Historical PM2.5 and NO2 estimates from a nationwide spatiotemporal model

  • Higher PM2.5 and NO2 concentrations associated with increased lung cancer risk

  • Associations stronger for squamous cell carcinoma and PM2.5; adenocarcinoma and NO2

  • Associations varied between the 8 cohort catchment areas

Acknowledgments

This research was supported [in part] by the Intramural Research Program of the National Institutes of Health (NIH). The contributions of the NIH author(s) are considered Works of the United States Government. The findings and conclusions presented in this paper are those of the author(s) and do not necessarily reflect the views of the NIH or the U.S. Department of Health and Human Services.

Footnotes

Conflicts of interest - The authors declare they have no conflicts of interest related to this work to disclose.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Beelen R, Hoek G, van den Brandt PA, Goldbohm RA, Fischer P, Schouten LJ, Armstrong B, Brunekreef B, 2008. Long-Term Exposure to Traffic-Related Air Pollution and Lung Cancer Risk. Epidemiology 19, 702. 10.1097/EDE.0b013e318181b3ca [DOI] [PubMed] [Google Scholar]
  2. Bell ML, Dominici F, Ebisu K, Zeger SL, Samet JM, 2007. Spatial and Temporal Variation in PM2.5 Chemical Composition in the United States for Health Effects Studies. Environ. Health Perspect. 115, 989–995. 10.1289/ehp.9621 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cao J, Yang C, Li J, Chen R, Chen B, Gu D, Kan H, 2011. Association between long-term exposure to outdoor air pollution and mortality in China: a cohort study. J. Hazard. Mater. 186, 1594–1600. 10.1016/j.jhazmat.2010.12.036 [DOI] [PubMed] [Google Scholar]
  4. Cheng I, Yang J, Tseng C, Wu J, Shariff-Marco S, Park S-SL, Conroy SM, Inamdar PP, Fruin S, Larson T, Setiawan VW, DeRouen MC, Gomez SL, Wilkens LR, Le Marchand L, Stram DO, Samet J, Ritz B, Wu AH, 2022. Traffic-related Air Pollution and Lung Cancer Incidence: The California Multiethnic Cohort Study. Am. J. Respir. Crit. Care Med. 206, 1008–1018. 10.1164/rccm.202107-1770OC [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Ciabattini M, Rizzello E, Lucaroni F, Palombi L, Boffetta P, 2021. Systematic review and meta-analysis of recent high-quality studies on exposure to particulate matter and risk of lung cancer. Environ. Res. 196, 110440. 10.1016/j.envres.2020.110440 [DOI] [PubMed] [Google Scholar]
  6. Cohen AJ, Brauer M, Burnett R, Anderson HR, Frostad J, Estep K, Balakrishnan K, Brunekreef B, Dandona L, Dandona R, Feigin V, Freedman G, Hubbell B, Jobling A, Kan H, Knibbs L, Liu Y, Martin R, Morawska L, Pope CA, Shin H, Straif K, Shaddick G, Thomas M, Dingenen R. van, Donkelaar A. van, Vos T, Murray CJL, Forouzanfar MH, 2017. 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. The Lancet 389, 1907–1918. 10.1016/S0140-6736(17)30505-6 [DOI] [Google Scholar]
  7. Devesa SS, Bray F, Vizcaino AP, Parkin DM, 2005. International lung cancer trends by histologic type: male:female differences diminishing and adenocarcinoma rates rising. Int. J. Cancer 117, 294–299. 10.1002/ijc.21183 [DOI] [PubMed] [Google Scholar]
