American Journal of Respiratory and Critical Care Medicine

Rationale: There is compelling evidence that acute and chronic exposure to ambient fine particulate matter (PM2.5) air pollution increases cardiopulmonary mortality. However, the role of PM2.5 in the etiology of lung cancer is less clear, particularly at concentrations that prevail in developed countries and in never-smokers.

Objectives: This study examined the association between mean long-term ambient PM2.5 concentrations and lung cancer mortality among 188,699 lifelong never-smokers drawn from the nearly 1.2 million Cancer Prevention Study–II participants enrolled by the American Cancer Society in 1982 and followed prospectively through 2008.

Methods: Mean metropolitan statistical area PM2.5 concentrations were determined for each participant based on central monitoring data. Cox proportional hazards regression models were used to estimate multivariate adjusted hazard ratios and 95% confidence intervals for lung cancer mortality in relation to PM2.5.

Measurements and Main Results: A total of 1,100 lung cancer deaths were observed during the 26-year follow-up period. Each 10 μg/m3 increase in PM2.5 concentrations was associated with a 15–27% increase in lung cancer mortality. The association between PM2.5 and lung cancer mortality was similar in men and women and across categories of attained age and educational attainment, but was stronger in those with a normal body mass index and a history of chronic lung disease at enrollment (P < 0.05).

Conclusions: The present findings strengthen the evidence that ambient concentrations of PM2.5 measured in recent decades are associated with small but measurable increases in lung cancer mortality.

Scientific Knowledge on the Subject

There is compelling evidence that acute and chronic exposure to ambient fine particulate matter (PM2.5) air pollution increases cardiopulmonary mortality. However, the role of PM2.5 in the etiology of lung cancer is less clear.

What This Study Adds to the Field

This study examined the association between mean long-term ambient PM2.5 concentrations and lung cancer mortality in a 26-year prospective study of a large cohort of lifelong never-smokers. Each 10 μg/m3 increase in PM2.5 concentrations was associated with a 15–27% increase in lung cancer mortality. These results strengthen the evidence that ambient concentrations of PM2.5 are associated with small but measurable increases in lung cancer mortality.

Time-series and prospective studies provide compelling evidence that acute and chronic exposure to ambient fine particulate matter (PM2.5) air pollution is associated with increased cardiopulmonary mortality (1). However, the role of PM2.5 in the etiology of lung cancer is less clear, particularly at concentrations that prevail in developed countries (∼ 5–35 μg/m3) and in never-smokers (2). In China, high levels of indoor air pollution caused by coal and biomass burning contribute to high lung cancer rates observed even among nonsmoking women (3). There are also high background concentrations (>100 μg/m3) of outdoor air pollution in some industrial regions of the country (2).

Given the strong relationship between cigarette smoking and lung cancer risk, evidence of an association between PM2.5 and lung cancer is more convincing when observed among never-smokers, compared with current or former smokers, because of possible residual confounding by cigarette smoking (4, 5). A previous analysis of the American Cancer Society Cancer Prevention Study-II (CPS-II), based on 16 years of follow-up data of approximately 500,000 included participants controlling for measured parameters of active smoking, found an 8% (95% confidence interval [CI], 1–16%) increase in lung cancer mortality for each 10 μg/m3 increase in PM2.5 concentrations (6). The risk was somewhat higher, although statistically insignificant, when restricted to the subgroup of never-smokers. An extended analysis of the Harvard Six Cities Study (n = 8,096) found a positive association between PM2.5 and lung cancer mortality (hazard ratio [HR] per each 10 μg/m3 = 1.27; 95% CI, 0.96–1.69) controlling for active smoking (7). Naess and coworkers (8) observed significant positive associations between PM2.5 and lung cancer mortality among Oslo women in a recent register-based study; however, no data on smoking history were available in this study.

Despite this, the World Health Organization has estimated that long-term PM2.5 exposure is responsible for approximately 5% of all cancers of the trachea, bronchus, and lung (9). To address the potential for residual confounding by cigarette smoking status, the present study examined associations between mean long-term ambient PM2.5 concentrations and lung cancer mortality in a 26-year (1982–2008) prospective follow-up of 188,699 lifelong never-smoking CPS-II participants.

Study Population

The CPS-II is a prospective study of nearly 1.2 million participants enrolled by over 77,000 volunteers in 1982. Ethics approval for the CPS-II was obtained from the Emory University School of Medicine Human Investigations Committee. Participants were recruited in all 50 states and the District of Columbia and Puerto Rico. Participants were largely friends and family members of the volunteers. For inclusion in CPS-II, participants had to be at least 30 years of age and have at least one family member aged 45 years or older. A four-page self-administered questionnaire completed at enrollment captured data on a range of demographic, lifestyle, medical, and other factors, including ZIP code of residence.

Follow-up for vital status has been conducted every 2 years. In 1984, 1986, and 1988, vital status was obtained from the study volunteers, and confirmed by obtaining the corresponding death certificate. Since 1989, computerized linkage to the National Death Index has been used for follow-up (10). Through 2008, a total of 637,033 (53.8%) participants were alive; 544,545 (46%) had died; and 2,840 (0.2%) had follow-up terminated in September of 1988 because of insufficient information to link to the National Death Index. Over 99% of all known deaths have been assigned a cause. Lung cancer deaths were classified according to the underlying cause of death using the International Classification of Disease-9 (162) and -10 (C33, C34) coding system (11, 12).

Of the 1,184,881 CPS-II participants, we excluded current or former cigarette smokers (702,427); individuals with missing data on vital status (46); prevalent cancer (except nonmelanoma skin cancer) at enrollment (33,852); missing ZIP code (39,093) or county (9,552) data; or missing data on any individual-level covariates of interest (24,828). A total of 375,083 lifelong never-smokers were retained for the present analysis, of which 188,699 resided in a Metropolitan Statistical Area (MSA) with available PM2.5 monitoring data (see below). A total of 1,100 lung cancer deaths were observed in 4,225,436 person-years of follow-up.

