Introduction and objectives We studied the changes in lung function and cardiovascular responses in healthy volunteers and patients with COPD exposed to the high pollution levels in a busy London street.

Methods Using a cross-over design, 37 healthy volunteers and 37 COPD patients (walked along Oxford Street (diesel only traffic) and, on a separate occasion, in Hyde Park (low or little traffic), London for two hours. Cardio-respiratory measurements were performed at baseline, and during and after each exposure, alongside personal particulate and gaseous exposure measurements.

Findings Compared to Hyde Park, mean exposures on Oxford Street had higher levels of black carbon (10.4 μm/m³ vs. 1.2 μm/m³, p < 0.001) and ultrafine particle counts (25472/cm³ vs 5709/cm³, p < 0.001).

In comparison with Hyde Park the healthy subjects had a mean fall in FEV₁ from baseline of6.05% (p = 0.01) 6 h and a fall of 4.17% (p = 0.01) 24 h after arrival in Oxford St. There was no associated drop in FVC. Arterial stiffness measured by pulse wave velocity (PWV) increased 24 h after arriving on Oxford Street.

In volunteers with COPD, there was a mean fall in FEV₁ of 4% (p = 0.01) with an associated drop in FVC of 3.4% (p = 0.02) one hour after the start of exposure on Oxford Street, compared to Hyde Park. Measurement of impulse oscillometry in volunteers with COPD demonstrated increased airway resistance at 5 Hz of 0.05 kPa/l/s (p = 0.01) four hours and at 20 Hz of 0.02 (p = 0.04) 24 h after exposure began on Oxford Street. PWV increased by 0.8 m/s and 0.5 m/s three hours and six hours after exposure started on Oxford street respectively.

There were no changes in FeNO in either group between the two sites.

Preliminary multivariate analysis has so far found no associations with individual particulate measurements.

Conclusions These findings show that airways obstruction occurred in both the healthy volunteers and COPD patients exposed to ambient levels of diesel pollution on a busy London Street. The associated vascular dysfunction was more prominent in COPD patients. Further analyses of markers of inflammation in the collected samples are now needed to ascertain the mechanistic cause of the pathophysiological findings.