A network analysis of risk and protective factors for body image in young adult women

This study employed a network analysis approach to model the complex interplay of risk and protective factors for body image dissatisfaction in young French women, with the objective of mapping the psychological system connecting these variables and identifying the most central factors. A sample of 233 female students completed an online questionnaire assessing 11 constructs, including risk factors like perfectionism, thin-ideal internalization, appearance comparison, and weight stigma, alongside protective factors such as self-compassion, intuitive eating, and body appreciation. This study also presents the first psychometric validation of the Physical Appearance Related Teasing Scale (PARTS) in the French language. A Gaussian Graphical Model (GGM) network analysis revealed that body dissatisfaction (BSQ-8C) has the highest strength and betweenness centrality, confirming its role as the core hub in the model's architecture and underscoring the relevance of the chosen variables for this study. The network showed strong direct positive links to body dissatisfaction from weight stigma (WSSQ) and appearance comparison (PACS-5), and strong negative links from the protective factors of body appreciation (BAS-2) and intuitive eating (IES-2). Weight-related teasing (PARTS) was established as a significant secondary risk factor through its robust connection with weight stigma. Sociocultural pressures (SATAQ-3) were identified as a critical bridging node, while variables such as self-compassion, social media use, and perfectionism occupied peripheral positions. This research advocates for a targeted, multi-component approach that actively works to dismantle the pillars of weight stigma and comparison while simultaneously building the distinct foundations of body appreciation and intuitive eating.Level of evidence; Level V, descriptive studiesOur manuscript describes a cross-sectional design that uses a network analysis approach to map the existing correlations between variables. As this methodology is a descriptive study and does not involve an intervention (ruling out Levels I & II) or a longitudinal/case-control design (ruling out Level III), it aligns with the journal's criteria for Level V.