Late-onset Hidradenitis Suppurativa: a cluster analysis of the National Italian Registry IRHIS

We analyzed data of 1,100 consecutive newly diagnosed Hidradenitis Suppurativa (HS) patients from 17 Italian outpatient clinics, participating in the IRHIS registry, with the aim of assessing the existence of clusters in the distribution of HS age-at-onset, and of evaluating differences between clusters in demographics, clinical characteristics and HS phenotypes. A detailed description of the methods is reported in Supplementary Materails.

In the analysis, we relied on both k-means and Gaussian mixture model (GMM) clustering algorithms. Both methods indicated that a two-clusters model best fitted our data, with early-onset HS peaking in the late teen years (mean= 17.0) and late-onset HS peaking at around age 40 (mean=38.4), with an optimal separation threshold at 28 years. Both univariate and age and sex-adjusted multivariable logistic regression analysis were performed.