Ecologic Measures of PM2.5

Study participants were assigned to a primary MSA of residence using five-digit ZIP code information provided at enrollment according to the ZIP code boundaries (STF3B) of the 1980 United States Census (13). Three different ecologic measures of PM2.5 were used as indicators of historical PM2.5 exposure. Average ambient PM2.5 concentrations for the 4-year period (1979–1983) encompassing the year of enrollment were obtained for 131,864 participants residing in 61 MSAs from the Inhalable Particle Monitoring Network, as compiled by the Health Effects Institute reanalysis team (14). Average ambient PM2.5 concentrations were also available in 1999 and in the first three quarters of 2000 for 177,752 participants residing in 117 MSAs from the Aerometric Information Retrieval System, implemented in response to the 1997 United States Environmental Protection Agency PM2.5 standard. Quarterly mean PM2.5 concentrations were determined by site and MSA and averaged when there were at least 50% of sixth-day samples and at least 45 total sampling days in one of the two corresponding quarters. Because there was no systematic monitoring of PM2.5 in the United States in the period spanning the early 1980s to the late 1990s, a third measure representing the average of PM2.5 concentrations in the two time periods (1979–1983 and 1999–2000) was also constructed for 120,917 participants in 53 MSAs. These indicators of ambient PM2.5 concentrations have been extensively examined in relation to mortality health effects in the CPS-II (6, 14, 15).

Ecologic Measures of Residential Radon

Mean county-level residential radon concentrations were obtained from the Lawrence Berkeley National Laboratory (16). Because long-term residential radon monitoring data in the United States are sparse, researchers at the Lawrence Berkeley National Laboratory used a variety of short- and long-term indoor radon monitoring data, along with a variety of geologic, soil, meteorologic, and housing data, to estimate the annual average radon concentrations in the main living areas of homes using an empirically constructed statistical model. Short-term screening data from the United States Environmental Protection Agency State Residential Radon Survey (mid- to late 1980s) were combined with geologic data, including estimated radium concentrations, and location of screening measurements within the home, along with a short- to long-term radon monitoring data conversion factor, to predict annual average radon concentrations in homes in 3,079 United States counties. We recently observed a significant positive association between mean county-level residential radon concentrations and lung cancer mortality in the CPS-II (17).

Sociodemographic Ecologic Covariates

Data on a range of social and demographic ecologic-level covariates were compiled for 18,731, 17,096, and 17,508 participant ZIP codes or zip code tabulation areas from the 1980, 1990, and 2000 United States Census, respectively (13, 18, 19). Variables included median household income, and percent air conditioning (1980 only), nonwhite, black, Hispanic, postsecondary education, unemployment, poverty, urban, moving, and homes with a well (1980 and 1990 only).

Statistical Analysis

Cox proportional hazards regression models were used to examine the independent effects of PM2.5 concentrations on lung cancer mortality in lifelong never-smokers. The proportional hazards models were stratified by 1-year age categories; sex; and race (white, black, or other). Follow-up time since enrollment (1982) was used as the time axis. The survival times of those still alive at the end of follow-up were treated as censored observations.

Estimated HRs and 95% CIs were adjusted for the following individual-level risk factors: education; marital status; body mass index (BMI); BMI squared; passive smoking (hours); quintiles of vegetable, fruit, and fiber and fat intake; occupational exposures (asbestos, chemicals-acids-solvents, coal or stone dusts, coal tar-pitch-asphalt, formaldehyde, and diesel engine exhaust); a previously developed occupational dirtiness index specifically designed for the CPS-II (14, 20); and mean county-level residential radon concentrations (17). Adjustment for prevalent chronic lung disease (CLD) (asthma, chronic bronchitis, or emphysema) or hay fever at enrollment produced virtually no change in the results.

Potential effect modification was assessed by including multiplicative interaction terms between PM2.5 concentrations and each risk factor in the proportional hazards models. Two-sided P values were calculated to assess the significance of the interaction term using the likelihood ratio statistic. To assess the impact of attained age, time-dependent variables were constructed by allowing participants to be included in the risk set at each death time only if they met the attained age criteria for the model (<70, 70–79, or ≥ 80 yr). The significance of an interaction term between PM2.5 and follow-up time was used to assess the plausibility of the proportional hazards assumption.

All analyses were conducted using SAS version 9.2 (SAS Institute Inc., Cary, NC) (21). Ethics approval was obtained from the Ottawa Hospital Research Ethics Board.

Mean (SD) PM2.5 concentrations ranged from 21.1 (4.7) μg/m3 in 1979–1983 to 14 (3) μg/m3 in 1999–2000 with an average of 17.6 (3.7) μg/m3 observed for the two time periods (Table 1). The three PM2.5 measures were strongly correlated (r = 0.72–0.96), whereas weak inverse correlations were observed between PM2.5 and radon (Table 2).

TABLE 1. DISTRIBUTION OF MEAN AMBIENT FINE PARTICULATE MATTER AIR POLLUTION AND RESIDENTIAL RADON CONCENTRATIONS, NEVER-SMOKERS, FOLLOW-UP 1982–2008, CPS-II COHORT, UNITED STATES

PM2.5/Radon ConcentrationParticipantsMSAsMean (SD)Minimum1st Quartile2nd Quartile3rd QuartileMaximum
PM2.5 (1979–1983), μg/m3131,8646121.1 (4.7)10.317.521.724.137.8
PM2.5 (1999–2000), μg/m3177,75211714 (3)5.811.814.31622.2
PM2.5 (1979–1983) and (1999–2000) average, μg/m3120,9175317.6 (3.7)914.418.220.227.7
Radon, Bq/m3375,08355.5 (39)6.327.443.374265.7

Definition of abbreviations: CPS-II = American Cancer Society Cancer Prevention Study-II; MSA = Metropolitan Statistical Area; PM2.5 = ambient fine particulate matter.

