| First author [ref.]/location | Year of study/ population/age of participants | Study design/ follow-up | Markers of exposure to air pollution | Definition of COPD | Handling of asthma | Effect estimate (95% CI) | Study limitations |
| Andersen [14]/Copenhagen and Aarhus, Denmark | 1993–2006/1786 patients admitted to hospital for COPD from cohort of 52 799 participants/50–64 years | Longitudinal study: Danish Diet, Cancer and Health cohort/13–17 years | Pollutants: modelled home NO2, NOx (dispersion model) Traffic proximity: presence of a major road within 50 m of residential address at baseline Traffic load: total number of kilometres travelled within 200 m of address at baseline | First hospital admission due to COPD (discharge diagnoses) between baseline (1993–1997) and June 27, 2006 | Data on participants admitted to hospital for asthma and other diseases were collected | HR for NO2 35-year mean level 1.08 (1.02–1.14) per 5.8 μg·m−3 HR per IQR increase for NO2 25-year mean level 1.07 (1.01–1.13) per 6.4 μg·m−3 HR per IQR increase for NO2 15-year mean level 1.05 (1.00–1.11) per 6.3 μg·m−3 HR per IQR increase for NOx 35-year mean level 1.05 (1.01–1.10) per 12.4 μg·m−3 HR per IQR increase for NOx 25-year mean level 1.04 (0.99–1.09) per 12.6 μg·m−3 HR per IQR increase for NOx 15-year mean level 1.03 (0.97–1.09) per 11.6 μg·m−3 HR per IQR increase for major road 1.04 (0.89–1.21) HR for traffic load 1.01 (0.97–1.05) per 5.8×103 vehicle km per day | Population was defined by hospital admission using discharge diagnosis of COPD, no lung function measurement |
| Karakatsani [20]/Athens, Greece | 1990–1996/Case series 1: 168 cases and 168 matched controls from Athens Case series 2: 84 cases, a subset of case series 1 that met criteria for clinical diagnosis of chronic bronchitis, emphysema or COPD/≥34 years | Nested case–control from EPIC Greece/Case series 2: 5 years | Pollutants: inverse distance-weighted mean from 3 nearest fixed-site monitors used to estimate NO2 concentrations for boroughs; past 5 and 20 years Traffic proximity: none | Case-series 1: reported history of COPD, chronic bronchitis, emphysema or respiratory symptoms Case-series 2: subset of above, with COPD defined based on clinical assessment (n=84), including 31 cases with FEV1/VC <88% (males) and 89% (females) | Asthmatics or subjects with wheezing in childhood and adulthood were excluded | OR per quartile of NO2, recent 5-years exposure: case series 1 (all), 1.18 (0.94–1.49); case series 2, 1.37 (1.05–1.79) OR per quartile of NO2, recent 20 years exposure: case series 1 (all), 1.10 (0.84–1.43); case series 2, 1.31 (0.95–1.79) OR for NO2, recent 5-year exposure for persons exposed to the highest quartile versus all others: case series 1 (all), 1.46 (0.82–2.59); case series 2, 2.01 (1.05–3.86) ORs for NO2, recent 20-year exposure for persons exposed to the highest quartile versus all others: case series 1 (all), 1.39 (0.73–2.67); case series 2, 1.46 (0.67–3.19) Data on the size of the quartiles are not available | No local traffic-related pollution information Only a subset of cases (n=31) were based on lung function alone; not GOLD cut-off. ERS criteria used to define COPD |
| Nuvolone [13]/Pisa-Cascina area, Italy | 1991–1993/2062 subjects from the general population living in the Pisa-Cascina/8–97 years (mean 45.9 and 48.9 years for males and females, respectively) | Cross sectional/Not indicated | Pollutants: none Traffic proximity: residential distance to a specified main road; exposure groups defined as highly exposed (living <100 m from the main road), moderately exposed (100–250 m) and unexposed (250–800 m) | GOLD: FEV1/FVC <0.7 using pre-bronchodilator lung function measurements | Not specified whether asthmatics were excluded from the analyses, though probably excluded as “asthma” was a separate outcome used in this analysis | For males living <100 m of main road, COPD diagnosis OR 1.80 (1.03–3.08); 100–250 m, OR 1.21 (0.69–2.13) For females living <100 m of main road, COPD diagnosis OR 1.60 (0.71–3.59); 100–250 m OR 0.99 (0.39–2.51) For males living <100 m of main road, reduced FEV1/FVC <70% OR 2.07 (1.11–3.87); 100–250 m, OR 2.53 (1.42–4.53) For females living <100 m of main road, reduced FEV1/FVC <70% OR 1.01 (0.48–2.14); 100–250 m OR 0.88 (0.41–1.89) | Only distance to the nearest road, no air pollutants, COPD defined using questionnaire-reported diagnosis of chronic bronchitis or emphysema Cross-sectional analyses |
