Helicobacter pylori infection but not small intestinal bacterial overgrowth may play a pathogenic role in rosacea.

ABSTRACT

BACKGROUND AND AIMS: Recent studies suggest a potential relationship between rosacea and Helicobacter pylori (H. pylori) infection or small intestinal bacterial overgrowth (SIBO), but there is no firm evidence of an association between rosacea and H. pylori infection or SIBO. We performed a prospective study to assess the prevalence of H. pylori infection and/or SIBO in patients with rosacea and evaluated the effect of H. pylori or SIBO eradication on rosacea. METHODS: We enrolled 90 patients with rosacea from January 2012 to January 2013 and a control group consisting of 90 patients referred to us because of mapping of nevi during the same period. We used the (13)C Urea Breath Test and H. pylori stool antigen (HpSA) test to assess H. pylori infection and the glucose breath test to assess SIBO. Patients infected by H. pylori were treated with clarithromycin-containing sequential therapy. Patients positive for SIBO were treated with rifaximin. RESULTS: We found that 44/90 (48.9%) patients with rosacea and 24/90 (26.7%) control subjects were infected with H. pylori (p = 0.003). Moreover, 9/90 (10%) patients with rosacea and 7/90 (7.8%) subjects in the control group had SIBO (p = 0.6). Within 10 weeks from the end of antibiotic therapy, the skin lesions of rosacea disappeared or decreased markedly in 35/36 (97.2%) patients after eradication of H. pylori and in 3/8 (37.5%) patients who did not eradicate the infection (p < 0.0001). Rosacea skin lesions decreased markedly in 6/7 (85.7%) after eradication of SIBO whereas of the two patients who did not eradicate SIBO, one (50%) showed an improvement in rosacea (p = 0.284). CONCLUSIONS: Prevalence of H. pylori infection was significantly higher in patients with rosacea than control group, whereas SIBO prevalence was comparable between the two groups. Eradication of H. pylori infection led to a significant improvement of skin symptoms in rosacea patients.