Exploratory profiles of orthorexia nervosa among medical students: A clinical continuum

Orthorexia Nervosa (ON) is a heterogeneous condition, yet traditional variable-centered approaches (e.g., correlation and regression analyses) often fail to distinguish between healthy dietary interests and maladaptive preoccupation. This study employed a data-driven, person-centered (cluster analysis) approach to identify distinct exploratory profiles of ON among medical students and to validate these profiles against independent personal psychological resources.

This cross-sectional study included 575 medical students. Participants completed the Orthorexia Nervosa Inventory (ONI), the Depression, Anxiety and Stress Scales-21 (DASS-21), the Rosenberg Self-Esteem Scale, and the Self-Confidence Scale. K-Means clustering, validated against Gaussian Mixture Modeling (GMM) was utilized to uncover latent profiles based on ON symptoms and psychological distress. Cluster separation was evaluated using the Silhouette coefficient, and stability was assessed via bootstrap resampling.

Model fit indices and theoretical interpretability supported a three-cluster solution: Protective (42.3%), High Risk (21.7%), and Compensated Profile (36.0%). Although the K-Means solution demonstrated a modest Silhouette coefficient (0.2338) reflecting a continuous spectrum, bootstrap analysis confirmed stable exploratory configurations. Despite displaying low psychological distress, the Compensated Profile exhibited significantly lower self-esteem and self-confidence compared to the Protective Profile, statistically indistinguishable from the High Risk Profile.

These exploratory profiles suggest ON represents individual differences along a clinical continuum. The observation that high orthorexic tendencies can co-occur with low self-esteem in seemingly stable individuals reveals a specific symptom configuration. Future longitudinal studies should investigate whether this pattern acts as a compensatory adaptation.