TABLE 2. CORRELATION BETWEEN MEAN AMBIENT FINE PARTICULATE MATTER AIR POLLUTION AND RESIDENTIAL RADON CONCENTRATIONS, NEVER-SMOKERS, FOLLOW-UP 1982–2008, CPS-II COHORT, UNITED STATES

PM2.5/Radon ConcentrationPM2.5(1979–1983)PM2.5(1999–2000)PM2.5 (1979–1983) and (1999–2000) AverageRadon
PM2.5 (1979–1983)0.720.96−0.22
PM2.5 (1999–2000)0.89−0.26
PM2.5 (1979–1983) and (1999–2000) average−0.31
Radon

Definition of abbreviations: CPS-II = American Cancer Society Cancer Prevention Study-II; PM2.5 = ambient fine particulate matter.

Table 3 presents the distribution of selected participant characteristics overall and in relation to PM2.5 (1999–2000) concentrations. Most participants were between 50 and 69 years of age, were female, and had some postsecondary education. There was a tendency for higher PM2.5 concentrations to be observed in participants who were nonwhite, had a lower level of educational attainment, had a higher BMI, were nonmarried, and had a lower intake of vegetables, fruit, and fiber.

TABLE 3. DISTRIBUTION (%) OF SELECTED PARTICIPANT CHARACTERISTICS AT ENROLLMENT (1982), NEVER-SMOKERS, CPS-II COHORT, UNITED STATES

Quartiles of PM2.5 (1999–2000) Concentration (μg/m3)
CharacteristicNo. (%)Q1 (5.8 to <11.8)Q2 (11.8 to <14.3)Q3 (14.3 to <16)Q4 (≥16)
Overall177,752 (100)42,397 (100)46,478 (100)42,258 (100)46,619 (100)
Age at enrollment, yr
 <409,510 (5.4)2,069 (4.9)2,512 (5.4)2,563 (6.1)2,366 (5.1)
 40–4938,438 (21.6)9,125 (21.5)10,012 (21.5)9,606 (22.7)9,695 (20.8)
 50–5961,553 (34.6)14,314 (33.8)16,148 (34.7)14,714 (34.8)16,377 (35.1)
 60–6945,028 (25.3)11,087 (26.2)11,675 (25.1)10,231 (24.2)12,035 (25.8)
 70–7918,618 (10.5)4,680 (11)4,920 (10.6)4,106 (9.7)4,912 (10.5)
 80+4,605 (2.6)1,122 (2.7)1,211 (2.6)1,038 (2.5)1,234 (2.7)
Race
 White166,222 (93.5)40,844 (96.3)43,447 (93.5)38,573 (91.3)43,358 (93)
 Black7,315 (4.1)754 (1.8)1,840 (4)2,732 (6.5)1,989 (4.3)
 Other4,215 (2.4)799 (1.9)1,191 (2.6)953 (2.3)1,272 (2.7)
Sex
 Male50,805 (28.6)12,538 (29.6)13,281 (28.6)12,156 (28.8)12,830 (27.5)
 Female126,947 (71.4)29,859 (70.4)33,197 (71.4)30,102 (71.2)33,789 (72.5)
Education
 Less than high school19,934 (11.2)4,230 (10)5,183 (11.2)4,733 (11.2)5,788 (12.4)
 High school57,345 (32.3)13,813 (32.6)15,216 (32.7)12,945 (30.6)15,371 (33)
 More than high school100,473 (56.5)24,354 (57.4)26,079 (56.1)24,580 (58.2)25,460 (54.6)
Marital status
 Single8,172 (4.6)1,761 (4.2)2,006 (4.3)2,042 (4.8)2,363 (5.1)
 Married144,111 (81.1)35,185 (83)37,762 (81.3)33,955 (80.4)37,209 (79.8)
 Other25,469 (14.3)5,451 (12.9)6,710 (14.4)6,261 (14.8)7,047 (15.1)
Body mass index, kg/m2
 <18.52,989 (1.7)681 (1.6)816 (1.8)713 (1.7)779 (1.7)
 18.5–24.993,163 (52.4)22,364 (52.8)24,591 (52.9)22,244 (52.6)23,964 (51.4)
 25–29.959,546 (33.5)14,387 (33.9)15,415 (33.2)13,954 (33)15,790 (33.9)
 30+22,054 (12.4)4,965 (11.7)5,656 (12.2)5,347 (12.7)6,086 (13.1)
Exposure to smoking, hr/d (SD)1.9 (3.1)1.7 (2.9)2 (3.1)2.1 (3.2)2 (3.2)
Vegetable, fruit, fiber consumption
 1st quintile25,002 (14.1)5,334 (12.6)6,392 (13.8)6,058 (14.3)7,218 (15.5)
 2nd quintile29,611 (16.7)6,716 (15.8)7,700 (16.6)7,153 (16.9)8,042 (17.3)
 3rd quintile32,986 (18.6)7,922 (18.7)8,646 (18.6)7,866 (18.6)8,552 (18.3)
 4th quintile36,865 (20.7)9,098 (21.5)9,647 (20.8)8,772 (20.8)9,348 (20.1)
 5th quintile38,001 (21.4)9,855 (23.2)9,995 (21.5)8,730 (20.7)9,421 (20.2)
Fat consumption
 1st quintile32,980 (18.6)6,874 (16.2)8,806 (19)8,192 (19.4)9,108 (19.5)
 2nd quintile35,075 (19.7)8,201 (19.3)9,155 (19.7)8,450 (20)9,269 (19.9)
 3rd quintile34,790 (19.6)8,644 (20.4)9,100 (19.6)8,101 (19.2)8,945 (19.2)
 4th quintile32,787 (18.5)8,377 (19.8)8,477 (18.2)7,631 (18.1)8,302 (17.8)
 5th quintile26,833 (15.1)6,829 (16.1)6,842 (14.7)6,205 (14.7)6,957 (14.9)
Unclassifiable diet15,287 (8.6)3,472 (8.2)4,098 (8.8)3,679 (8.7)4,038 (8.7)
Industrial exposures (%)26,746 (15.1)6,680 (15.8)7,037 (15.1)6,337 (15)6,692 (14.4)
Occupational Dirtiness Index
 Level 094,866 (53.4)22,241 (52.5)24,776 (53.3)22,444 (53.1)25,405 (54.5)
 Level 127,771 (15.6)6,715 (15.8)7,347 (15.8)6,793 (16.1)6,916 (14.8)
 Level 218,268 (10.3)4,581 (10.8)4,745 (10.2)4,383 (10.4)4,559 (9.8)
 Level 36,977 (3.9)1,788 (4.2)1,897 (4.1)1,606 (3.8)1,686 (3.6)
 Level 49,531 (5.4)2,498 (5.9)2,588 (5.6)2,167 (5.1)2,278 (4.9)
 Level 54,756 (2.7)1,215 (2.9)1,284 (2.8)1,065 (2.5)1,192 (2.6)
 Level 61,272 (0.7)318 (0.8)323 (0.7)258 (0.6)373 (0.8)
 Not able to ascertain14,311 (8.1)3,041 (7.2)3,518 (7.6)3,542 (8.4)4,210 (9)
Radon concentrations, Bq/m3 (mean SD)53.8 (39.9)75.4 (44.9)48.9 (37.9)43.3 (33.7)48.3 (34.4)
Asthma, %8,158 (4.6)2,049 (4.8)2,242 (4.8)1,914 (4.5)1,953 (4.2)
Hay fever, %25,621 (14.4)7,064 (16.7)6,895 (14.8)5,858 (13.9)5,804 (12.5)
Chronic obstructive pulmonary disease, %5,318 (3)1,242 (2.9)1,368 (2.9)1,273 (3)1,435 (3.1)
Region
 Northeast42,925 (24.2)9,212 (21.7)16,528 (35.6)11,802 (27.9)5,383 (11.6)
 South35,479 (20)3,195 (7.5)10,288 (22.1)14,602 (34.6)7,394 (15.9)
 Midwest54,342 (30.6)10,590 (25)9,143 (19.7)11,531 (27.3)23,078 (49.5)
 West45,006 (25.3)19,400 (45.8)10,519 (22.6)4,323 (10.2)10,764 (23.1)