  8. Fischer PH, Marra M, Ameling CB, Hoek G, Beelen R, de Hoogh K, Breugelmans O, Kruize H, Janssen NAH, Houthuijs D, 2015. Air Pollution and Mortality in Seven Million Adults: The Dutch Environmental Longitudinal Study (DUELS). Environ. Health Perspect. 123, 697–704. 10.1289/ehp.1408254 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Hamra GB, Guha N, Cohen A, Laden F, Raaschou-Nielsen O, Samet JM, Vineis P, Forastiere F, Saldiva P, Yorifuji T, Loomis D, 2014. Outdoor particulate matter exposure and lung cancer: a systematic review and meta-analysis. Environ. Health Perspect. 122, 906–911. 10.1289/ehp.1408092 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Hamra GB, Laden F, Cohen AJ, Raaschou-Nielsen O, Brauer M, Loomis D, 2015. Lung Cancer and Exposure to Nitrogen Dioxide and Traffic: A Systematic Review and Meta-Analysis. Environ. Health Perspect. 123, 1107–1112. 10.1289/ehp.1408882 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Heinrich J, Thiering E, Rzehak P, Krämer U, Hochadel M, Rauchfuss KM, Gehring U, Wichmann H-E, 2013. Long-term exposure to NO2 and PM10 and all-cause and cause-specific mortality in a prospective cohort of women. Occup. Environ. Med. 70, 179–186. 10.1136/oemed-2012-100876 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hystad P, Demers PA, Johnson KC, Carpiano RM, Brauer M, 2013. Long-term residential exposure to air pollution and lung cancer risk. Epidemiol. Camb. Mass 24, 762–772. 10.1097/EDE.0b013e3182949ae7 [DOI] [Google Scholar]
  13. IARC, 2015. Outdoor Air Pollution - IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. [Google Scholar]
  14. Jones RR, Fisher JA, Hasheminassab S, Kaufman JD, Freedman ND, Ward MH, Sioutas C, Vermeulen R, Hoek G, Silverman DT, 2024. Outdoor Ultrafine Particulate Matter and Risk of Lung Cancer in Southern California. Am. J. Respir. Crit. Care Med. 209, 307–315. 10.1164/rccm.202305-0902OC [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Katanoda K, Sobue T, Satoh H, Tajima K, Suzuki T, Nakatsuka H, Takezaki T, Nakayama T, Nitta H, Tanabe K, Tominaga S, 2011. An association between long-term exposure to ambient air pollution and mortality from lung cancer and respiratory diseases in Japan. J. Epidemiol. 21, 132–143. 10.2188/jea.je20100098 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Keller JP, Olives C, Kim S-Y, Sheppard L, Sampson PD, Szpiro AA, Oron AP, Lindström J, Vedal S, Kaufman JD, 2015. A unified spatiotemporal modeling approach for predicting concentrations of multiple air pollutants in the multi-ethnic study of atherosclerosis and air pollution. Environ. Health Perspect. 123, 301–309. 10.1289/ehp.1408145 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kim S-Y, Olives C, Sheppard L, Sampson PD, Larson TV, Keller JP, Kaufman JD, 2017. Historical Prediction Modeling Approach for Estimating Long-Term Concentrations of PM2.5 in Cohort Studies before the 1999 Implementation of Widespread Monitoring. Environ. Health Perspect. 125, 38–46. 10.1289/EHP131 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kirwa K, Szpiro AA, Sheppard L, Sampson PD, Wang M, Keller JP, Young MT, Kim S-Y, Larson TV, Kaufman JD, 2021. Fine-scale air pollution models for epidemiologic research: Insights from approaches developed in the Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air). Curr. Environ. Health Rep. 8, 113–126. 10.1007/s40572-021-00310-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Koehler C, Thielen S, Ginzkey C, Hackenberg S, Scherzed A, Burghartz M, Paulus M, Hagen R, Kleinsasser NH, 2013. Nitrogen dioxide is genotoxic in urban concentrations. Inhal. Toxicol. 25, 341–347. 10.3109/08958378.2013.788104 [DOI] [PubMed] [Google Scholar]
  20. Krewski D, Burnett RT, Goldberg MS, Hoover K, Siemiatycki J, Jerrett M, Abrahamowicz M, White WH, 2001. Reanalysis of the Harvard Six Cities Study and the American Cancer Society Study of Particulate Air Pollution and Mortality. [Google Scholar]
