Assessment of major comorbidities in adults with atopic dermatitis using the Charlson comorbidity index

Background There is a growing interest in comorbidities of adults with atopic dermatitis (AD). Objectives To examine the burden of comorbidities in adult patients with AD using the Charlson comorbidity index (CCI) in nationwide registries. Methods All Danish patients ≥18 years on January 1, 2012 with AD diagnosed by a hospital dermatologist were included. Patients were age-and sex-matched in a 1:4 ratio with general population controls. Severity was determined by systemic AD treatment and analyzed by conditional logistic regression. Results In total, 10,738 adult patients with AD and 42,952 controls were analyzed. CCI score was significantly increased in smokers with AD compared with controls (0.41 vs 0.13, P < .001). Nonsmokers with AD had a similar CCI score as controls (0.09 vs 0.08, P = .12). In analyses restricted to patients with severe AD, a stronger difference in CCI score was observed for smokers (0.48 vs 0.14, P < .001) than for nonsmokers (0.10 vs 0.08, P = .01). Limitations Observational studies do not establish cause and effect. Conclusion On the basis of nationwide data, the risk for major comorbidities was significantly increased in adult patients with AD compared with controls. The risk difference was predominantly found in patients with severe disease and among smokers. There is a growing interest in comorbidities of adults with atopic dermatitis (AD). To examine the burden of comorbidities in adult patients with AD using the Charlson comorbidity index (CCI) in nationwide registries. All Danish patients ≥18 years on January 1, 2012 with AD diagnosed by a hospital dermatologist were included. Patients were age-and sex-matched in a 1:4 ratio with general population controls. Severity was determined by systemic AD treatment and analyzed by conditional logistic regression. In total, 10,738 adult patients with AD and 42,952 controls were analyzed. CCI score was significantly increased in smokers with AD compared with controls (0.41 vs 0.13, P < .001). Nonsmokers with AD had a similar CCI score as controls (0.09 vs 0.08, P = .12). In analyses restricted to patients with severe AD, a stronger difference in CCI score was observed for smokers (0.48 vs 0.14, P < .001) than for nonsmokers (0.10 vs 0.08, P = .01). Observational studies do not establish cause and effect. On the basis of nationwide data, the risk for major comorbidities was significantly increased in adult patients with AD compared with controls. The risk difference was predominantly found in patients with severe disease and among smokers.