Definition of abbreviations: CPS-II = American Cancer Society Cancer Prevention Study-II; PM2.5 = ambient fine particulate matter.

Adjusted HRs (95% CIs) for lung cancer mortality in relation to mean PM2.5 concentrations are presented in Table 4. In the partially adjusted model, each 10 μg/m3 increase in PM2.5 was associated with a significant 19–30% increase in the risk of lung cancer death, depending on the specific PM2.5 measure used. Results were similar although slightly attenuated in the fully adjusted model, which included an additional term for mean county-level residential radon concentrations. In the fully adjusted model, a HR of 1.15 (95% CI, 0.99–1.35) was observed for lung cancer mortality associated with each 10 μg/m3 increase in PM2.5 (1979–1983) concentrations. A significant positive association (HR per each 10 μg/m3 = 1.27; 95% CI, 1.03–1.56) was observed for PM2.5 (1999–2000). Figure 1 presents adjusted HRs (95% CIs) for lung cancer mortality according to categorical indicators of PM2.5 (1999–2000) concentrations. No association was observed between PM2.5 and mortality from nonmalignant respiratory disease overall (see Table E1 in the online supplement). There was no evidence that the proportional hazards assumption was violated (P > 0.05).

TABLE 4. ADJUSTED HR (95% CI) FOR LUNG CANCER MORTALITY IN RELATION TO EACH 10 μg/m3 INCREASE IN MEAN AMBIENT FINE PARTICULATE MATTER AIR POLLUTION CONCENTRATIONS, FOLLOW-UP 1982–2008, NEVER-SMOKERS, CPS-II COHORT, UNITED STATES

PM2.5 ConcentrationNo. of Subjects (deaths)Minimally Adjusted HR (1) (95% CI)*Partially Adjusted HR (2) (95% CI)*Fully Adjusted HR (3) (95% CI)*
PM2.5 (1979–1983)131,864 (772)1.21 (1.04–1.41)1.19 (1.02–1.38)1.15 (0.99–1.35)
PM2.5 (1999–2000)177,752 (1,042)1.31 (1.07–1.60)1.30 (1.06–1.59)1.27 (1.03–1.56)
PM2.5 (1979–1983) and (1999–2000) average120,917 (714)1.29 (1.06–1.57)1.26 (1.03–1.54)1.19 (0.97–1.47)

Definition of abbreviations: CI = confidence interval; CPS-II = American Cancer Society Cancer Prevention Study-II; HR = hazard ratio; PM2.5 = ambient fine particulate matter.

* Minimally adjusted HR (1): age, race, and sex stratified. Partially adjusted HR (2): age, race, and sex stratified and adjusted for education, marital status, body mass index, body mass index squared, passive smoking, vegetable/fruit/fiber consumption, fat consumption, industrial exposures, and occupation dirtiness index. Fully adjusted HR (3): age, race, and sex stratified and adjusted for education, marital status, body mass index, body mass index squared, passive smoking, vegetable/fruit/fiber consumption, fat consumption, industrial exposures, occupation dirtiness index, and mean county-level residential radon concentrations.

Mean PM2.5 (1999–2000) concentrations were weakly correlated with sociodemographic ecologic covariates (r ranged from −0.22 to 0.22) (see Table E2). There was little change in results observed with the inclusion of ecologic covariates from any time period in the model.

Table 5 presents adjusted HRs (95% CIs) for lung cancer mortality in relation to mean PM2.5 (1999–2000) concentrations stratified according to selected participant characteristics at enrollment. Similar results were observed in men and women and across categories of attained age and educational attainment. However, results varied across categories of BMI, with a stronger association observed in participants with a normal BMI (18.5–24.9 kg/m2) (HR per each 10 μg/m3 = 1.42; 95% CI, 1.07–1.88) compared with other BMI groups (P < 0.05). Results were also found to vary by a history of self-reported physician-diagnosed asthma, or any CLD, at enrollment with stronger associations observed in those with a positive history of asthma (HR per each 10 μg/m3 = 5.18; 95% CI, 1.96–13.71) or any CLD (HR per each 10 μg/m3 = 3.78; 95% CI, 1.69–8.43) compared with those without (P < 0.05).

