Clustering of Health Behaviours in University Students in Cyprus: The PrePARE for Life Study

Introduction

Transition to university life is often associated with adverse health behaviours that may persist in later life, affecting chronic disease trajectories. The PrePARE for LIFE study assessed the clustering and social patterning of health behaviours amongst University students.

Methods

Six behaviours were assessed using validated tools: adherence to Mediterranean diet (MEDAS), physical activity (IPAQ), sleep quality (PSQI), social support (MOS-SSS), alcohol use (AUDIT) and tobacco use (GATS). To operationalize social position, nine individual-level (own and parental income, parental education, subjective individual and family social ladder position, student occupational status, perceived household financial situation) and one area-level measure (postcode Socioeconomic Disadvantage Index) were considered. K-means cluster analysis was performed, inspecting the total within-cluster sum of squares across 2–10 cluster solutions. Cluster membership and social position associations were examined using chi-square tests.

Results

Among 396participants, (mean age 20 yrs, 67% female), six clusters were identified: Overall-balanced (n = 98; overall better profile), Run-Down (n = 63, poor sleep, diet and physical activity but good support), Well-rested (n = 120, good sleep but poorer diet and physical activity), Feeling unsupported (n = 79, low social support; other behaviours suboptimal), and two smaller groups: Fit-Unfit (n = 17, very high physical activity but average in terms of other behaviours), and One too many (n = 19, overall poorer profile with also high alcohol use). Of the social position variables, only subjective family social position was associated with cluster membership (p = 0.02), with higher position more common in the well-rested and overall-balanced clusters.

Conclusions

The identification of distinct health behaviour clusters highlights the need for interventions that target multiple health habits simultaneously tailored to behavioural phenotypes.

Key messages

• Multiple health behaviours cluster into distinct profiles in university students, highlighting the need for interventions targeting interconnected behavioural patterns.

• Limited social patterning of behavioural profiles warrants further investigation, with only perceived family social position emerging as a modest differentiating factor.