Impact of smoking status on Sjögren's disease diagnosis and phenotype: results from a cohort study with matched population controls.

Abstract

To examine associations between smoking and Sjögren's disease (SjD) diagnosis, compared with non-Sjögren's sicca syndrome (sicca) and population controls, and to explore immunomodulatory effects of smoking on disease phenotype. Adults referred for suspected SjD to the OASIS cohort from 2014 until 2023 were classified as SjD or sicca. Smoking status, exposure (pack-years), and passive smoking history were recorded through standardised questionnaires. Associations with diagnosis, clinical and serological parameters, and patient-reported outcomes were analysed using multivariate logistic and linear regression. Smoking patterns were further compared to matched controls from the 2018 UK Annual Population Survey. 317 SjD patients and 170 sicca controls were included. Of these, 257 SjD met 2016 classification criteria, and were used for primary analyses. Smoking data were available for 85.0% of the cohort. SjD patients were more frequently never smokers than sicca (70.5% vs 53.1%, p = 0.002). In the multivariate analysis, current smoking (OR 0.37, 95%CI 0.15-0.92) was independently associated with lower odds of SjD. Never smokers exhibited higher IgG levels (p = 0.003) and more frequent biological ESSDAI activity (p = 0.038). Smoking exposure correlated with reduced IgG, rheumatoid factor levels, and seropositivity, independent of DMARD use. Compared with population controls, sicca, but not SjD, showed higher ever-smoking rates (49.6% vs 39.1%, p = 0.049). Smoking was negatively associated with SjD diagnosis compared to sicca, but not population controls. Smoking exposure linked to lower immunoglobulin and autoantibody levels, suggesting an immunomodulatory effect rather than true disease protection. Findings support a potential metabolically-influenced, non-autoimmune, sicca phenotype linked to smoking.