TABLE 5. FULLY ADJUSTED HR (95% CI) FOR LUNG CANCER MORTALITY IN RELATION TO EACH 10 μg/m3 INCREASE IN PM2.5 (1999–2000) CONCENTRATIONS ACCORDING TO SELECTED RISK FACTORS, NEVER-SMOKERS, MULTIPLICATIVE SCALE, FOLLOW-UP 1982–2008, CPS-II COHORT, UNITED STATES

CharacteristicNo. of DeathsFully Adjusted HR (95% CI)*P Value
Age at enrollment
 <65 yr6741.30 (1.00–1.69)
 ≥65 yr3681.22 (0.87–1.73)1.00
Attained age
 <70 yr5161.37 (0.91–2.07)
 70–79 yr6401.14 (0.80–1.63)
 ≥80 yr8491.33 (0.96–1.84)0.77
Race
 White9651.26 (1.01–1.56)
 Other771.63 (0.77–3.45)0.72
Sex
 Male3341.19 (0.83–1.73)
 Female7081.30 (1.01–1.68)0.84
Education
 Less than high school1641.50 (0.86–2.62)
 High school3611.29 (0.90–1.85)
 More than high school5171.21 (0.90–1.61)0.96
Marital status
 Married8141.24 (0.98–1.57)
 Other2281.44 (0.91–2.26)0.51
Body mass index
 18.5–24.9 kg/m25361.42 (1.07–1.88)
 25–29.9 kg/m23621.28 (0.89–1.83)
 ≥30 kg/m21210.54 (0.27–1.07)0.01
Passive smoking (any)
 None4991.39 (1.03–1.87)
 Any5431.17 (0.88–1.57)0.67
Vegetable, fruit, fiber consumption
 1st Tertile3781.15 (0.76–1.74)
 2nd Tertile2881.70 (1.13–2.54)
 3rd Tertile3761.03 (0.73–1.44)0.27
Fat consumption
 1st Tertile4411.29 (0.91–1.82)
 2nd Tertile3371.31 (0.90–1.90)
 3rd Tertile2641.06 (0.69–1.62)0.59
Industrial exposures
 Yes1431.17 (0.66–2.09)
 No8991.29 (1.03–1.61)0.79
Residential radon concentrations
 <148 Bq/m31,0051.26 (1.03–1.55)
 148+ Bq/m3373.97 (1.14–13.83)0.11
Asthma
 No9951.18 (0.96–1.47)
 Yes475.18 (1.96–13.71)0.005
Hay fever
 No9121.30 (1.04–1.62)
 Yes1301.07 (0.60–1.90)0.66
Chronic obstructive pulmonary disease
 No1,0141.24 (1.01–1.54)
 Yes282.08 (0.55–7.90)0.16
Any chronic lung disease
 No9721.17 (0.94–1.45)
 Yes703.78 (1.69–8.43)0.003
Region
 Northeast2562.08 (0.98–4.43)
 South2210.96 (0.48–1.92)
 Midwest3161.18 (0.72–1.93)
 West2491.24 (0.89–1.73)0.23

Definition of abbreviations: CI = confidence interval; CPS-II = American Cancer Society Cancer Prevention Study-II; HR = hazard ratio; PM2.5 = ambient fine particulate matter.

* Fully adjusted model: age, race, and sex stratified and adjusted for education, marital status, body mass index, body mass index squared, passive smoking, vegetable/fruit/fiber consumption, fat consumption, industrial exposures, occupation dirtiness index, and mean county-level residential radon concentrations where appropriate.

This large prospective study showed positive associations between mean long-term ambient fine particulate matter air pollution concentrations and lung cancer mortality in lifelong never-smokers. Each 10 μg/m3 increase in PM2.5 concentrations was associated with a 15–27% increase in the relative risk of lung cancer death after detailed adjustment for a number of potential confounders including passive smoking, occupational exposures, and radon. The association was similar in men and women and across categories of attained age and educational attainment but was stronger in those with a normal BMI or a history of asthma or any CLD at enrollment. Findings were robust to the adjustment of a variety of sociodemographic ecologic covariates at different time points in the model.

Strengths of this study include the examination of lung cancer mortality in a large cohort of 188,699 lifelong never-smokers to eliminate potential residual confounding by cigarette smoking status; an extended 26-year follow-up time period (1982–2008) with a total of 1,100 observed lung cancer deaths; detailed prospectively collected individual-level lung cancer risk factor data; and the availability of ecologic measures of residential radon concentrations and sociodemographic characteristics to examine potential confounding by radon and community-level factors.

Although previous studies examining associations between PM2.5 and lung cancer adjusting for cigarette smoking history have generally reported positive findings (6, 7, 15), there remains concern regarding potential residual confounding by cigarette smoking status; previous studies of nonsmokers were also limited by the small numbers of lung cancer cases. Results from a prospective investigation of 3,769 participants from the Adventist Health Study of Smog, a cohort of nonsmoking California Seventh-Day Adventists followed-up from 1977–1992, reported a positive, although imprecise, association between estimated PM2.5 concentrations and lung cancer mortality in males (HR per each 24.3 μg/m3 = 2.23; 95% CI, 0.56–8.94); however only 13 lung cancer deaths were observed (22). A previous 16-year follow-up of never-smoking CPS-II participants reported a positive, although statistically insignificant, association between PM2.5 and lung cancer death (6).

Several European studies have examined associations between measures of traffic air pollution and lung cancer incidence or mortality (8, 2333). Beelen and coworkers (23) observed positive associations between measures of black smoke concentrations and traffic intensity and lung cancer incidence in 40,114 never-smoking participants in the Netherlands Cohort Study on Diet and Cancer. A total of 252 lung cancer cases were observed in the 11-year follow-up time period. Vineis and coworkers (32) reported a significant positive association between NO2 concentrations (upper vs. lowest and intermediate tertiles combined) and lung cancer incidence (odds ratio = 1.37; 95% CI, 1.06–1.75), but not PM10 or SO2, in a case-control study of 271 nonsmoking lung cancer cases nested within the European Prospective Investigation on Cancer and Nutrition. There was also some evidence for higher air pollution relative risk estimates in nonsmokers compared with current or former smokers in other recent work (30, 31, 33).

