Smoking and Subclinical ILD in RA versus the Multi-Ethnic Study of Atherosclerosis

A population-based cohort showed an association between cigarette smoking and subclinical parenchymal lung disease defined as regions of increased computed tomography (CT) lung densitometry. This technique has not been applied to the rheumatoid arthritis (RA) population where associated ILD is highly prevalent. The association between cumulative cigarette smoking and volume of areas of high attenuation (HAA: >-600 and <-250 Hounsfield Units) on full inspiratory CT was compared in 172 RA participants and 3,969 controls in a general population sample. Multivariable regression models were used to adjust for demography, anthropometrics, percent emphysema, and CT parameters. The mean cumulative cigarette smoking exposure was 25 (IQR 10-42) and 15(IQR 5-31) pack-years for the RA and non-RA cohorts, respectively. Mean HAA was 153(±57) cm3 and 129(±50) cm3 in the RA and non-RA cohorts, respectively. Each 10 cigarette pack-year increment was associated with a higher HAA by 0.03% (95% CI, 0.007-0.05%) in RA patients and by 0.008% (95% CI, 0.003-0.01%) in those without RA (interaction p = 0.001). Cigarette smoking was associated with higher lung attenuation; with a magnitude of association more pronounced in those with RA than in the general population. These data suggest that cigarette smoking may be a more potent ILD risk factor for RA patients than in the general population.