Rosacea Severity Linked to Metabolic Syndrome in New Cohort Study
Key Takeaways
Metabolic syndrome was identified in 21.5% of 298 patients with rosacea, with higher inflammatory markers and rosacea severity scores among affected patients.
A machine-learning model using routine laboratory indicators achieved an area under the curve (AUC) of 0.734 for identifying metabolic syndrome, with C-reactive protein (CRP), total cholesterol, alanine aminotransferase (ALT), and hemoglobin A1c (HbA1c) emerging as leading predictors of rosacea.
Changes in fasting glucose correlated with changes in rosacea severity over time.
New research shows associations between rosacea severity, systemic inflammation, and metabolic syndrome.
Published in Clinical, Cosmetic and Investigational Dermatology, the study included 298 patients with rosacea treated at a tertiary dermatology clinic in China between 2021 and 2026. Researchers analyzed 5,420 laboratory records encompassing repeated measurements across nine metabolic indicators.
Metabolic Syndrome and Rosacea Severity
The authors identified metabolic syndrome in 64 patients (21.5%). Abnormal fasting glucose appeared in 23.7% of laboratory records; abnormal triglycerides occurred in 31.4% of patients. Those with metabolic syndrome had significantly higher CRP concentrations than those without the condition (6.8 ± 3.1 vs 2.4 ± 1.6 mg/L; P < 0.001). Rosacea severity scores were also higher (14.2 ± 3.5 vs 10.1 ± 3.0; P = 0.003).
Investigators developed a random forest classifier using seven metabolic indicators, excluding fasting glucose and triglycerides from the predictors. The model was trained on 164 patients and evaluated in a time-stratified holdout cohort of 134 patients.
The model achieved an AUC of 0.734 (95% CI, 0.628 to 0.824). CRP, total cholesterol, ALT, and HbA1c were the leading contributors to prediction.
Among 83 patients with repeated follow-up visits, changes in fasting glucose correlated positively with changes in rosacea severity (ρ = 0.31; P = 0.004).
"Outpatients with rosacea frequently show systemic metabolic disturbances aligned with metabolic syndrome status and clinical severity," the authors wrote. "A machine-learning model built from routinely available, non-diagnostic indicators retained discrimination on a time-stratified holdout, suggesting a low-cost approach to flagging metabolic risk in this population, which requires external validation in independent, multicenter cohorts before clinical use."
Source
Wu M, et al. Clin Cosmet Investig Dermatol. 2026. Doi:10.2147/CCID.S634954