Ambient fine particulate matter comprises a diverse group of air pollutants that may be deposited and retained in the deep branches of the respiratory system, the chemical composition of which varies widely and may include a variety of adsorbed organic compounds, transition metals, ions, and minerals capable of inducing toxic biologic effects (34). Long-term exposure to fine particulate air pollution may lead to increased lung cancer risk through inflammatory injury, reactive oxygen species production, and oxidative damage to DNA (35). Genotoxic and mutagenic effects have also been demonstrated in laboratory studies (34, 36). In 1989, the International Agency for Research on Cancer identified diesel engine exhaust as a probable human carcinogen, based largely on findings from animal-based studies (37). Studies of occupational diesel exposure have also reported positive associations with lung cancer, although uncertainties with respect to exposure-response and residual confounding by cigarette smoking status remain (38).

Although potential mechanisms surrounding the stronger PM2.5-lung cancer mortality association observed in those with a normal BMI are unclear, there may be other more important influences on the mortality experience of overweight and obese individuals that may compete with lung cancer, including elevated underlying cardiovascular disease risk factors (39). Stronger associations were also observed in individuals with a history of asthma or any CLD at enrollment. Although these results should be interpreted cautiously because of the small number of participants with CLD in the present study, findings may be caused by an increased susceptibility to the carcinogenic effects of fine particulate air pollution in those with underlying respiratory disease, possibly as a result of impaired clearance or defense mechanisms (35, 40), or some form of common underlying exposure that may be independently associated with both CLD and lung cancer. Impaired pulmonary function and CLDs have been associated with ambient air pollution (4, 4143). CLDs may also be independently associated with lung cancer because of local mechanisms of inflammation and repair (4446). No information was available on CLD from enrollment. Although some studies have also suggested potential modifying effects of educational attainment and fruit and vegetable consumption on air pollution–mortality associations (6, 14, 15, 23, 26, 30, 31), this was not observed in the present study.

Limitations include the assignment of PM2.5 data to study participants at a coarse geographic scale, at the level of the MSA of residence at enrollment, rather than at the individual- or household-level. Previous research in the CPS-II examining mortality health effects at the intraurban scale in Los Angeles, California, revealed relative risk estimates approximately threefold greater than those estimated using between-city contrasts (47). There was also limited historical PM2.5 monitoring data, with widespread systematic PM2.5 monitoring occurring only in the late 1990s, nearly two decades after cohort enrollment. Air pollution exposures experienced over an extended historical time period are likely more relevant to the etiology of lung cancer than air pollution exposures experienced in the more recent past (7, 29). Although PM2.5 concentrations have declined in recent decades, with an approximate in 33% decline in mean PM2.5 concentrations observed from 1979–1983 to 1999–2000 in the 53 MSAs with data available on both time periods, PM2.5 data from both historical monitoring time periods were strongly correlated (r >0.7) and the relative ranking of MSAs in terms of PM2.5 concentrations was generally retained over time. Similar findings for lung cancer mortality were also observed using either PM2.5 (1979–1983) or PM2.5 (1999–2000) among participants residing in one of the 53 MSAs common to both measures (fully adjusted HR per each 10 μg/m3 PM2.5 [1979–1983] = 1.16, 95% CI 0.99–1.35; PM2.5 [1999–2000] = 1.15, 95% CI 0.89–1.48).

The present results provide no information as to whether there may be a critical exposure time window that may be most relevant for lung cancer etiology. However, previous work in a subset of the CPS-II using estimated yearly PM2.5 (1972–2000) concentrations, derived from concentrations of PM10 and total suspended particulates, examining the relative importance of different exposure time windows for all-cause and cause-specific mortality, including lung cancer, was largely uninformative because of limitations in study design and modest spatiotemporal variation in PM2.5 concentrations over time (15).

There was no information on residential mobility after enrollment; however, never-smoking participants reported living in their current neighborhood at enrollment for a mean number (SD) of 20.7 (14.9) years. Misclassification because of residential mobility would also likely be nondifferential, biasing estimated RR estimates toward unity. No updated data on cigarette smoking or other individual-level covariates of interest were collected from enrollment in the full CPS-II; however, it is unlikely that lifelong never-smokers in the cohort with an average age at enrollment of 57 years would begin smoking during follow-up. There may also have been changes in other sociodemographic ecologic-level factors over time; however, little change in results was observed on the inclusion of sociodemographic ecologic covariates in the model from any three of the time periods considered (1980s, 1990s, or 2000s).

Although the present study was based on mortality, inferences about the incidence of highly fatal diseases, such as lung cancer, may be reasonably approximated using mortality-based data. Similar associations between ambient air pollution and lung cancer incidence and mortality were also observed in other recent work (23, 24, 28, 29). There was no information available on the histologic subtype of lung cancer. Results from a Danish study reported stronger associations between estimated NOx concentrations and incident small-cell carcinoma and squamous cell carcinoma than adenocarcinoma (30).

Finally, this study used ecologic measures of residential radon to adjust for the potential confounding effects of residential radon exposure, rather than residential radon concentrations measured in individual homes. However, in previous work, estimates of increased lung cancer mortality caused by environmental radon observed in the CPS-II were compatible with estimates obtained in combined analyses of residential case-control studies (17). Mean radon concentrations were also weakly (and inversely) correlated with PM2.5, suggesting that any potential confounding effect of residential radon concentrations on PM2.5–lung cancer associations is likely small.

In conclusion, results from this large prospective study showed positive associations between mean long-term ambient PM2.5 concentrations and lung cancer mortality in lifelong never-smokers, further strengthening the evidence that ambient concentrations of PM2.5 measured in recent decades are associated with small but measurable increases in lung cancer mortality. Results also demonstrate that the magnitude of lung cancer risk associated with exposure to PM2.5 is notably smaller than that caused by active smoking (48).

The authors thank Dr. Jeanne Calle for valuable contributions in the development of the study and Dr. Richard Burnett for helpful discussions.