| Pujades-Rodriguez [11]/Nottingham, UK | 1991/2599 subjects from the general population of the Gedling area, Nottingham/18–70 years | Cross sectional analyses/9 years | Pollutants: modelled NO2 at home (dispersion model) Traffic proximity: distance of residence to the nearest road (150 m cut-off and 3 distance bands of 50 m from 0–150 m) | GOLD: FEV1/FVC <0.7 using pre-bronchodilator lung function measurements | Not specified whether asthmatics were excluded from the analyses, though probably excluded as “asthma” was a separate outcome used in this analysis | OR for ≤150 m distance (>150 m as ref.): 0.97 (0.68–1.37) OR for distance bands (100–150 m as ref.): <50 m 1.54 (0.69–3.45); for 50–100 m 1.67 (0.79–3.49) OR for quintiles of modelled NO2 (<33.92 μg·m−3 as ref.): for 33.92 – 34.23 μg·m−3, 1.09 (0.68–1.73); for 34.23–34.73 μg·m−3, 0.95 (0.60–1.52); for 34.73–36.79 μg·m−3, 0.91 (0.57–1.45); for >36.79 μg·m−3, 1.07 (0.68–1.68) | Modelled residential NO2 had extremely small contrasts with the 5 categories ranging from <33.92 μg·m−3 (lowest) to >36.79 μg·m−3 (highest) Unclear to what extent distance reflected true contrasts in exposure Cross-sectional analysis |
| Pujades-Rodriguez [21]/England, UK | Assessments in 1995, 1996 and 2001/48 145 adults from the general population/>16 years | Cross-sectional surveys, Health urvey for England/5 years | Pollutants: none Traffic proximity: residential proximity to the nearest main road | GOLD: FEV1/FVC <0.7 using pre-bronchodilator lung function measurements | Not specified whether asthmatics were excluded from the analyses, though probably excluded as “asthma” was a separate outcome used in this analysis | OR ≤150 m of a main road (>150 m as ref.): 0.96 (0.89–1.03) OR for bands of distance (120–150 m as ref.): <30 m, 0.94 (0.77–1.13); 30–60 m, 0.97 (0.80–1.18); 60–90 m, 0.94 (0.78–1.14); 90–120 m, 0.97 (0.81–1.17) | Using postcode to identify residential address No pollutant estimates |
| Schikowski [22]/Ruhr area, Germany | 1985–1994/4262 females of general population of Ruhr/54–55 years at baseline | Consecutive cross-sectional studies (SALIA)/5 years | Pollutants: background concentrations of PM10 and NO2, measured within up to 8 km of each participant’s residence Traffic proximity: distance of residence to nearest major road | GOLD: FEV1/FVC <0.7 using pre-bronchodilator lung function measurements | Asthmatics with physician diagnosis or those using asthma medication were excluded from the analyses. | OR for <100 m of road: 1.79 (1.06–3.02) OR for IQR increase of 16 μg·m−3 in annual mean NO2: 1.39 (1.20–1.63) OR for IQR increase of 16 μg·m−3 in 5-year mean NO2: 1.43 (1.23–1.66) OR for IQR increase of 7 μg·m−3 in annual mean PM10: 1.37 (0.98–1.92) OR for IQR increase of 7 μg·m−3 in 5-year mean PM10: 1.33 (1.03–1.72) | Only females Narrow age group Cross-sectional analyses |
| Schikowski [23]/Ruhr area, Germany | Baseline: 1985–1994/4874 females from the general population of Ruhr/54–55 years at baseline | Cohort study (SALIA)/12–20 years | Pollutants: background concentrations of PM10, measured ≤8 km of each participant’s residence Traffic proximity: distance of residence to nearest major road (≤100 m with >10 cars per day) | GOLD: FEV1/FVC <0.7 using pre-bronchodilator lung function measurements | Asthmatics with physician diagnosis or those using asthma medication were excluded from the analyses. | OR for <100 m of road: 1.69 (0.90–3.18) OR for IQR increase of 7 μg·m−3 in 5-year mean PM10: 1.25 (0.79–1.99) | Only females Small study sample for follow-up |
| Schikowski [12]/Ruhr area, Germany | Baseline: 1985–1994 Follow-up: 2006–2009/Baseline: 4874 females of the general population living in Ruhr Follow-up: 395 females with spirometry at follow-up of the general population/Mean at baseline: 54 years Mean at follow-up: >70 years | Cohort study (SALIA)/12–20 years | Pollutants: background concentrations of PM10 and NO2, measured within up to 8 km of each participant’s residence Traffic proximity: distance of residence to nearest major road | GOLD: FEV1/FVC <0.7 using pre-bronchodilator lung function measurements | Asthmatics with physician diagnosis or those using asthma medication were excluded from the analyses. | Parameter estimate for follow-up time (unit 10 years) for mild COPD and PM10 exposure by 20 μg·m−3: 20.61 (7.81–33.41) NO2 exposure by 10 μg·m−3: 9.12 (4.78–13.46) Moderate COPD and PM10 exposure by 20 μg·m−3: 8.02 (0.01–16.03) NO2 exposure by 10 μg·m−3: 2.73 (0.03–5.43) Parameter estimate for mild COPD and a decline in NO2 exposure by 10 μg·m−3 per 10 years: −4.64 (−8.03– −1.26) Decline in PM10 by 20 μg·m−3 per 10 years: −14.62 (−25.88– −3.36) Moderate COPD and NO2 decline: −1.66 (−3.8– −0.048) PM10 decline: −6.20 (−13.33–0.94) | Only females Small study sample for follow-up Design issues |
EPIC: European Prospective Investigation into Cancer and Nutrition; SALIA: Study on the Influence of Air Pollution on Lung, Inflammation and Aging; NOx: nitrogen oxides; PM10: particulate matter with an aerodynamic cut-off diameter ≤10 μm; FEV1: forced expiratory volume in 1 s; VC: vital capacity; GOLD: Global Initiative for Chronic Obstructive Lung Disease; FVC: forced vital capacity; HR: hazard ratio; IQR: interquartile range; ERS: European Respiratory Society.