  21. Laden F, Schwartz J, Speizer FE, Dockery DW, 2006. Reduction in fine particulate air pollution and mortality: Extended follow-up of the Harvard Six Cities study. Am. J. Respir. Crit. Care Med. 173, 667–672. 10.1164/rccm.200503-443OC [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. McDonnell WF, Nishino-Ishikawa N, Petersen FF, Chen LH, Abbey DE, 2000. Relationships of mortality with the fine and coarse fractions of long-term ambient PM10 concentrations in nonsmokers. J. Expo. Anal. Environ. Epidemiol. 10, 427–436. 10.1038/sj.jea.7500095 [DOI] [PubMed] [Google Scholar]
  23. Medgyesi DN, Fisher JA, Flory AR, Hayes RB, Thurston GD, Liao LM, Ward MH, Silverman DT, Jones RR, 2021. Evaluation of a commercial database to estimate residence histories in the Los Angeles Ultrafines Study. Environ. Res. 197, 110986. 10.1016/j.envres.2021.110986 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Meza R, Cao P, Jeon J, Warner KE, Levy DT, 2023. Trends in US Adult Smoking Prevalence, 2011 to 2022. JAMA Health Forum 4, e234213. 10.1001/jamahealthforum.2023.4213 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Michaud D, Midthune D, Hermansen S, Leitzmann MF, Harlan LC, Kipnis V, Schatzkin A, 2005. Comparison of cancer registry case ascertainment with SEER estimates and self-reporting in a subst of the NIH-AARP Diet and Health Study. J Regist. Manag 32, 70–75. [Google Scholar]
  26. Moon DH, Kwon SO, Kim S-Y, Kim WJ, 2020. Air Pollution and Incidence of Lung Cancer by Histological Type in Korean Adults: A Korean National Health Insurance Service Health Examinee Cohort Study. Int. J. Environ. Res. Public. Health 17, 915. 10.3390/ijerph17030915 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. National Cancer Institute, 2025. Common Cancer Sites - Cancer Stat Facts [WWW Document]. SEER. URL https://seer.cancer.gov/statfacts/html/common.html (accessed 5.14.25). [Google Scholar]
  28. Pesch B, Kendzia B, Gustavsson P, Jöckel K-H, Johnen G, Pohlabeln H, Olsson A, Ahrens W, Gross IM, Brüske I, Wichmann H-E, Merletti F, Richiardi L, Simonato L, Fortes C, Siemiatycki J, Parent M-E, Consonni D, Landi MT, Caporaso N, Zaridze D, Cassidy A, Szeszenia-Dabrowska N, Rudnai P, Lissowska J, Stücker I, Fabianova E, Dumitru RS, Bencko V, Foretova L, Janout V, Rudin CM, Brennan P, Boffetta P, Straif K, Brüning T, 2012. Cigarette smoking and lung cancer—relative risk estimates for the major histological types from a pooled analysis of case–control studies. Int. J. Cancer 131, 1210–1219. 10.1002/ijc.27339 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pope CA, Burnett RT, Turner MC, Cohen A, Krewski D, Jerrett M, Gapstur SM, Thun MJ, 2011. Lung Cancer and Cardiovascular Disease Mortality Associated with Ambient Air Pollution and Cigarette Smoke: Shape of the Exposure–Response Relationships. Environ. Health Perspect. 119, 1616–1621. 10.1289/ehp.1103639 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Pope CA, Lefler JS, Ezzati M, Higbee JD, Marshall JD, Kim S-Y, Bechle M, Gilliat KS, Vernon SE, Robinson AL, Burnett RT, 2019. Mortality Risk and Fine Particulate Air Pollution in a Large, Representative Cohort of U.S. Adults. Environ. Health Perspect. 127, 77007. 10.1289/EHP4438 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Puett RC, Hart JE, Yanosky JD, Spiegelman D, Wang M, Fisher JA, Hong B, Laden F, 2014. Particulate matter air pollution exposure, distance to road, and incident lung cancer in the nurses’ health study cohort. Environ. Health Perspect. 122, 926–932. 