1. Brook RD, Rajagopalan S, Pope CA, Brook JR, Bhatnagar A, Diez-Roux AV, Holguin F, Hong Y, Luepker RV, Mittleman MA, et al.., American Heart Association Council on Epidemiology and Prevention, Council on the Kidney in Cardiovascular Disease, and Council on Nutrition, Physical Activity and Metabolism. Particulate matter air pollution and cardiovascular disease: an update to the scientific statement from the American Heart Association. Circulation 2010;121:23312378.
2. van Donkelaar A, Martin RV, Brauer M, Kahn R, Levy R, Verduzco C, Villeneuve PJ. Global estimates of ambient fine particulate matter concentrations from satellite-based aerosol optical depth: development and application. Environ Health Perspect 2010;118:847855.
3. Zhang JJ, Smith KR. Household air pollution from coal and biomass fuels in China: measurements, health impacts, and interventions. Environ Health Perspect 2007;115:848855.
4. Pope CA, Dockery DW. Health effects of fine particulate air pollution: lines that connect. J Air Waste Manage Assoc 2006;56:709742.
5. Cohen AJ. Air pollution and lung cancer: what more do we need to know? Thorax 2003;58:10101012.
6. Pope CA, Burnett RT, Thun MJ, Calle EE, Krewski D, Ito K, Thurston GD. Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. JAMA 2002;287:11321141.
7. Laden F, Schwartz J, Speizer FE, Dockery DW. Reduction in fine particulate air pollution and mortality: extended follow-up of the Harvard Six Cities study. Am J Respir Crit Care Med 2006;173:667672.
8. Naess O, Nafstad P, Aamodt G, Claussen B, Rosland P. Relation between concentration of air pollution and cause-specific mortality: four-year exposures to nitrogen dioxide and particulate matter pollutants in 470 neighborhoods in Oslo, Norway. Am J Epidemiol 2007;165:435443.
9. Cohen AJ, Anderson HR, Ostro B, Pandey KD, Krzyzanowski M, Kuenzli N, Gutschmidt K, Pope CA, Romieu I, Samet JM, et al.. Urban air pollution. In: , Ezzati M, Lopez AD, Rodgers A, Murray CUJL, editors. Comparative quantification of health risks: global and regional burden of disease due to selected major risk factors. Vol. 2. Geneva: World Health Organization; 2004. pp. 13531433.
10. Calle EE, Terrell DD. Utility of the National Death Index for ascertainment of mortality among Cancer Prevention Study II participants. Am J Epidemiol 1993;137:235241.
11. World Health Organization. International classification of diseases: manual of the international statistical classification of diseases, injuries, and causes of death. Geneva: World Health Organization; 1977.
12. World Health Organization. ICD-10: International statistical classification of diseases and related health problems. Geneva: World Health Organization; 1992.
13. US Department of Commerce, Bureau of the Census. Census of population and housing, 1980 (United States): summary tape file 3B. ICPSR version. Washington; 2008.
14. Krewski D, Burnett RT, Goldberg MS, Hoover K, Siemiatycki J, Jerrett M, Abrahamowicz M, White WH. Reanalysis of the Harvard Six Cities study and the American Cancer Society study of particulate air pollution and mortality. Special report. Cambridge, MA: Health Effects Institute; 2000.
15. Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi Y, Turner MC, Pope CA, Thurston G, Calle EE, et al.. Extended follow-up and spatial analysis of the American Cancer Society study linking particulate air pollution and mortality. Res Rep Health Eff Inst 2009;140:5114, discussion 115–136.
16. Price PN, Nero A, Revzan K, Apte M, Gelman A, Boscardin WJ. Predicted GM and other parameters by county in the US Data provided through the E.O. Lawrence Berkeley National Laboratory web-site [accessed 20 May 2010]. Available from: http://eande.lbl.gov/IEP/high-radon/files.html.
17. Turner MC, Krewski D, Chen Y, Pope CA, Gapstur S, Thun MJ. Radon and lung cancer in the American Cancer Society Cohort. Cancer Epidemiol Biomarkers Prev 2011;20:438448.
18. US Department of Commerce, Bureau of the Census. Census of population and housing, 1990 (United States): summary tape file 3B. ICPSR version. Washington; 1993.
19. US Department of Commerce, Bureau of the Census. Census of population and housing, 2000 (United States): selected subsets from summary file 3. ICPSR version. Washington; 2004.
20. Siemiatycki J, Krewski D, Shi Y, Goldberg MS, Nadon L, Lakhani R. Controlling for potential confounding by occupational exposures. J Toxicol Environ Health A 2003;66:15911603.
21. SAS Institute, Inc. SAS, version 9.2. Cary, NC: SAS Institute; 2008.
22. McDonnell WF, Nishino-Ishikawa N, Petersen FF, Chen LH, Abbey DE. Relationships of mortality with the fine and coarse fractions of long-term ambient PM10 concentrations in nonsmokers. J Expo Anal Environ Epidemiol 2000;10:427436.
23. Beelen R, Hoek G, van den Brandt PA, Goldbohm RA, Fischer P, Schouten LJ, Armstrong B, Brunekreef B. Long-term exposure to traffic-related air pollution and lung cancer risk. Epidemiology 2008;19:702710.
24. Beelen R, Hoek G, van den Brandt PA, Goldbohm RA, Fischer P, Schouten LJ, Jerrett M, Hughes E, Armstrong B, Brunekreef B. Long-term effects of traffic-related air pollution on mortality in a Dutch cohort (NLCS-AIR Study). Environ Health Perspect 2008;116:196202.
25. Filleul L, Rondeau V, Vandentorren S, Le Moual N, Cantagrel A, Annesi-Maesano I, Charpin D, Declercq C, Neukirch F, Paris C, et al.. Twenty five year mortality and air pollution: results from the French PAARC survey. Occup Environ Med 2005;62:453460.
26. Hoek G, Brunekreef B, Goldbohm S, Fischer P, van den Brandt PA. Association between mortality and indicators of traffic-related air pollution in the Netherlands: a cohort study. Lancet 2002;360:12031209.
