Deriving Subtypes From the Insomnia Severity Index: A Latent Class Analysis and Comparison of Features

Insomnia and its association with mental health problems are prevalent in young populations. While person-centered statistical methods have identified insomnia phenotypes using a wide range of variables, the potential of common screening tools like the Insomnia Severity Index (ISI) for subtyping insomnia is underexplored. This study aimed to investigate insomnia subtypes in university students using ISI items.

In a cross-sectional online survey, 1,234 Italian university students (mean age: 23.3 ± 2.4 years) completed the ISI, the Pittsburgh Sleep Quality Index, the Depression Anxiety Stress Scale, and the Short Form-12 health survey. Latent class analysis (LCA) was performed using ISI items as indicators, and class differences in sleep quality, psychological distress, and health-related quality of life were assessed.

A four-class solution was identified: "no insomnia" (NI; 31.4%) with no significant sleep complaints; "high insomnia risk" (HI; 17.7%) showing severe nighttime and daytime symptoms; "subthreshold insomnia" (SI; 37.0%) characterized by moderate nighttime symptoms and sleep dissatisfaction; and "predominant daytime symptoms" (DS; 13.9%) featuring pronounced daytime dysfunction without major nighttime issues. The HI group exhibited the worst sleep quality and highest psychological distress. NI had the best overall outcomes, with SI and DS in intermediate positions. DS had worse mental well-being, higher daytime dysfunction, and more psychological distress compared to SI.

LCA identified four insomnia subtypes based on ISI scores, delineating a continuum from no insomnia to high risk, with one subtype marked primarily by daytime impairments. These findings could guide tailored interventions for different clinical presentations.