Modification of associations between indoor particulate matter and systemic inflammation in individuals with COPD

Rationale: Little is known about personal characteristics and systemic responses to particulate pollution in patients with COPD.

Objectives: Assess whether diabetes, obesity, statins and non-steroidal anti-inflammatory medications (NSAIDs) modify associations between indoor black carbon (BC) and fine particulate matter ≤2.5 μm in diameter (PM_2.5) on systemic inflammation and endothelial activation.

Methods: 144 individuals with COPD without current smoking and without major in-home combustion sources were recruited at Veterans Affairs Boston Healthcare System. PM_2.5 and BC were measured in each participant's home seasonally for a week (up to 4 times; 482 observations) and plasma biomarkers of systemic inflammation [C-reactive protein (CRP); interleukin-6 (IL-6)] and endothelial activation [soluble vascular adhesion molecule-1 (sVCAM-1)] measured. Linear mixed effects regression with a random intercept was used, and effect modification assessed with multiplicative interaction terms and stratum specific estimates.

Results: Median (25%ile, 75%ile) indoor BC and PM_2.5 were 0.6 (0.5,0.7) μg/m³ and 6.8 (4.8,10.4) μg/m³, respectively. Although p-values for effect modification were not statistically significant, there were positive associations (%-increase/interquartile range; 95% CI) between CRP and BC greater among non-statin (18.8%; 3.6-36.3) than statin users (11.1%; 2.1-20.9). There were also positive associations greater among non-statin users between PM_2.5 and CRP. For IL-6, associations with BC and PM_2.5 were also greater among non-statin users. Associations between CRP and BC were greater (20.3%; 4.5-38.5) in persons with diabetes than without diabetes (10.3%; 0.92-20.6) with similar effects of PM_2.5. There were no consistent associations that differed based on obesity. Effect modification was not observed for NSAID use, or with any factor considered with sVCAM-1.

Conclusions: Associations between indoor BC and PM_2.5 and CRP were greater in patients with diabetes and those not taking statins, and with IL-6 if not taking statins. These results suggest that these characteristics may modify the systemic response to indoor BC and PM_2.5 in persons with COPD.