Loneliness and Social-Communication Facets of Autistic Traits in Emerging Adulthood: A Latent Profile and Network Analysis

Loneliness is common in emerging adulthood, yet it remains unclear which facets of autistic traits are most strongly associated with it. Undergraduate students aged 18–23 years ( N = 1,172) completed the Autism-Spectrum Quotient (AQ; Likert scoring) and the UCLA Loneliness Scale. We used latent profile analysis of AQ Social Skills, Communication, and Attention to Detail (six parcels) and regularized partial-correlation networks to examine conditional associations with loneliness. The LPA identified four profiles that primarily differed in social-communication difficulty, with Attention to Detail showing an inverse pattern. Loneliness was lowest in a profile marked by low social-communication difficulty and high Attention to Detail, and highest in a profile marked by high social-communication difficulty and low Attention to Detail. In the full-sample network, loneliness showed its strongest conditional associations with Social Skills (edge = .31) and Communication (edge = .22), a small association with Attention Switching (edge = .10), and no direct conditional associations with Attention to Detail or Imagination. A reduced four-node network excluding lower-reliability AQ domains reproduced this pattern. Exploratory profile-specific networks suggested similar loneliness-related edge patterns across the larger profiles, although overall connectivity differed. Among Chinese university students in emerging adulthood, loneliness was most consistently associated with social-communication facets of autistic traits. The findings are consistent with an interactional-fit account and highlight the need for longitudinal, context-sensitive research.