Background Patients of color are more likely to have systemic lupus erythematosus (SLE) and a smoking history. Prior literature notes that both smoking and race impact odds of cutaneous manifestations. Therefore, we sought to examine the impact of cumulative smoking and race on cutaneous manifestations of SLE.

Methods Our cohort study included 631 consecutive SLE patients at a single academic center. Adults with at least one ambulatory rheumatology encounter with an SLE ICD- 9 or 10 code from 2008–16 were identified. Electronic health records were manually abstracted to include patients meeting ACR 1997 or SLICC 2012 classification criteria. The primary outcomes were ACR or SLICC cutaneous criteria and SLICC Damage Index (DI) cutaneous criteria. The primary explanatory variable was smoking exposure defined as low (<5 pack-years), medium (5–10 pack-years), and high (>10 pack-years), compared to nonsmokers. Covariates included age category at diagnosis (early onset <18 years old, 18–50 years, or late onset >50 years), sex, and race. Analysis was performed using multivariate logistic regression to calculate odds ratios and 95% confidence intervals (OR, (95% CI)).

Results Among 631 SLE patients, mean age was 42, 91% female, 82% white, and 40% ever smokers. Patients with low smoking exposure were nine times more likely to develop any mucocutaneous manifestations (OR 9.0, (1.2, 67.7)), four times more likely to meet any SLICC cutaneous criteria (OR 3.7, (OR 1.3, 10.6)), and twice as likely to meet ACR cutaneous criteria (OR 2.0 (1.0, 3.8)) compared to non-smokers (table 1). Patients with medium smoking exposure were twice as likely to meet acute cutaneous SLICC criteria (OR 2.3, (1.1, 5.1)), whereas those with high smoking exposure had two-fold higher odds of discoid lupus (OR 2.1, (1.1, 4.1) data not shown). Chronic cutaneous SLICC criteria and DI cutaneous criteria showed linear pack-year trends that met significance with high smoking exposure (OR 2.2, (1.2, 4.2); OR 4.2, (0.9, 9.2)). Patients of color had increased risk for alopecia, discoid lupus, chronic cutaneous lupus, and DI skin damage. Limitations included sample size and just 18% patients of color.

Conclusions Any smoking exposure was an independent risk factor for nearly all cutaneous SLE manifestations whereas high smoking exposure and patients of color had significantly increased risk of chronic cutaneous manifestations and skin damage. Findings suggest a dose relationship between smoking exposure and cutaneous manifestations/damage, making cessation an important strategy to potentially reduce disparities and improve cutaneous outcomes in SLE.

Funding Source(s): Support: Rheumatology Research Foundation