Multidimensional Loneliness Among University Students: A Latent Profile Approach

TL;DR

This study identifies different types of loneliness among university students and how they relate to mental health and personality traits.

Contribution

The study introduces a new method to classify loneliness profiles and links them to mental health and personality factors.

Findings

Four distinct loneliness profiles were identified, with 'Social and emotional lonely' showing the highest levels of depression, anxiety, and stress.

Neuroticism was the strongest personality trait predicting loneliness profile membership.

Being in a relationship reduced the likelihood of being in the 'Severe romantic lonely' profile.

Abstract

Background: An increasing number of university students report feeling lonely, a negative experience arising from a mismatch between perceived and actual social relationships. Loneliness has been linked to poorer mental health. However, the relationship between qualitative (sources of loneliness) and quantitative (high or low) differences in loneliness and mental health is under researched. The aims of this research were to (a) identify profiles of loneliness among university students across three indicators of loneliness, namely, social, family, and romantic indicators, using latent profile analysis (LPA); (b) examine the differences among identified profiles based on dimensions of mental health indicators (depression, anxiety, and stress), social support, and life satisfaction; and (c) assess profile membership based on demographic variables (gender, social isolation, relationship status, and education characteristics) and the Big Five personality traits (extraversion, agreeableness, openness, conscientiousness, and neuroticism). Method: A cross-sectional survey was conducted on 912 university students from five cities in Uttar Pradesh, India. Participants completed questionnaires covering demographic details and validated measures assessing loneliness, depression, stress, anxiety, social support, life satisfaction, and the Big Five personality traits. Data were analyzed using the latent profile module in Jamovi and fit indices, namely, BIC, AIC, and BLRT, and entropy was used to select the best profile. Results: The latent profile analysis identified four profiles for university student loneliness, including Social and emotional lonely (31.4%), Moderate romantic lonely (23.8%), Moderate social lonely (8.2%), and Severe romantic lonely (36.6%). Moreover, the Social and emotional lonely profile scored the highest on depression, anxiety, and stress. The Moderate romantic lonely profile scored the highest on life satisfaction and social support. Being in a relationship decreased the likelihood of being categorized as Severe romantic lonely. In terms of personality, neuroticism was the strongest predictor of profile membership. This study is a step towards identifying at-risk lonely individuals with varying sources of loneliness. Identifying different profiles of lonely individuals will have direct implications for designing interventions that cater to a particular group rather than a one-size-fits-all approach.