Cigarette smoking and risk of palindromic rheumatism: A propensity score matching analysis

Present study was conducted to investigate smoking status in palindromic rheumatism (PR) patients compared to healthy individuals as well as to assess the effect of smoking on clinical features and outcomes of PR. One hundred and forty-six patients with diagnosis of PR and 346 healthy controls were included in this study. Demographic, clinical, and laboratory characteristics and the smoking history of PR patients at the cohort entry were obtained from patients' records. Demographic and smoking history of the control group were obtained by direct interview. In order to reduce heterogeneity between the studied groups, propensity score matching (PSM) analyses was performed. Matching was achieved by considering age, gender, educational status, and marital status. After PSM, we carried out a multivariate analysis with PR as the main outcome variable, ever smoking as the main predictor variable and age, gender, educational status, and marital status as covariates. PSM resulted in 123 PR patients and 246 matched controls. Multivariate analysis did not show a significant increase in the risk of PR in ever smokers. Seventy-six patients were anti-citrullinated protein/peptide antibody positive (ACPA-positive). Multivariate logistic regression showed a significant increase in the risk of PR in ACPA-positive ever smokers. Except lower sustained remission rate in ever smokers, no significant differences were observed in clinical manifestations and outcomes of PR between ever and never smokers. In conclusion, smoking is a risk factor for ACPA-positive PR.