27. Nafstad P, Haheim LL, Oftedal B, Gram F, Holme I, Hjermann I, Leren P. Lung cancer and air pollution: a 27 year follow up of 16 209 Norwegian men. Thorax 2003;58:10711076.
28. Nafstad P, Haheim LL, Wisloff T, Gram F, Oftedal B, Holme I, Hjermann I, Leren P. Urban air pollution and mortality in a cohort of Norwegian men. Environ Health Perspect 2004;112:610615.
29. Nyberg F, Gustavsson P, Jarup L, Bellander T, Berglind N, Jakobsson R, Pershagen G. Urban air pollution and lung cancer in Stockholm. Epidemiology 2000;11:487495.
30. Raaschou-Nielsen O, Bak H, Sorensen M, Jensen SS, Ketzel M, Hvidberg M, Schnohr P, Tjonneland A, Overvad K, Loft S. Air pollution from traffic and risk for lung cancer in three Danish cohorts. Cancer Epidemiol Biomarkers Prev 2010;19:12841291.
31. Raaschou-Nielsen O, Andersen ZJ, Hvidberg M, Jensen SS, Ketzel M, Sorensen M, Loft S, Overvad K, Tjonneland A. Lung cancer incidence and long-term exposure to air pollution from traffic. Environ Health Perspect 2011;119:860865.
32. Vineis P, Hoek G, Krzyzanowski M, Vigna-Taglianti F, Veglia F, Airoldi L, Autrup H, Dunning A, Garte S, Hainaut P, et al.. Air pollution and risk of lung cancer in a prospective study in Europe. Int J Cancer 2006;119:169174.
33. Yorifuji T, Kashima S, Tsuda T, Takao S, Suzuki E, Doi H, Sugiyama M, Ishikawa-Takata K, Ohta T. Long-term exposure to traffic-related air pollution and mortality in Shizuoka, Japan. Occup Environ Med 2010;67:111117.
34. Valavanidis A, Fiotakis K, Vlachogianni T. 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 C Environ Carcinog Ecotoxical Rev 2008;26:339362.
35. Moller P, Jacobsen NR, Folkmann JK, Danielsen PH, Mikkelsen L, Hemmingsen JG, Vesterdal LK, Forchhammer L, Wallin H, Loft S. Role of oxidative damage in toxicity of particulates. Free Radic Res 2010;44:146.
36. André V, Billet S, Pottier D, Le Goff J, Pottier I, Garçon G, Shirali P, Sichel F. Mutagenicity and genotoxicity of PM2.5 issued from an urbano-industrialized area of Dunkerque (France). J Appl Toxicol 2011;31:131138.
37. International for Research on Cancer. Diesel and gasoline engine exhausts and some nitroarenes. Lyon, France: International Agency for Research on Cancer; 1989.
38. Ris CUS. EPA health assessment for diesel engine exhaust: a review. Inhal Toxicol 2007;19:229239.
39. Giles LV, Barn P, Kunzli N, Romieu I, Mittleman MA, van Eeden S, Allen R, Carlsten C, Stieb D, Noonan C, et al.. From good intentions to proven interventions: effectiveness of actions to reduce the health impacts of air pollution. Environ Health Perspect 2011;119:2936.
40. Sacks JD, Stanek LW, Luben TJ, Johns DO, Buckley BJ, Brown JS, Ross M. Particulate matter-induced health effects: who is susceptible? Environ Health Perspect 2011;119:446454.
41. Braback L, Forsberg B. Does traffic exhaust contribute to the development of asthma and allergic sensitization in children: findings from recent cohort studies. Environ Health 2009;8:17.
42. Eisner MD, Anthonisen N, Coultas D, Kuenzli N, Perez-Padilla R, Postma D, Romieu I, Silverman EK, Balmes JR, Committee on Nonsmoking COPD, Environmental and Occupational Health Assembly. An official American Thoracic Society public policy statement: novel risk factors and the global burden of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2010;182:693718.
43. Sint T, Donohue JF, Ghio AJ. Ambient air pollution particles and the acute exacerbation of chronic obstructive pulmonary disease. Inhal Toxicol 2008;20:2529.
44. Turner MC, Chen Y, Krewski D, Calle EE, Thun MJ. Chronic obstructive pulmonary disease is associated with lung cancer mortality in a prospective study of never smokers. Am J Respir Crit Care Med 2007;176:285290.
45. Turner MC, Chen Y, Krewski D, Ghadirian P. An overview of the association between allergy and cancer. Int J Cancer 2006;118:31243132.
46. Punturieri A, Szabo E, Croxton TL, Shapiro SD, Dubinett SM. Lung cancer and chronic obstructive pulmonary disease: needs and opportunities for integrated research. J Natl Cancer Inst 2009;101:554559.
47. Jerrett M, Burnett RT, Ma R, Pope CA, Krewski D, Newbold KB, Thurston G, Shi Y, Finkelstein N, Calle EE, et al.. Spatial analysis of air pollution and mortality in Los Angeles. Epidemiology 2005;16:727736.
48. Pope CA, Burnett RT, Turner MC, Cohen A, Krewski D, Jerrett M, Gapstur S, Thun MJ. 2011. Lung cancer and cardiovascular disease mortality associated with particulate matter exposure from ambient air pollution and cigarette smoke: Shape of the exposure-response relationships. Environ Health Perspect 2011;119:16161621.
Correspondence and requests for reprints should be addressed to Michelle C. Turner, M.Sc., McLaughlin Centre for Population Health Risk Assessment, Institute of Population Health, University of Ottawa, One Stewart Street, Room 313, Ottawa, ON, K1N 6N5 Canada. E-mail:

Supported by Canada Graduate Scholarship from the Canadian Institutes of Health Research (M.C.T.). D.K. is the Natural Sciences and Engineering Research Council Chair in Risk Science at the University of Ottawa.

Author Contributions: Conception and design, M.C.T., D.K., C.A.P., Y.C., S.M.G., and M.J.T.; analysis and interpretation, M.C.T., D.K., C.A.P., Y.C., S.M.G., and M.J.T.; drafting the manuscript for important intellectual content, M.C.T., D.K., C.A.P., Y.C., S.M.G., and M.J.T.

This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org

Originally Published in Press as DOI: 10.1164/rccm.201106-1011OC on October 6, 2011

Author disclosures

Related

No related items
American Journal of Respiratory and Critical Care Medicine
184
12

Click to see any corrections or updates and to confirm this is the authentic version of record