10.1289/ehp.1307490 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Raaschou-Nielsen O, Andersen ZJ, Beelen R, Samoli E, Stafoggia M, Weinmayr G, Hoffmann B, Fischer P, Nieuwenhuijsen MJ, Brunekreef B, Xun WW, Katsouyanni K, Dimakopoulou K, Sommar J, Forsberg B, Modig L, Oudin A, Oftedal B, Schwarze PE, Nafstad P, De Faire U, Pedersen NL, Ostenson C-G, Fratiglioni L, Penell J, Korek M, Pershagen G, Eriksen KT, Sørensen M, Tjønneland A, Ellermann T, Eeftens M, Peeters PH, Meliefste K, Wang M, Bueno-de-Mesquita B, Key TJ, de Hoogh K, Concin H, Nagel G, Vilier A, Grioni S, Krogh V, Tsai M-Y, Ricceri F, Sacerdote C, Galassi C, Migliore E, Ranzi A, Cesaroni G, Badaloni C, Forastiere F, Tamayo I, Amiano P, Dorronsoro M, Trichopoulou A, Bamia C, Vineis P, Hoek G, 2013. Air pollution and lung cancer incidence in 17 European cohorts: prospective analyses from the European Study of Cohorts for Air Pollution Effects (ESCAPE). Lancet Oncol. 14, 813–822. 10.1016/S1470-2045(13)70279-1 [DOI] [PubMed] [Google Scholar]
  33. Schatzkin A, Subar AF, Thompson FE, Harlan LC, Tangrea J, Hollenbeck AR, Hurwitz PE, Coyle L, Schussler N, Michaud DS, Freedman LS, Brown CC, Midthune D, Kipnis V, 2001. Design and Serendipity in Establishing a Large Cohort with Wide Dietary Intake Distributions : The National Institutes of Health–American Association of Retired Persons Diet and Health Study. Am. J. Epidemiol. 154, 1119–1125. 10.1093/aje/154.12.1119 [DOI] [PubMed] [Google Scholar]
  34. Sigsgaard T, Hoffmann B, 2024. Assessing the health burden from air pollution. Science 384, 33–34. 10.1126/science.abo3801 [DOI] [PubMed] [Google Scholar]
  35. Spitz MR, Wu X, WILKINSON A, WEI Q, 2006. Cancer of the Lung, in: Schottenfeld D, Fraumeni JF (Eds.), Cancer Epidemiology and Prevention. Oxford University Press, p. 0. 10.1093/acprof:oso/9780195149616.003.0033 [DOI] [Google Scholar]
  36. Thun MJ, Henley SJ, Travis WD, 2017. Lung Cancer, in: Thun M, Linet MS, Cerhan JR, Haiman CA, Schottenfeld D (Eds.), Cancer Epidemiology and Prevention. Oxford University Press, p. 0. 10.1093/oso/9780190238667.003.0028 [DOI] [Google Scholar]
  37. Thurston G, Ito K, Lall R, 2013. NPACT Study 4. Mortality and Long-Term Exposure to PM2.5 and Its Components in the American Cancer Society’s Cancer Prevention Study II Cohort. In: National Particle Component Toxicity (NPACT) Initiative: Integrated Epidemiologic and Toxicologic Studies of the Health Effects of Particulate Matter Components. (No. 332). Health Effects Institute, Boston, MA. [Google Scholar]
  38. Valavanidis A, Fiotakis K, Vlachogianni T, 2008. Airborne particulate matter and human health: toxicological assessment and importance of size and composition of particles for oxidative damage and carcinogenic mechanisms. J. Environ. Sci. Health Part C Environ. Carcinog. Ecotoxicol. Rev. 26, 339–362. 10.1080/10590500802494538 [DOI] [Google Scholar]
  39. Victorin K, 1994. Review of the genotoxicity of nitrogen oxides. Mutat. Res. Genet. Toxicol. 317, 43–55. 10.1016/0165-1110(94)90011-6 [DOI] [Google Scholar]
  40. Wang X, Wang T, Hua J, Cai M, Qian Z, Wang C, Li H, McMillin SE, Aaron HE, Xie C, Lin H, 2023. Histological types of lung cancer attributable to fine particulate, smoking, and genetic susceptibility. Sci. Total Environ. 858, 159890. 10.1016/j.scitotenv.2022.159890 [DOI] [PubMed] [Google Scholar]
  41. Wong CM, Tsang H, Lai HK, Thomas GN, Lam KB, Chan KP, Zheng Q, Ayres JG, Lee SY, Lam TH, Thach TQ, 2016. Cancer Mortality Risks from Long-term Exposure to Ambient Fine Particle. Cancer Epidemiol. Biomark. Prev. Publ. Am. Assoc. Cancer Res. Cosponsored Am. Soc. Prev. Oncol. 25, 839–845. 10.1158/1055-9965.EPI-15-0626 [DOI] [Google Scholar]

Associated Data

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

Supplementary Materials

1
NIHMS2112463-supplement-1.docx (188.2KB, docx)

RESOURCES