Introduction
The Sociocultural Attitudes Towards Appearance Questionnaire-4 Revised (SATAQ-4R) has been widely used in Western countries to link body appearance that is related to eating disorders and body dissatisfaction being commonly reported by adolescents. However, a comprehensive psychometric validation of the SATAQ-4R in Chinese adolescent samples is still lacking. To this end, the aim of the current study was to validate the gender-appropriate SATAQ-4R in a sample of Chinese adolescents, following by an investigation of its associations with body-related outcomes and eating disorder symptoms.


Methods
Two gender-specific studies were conducted to examine the psychometric properties of the SATAQ-4R-Female and SATAQ-4R-Male respectively among adolescent girls (Study1, N=344, with 73 participants at retest) and boys (Study2, N=335, with 64 participants at retest). Confirmatory factor analysis was employed to examine the factor structure and their test-retest reliability, the internal consistency and convergent validity were subsequently examined.


Results
For the SATAQ-4R-Females, the seven-factor model has a reasonable fit, with Chi-square =1112.769 (p < 0.001), CFI = 0.91, RMSEA = 0.071, SRMR = 0.067. For the SATAR-4R-Males, an acceptable seven-factor model with Chi-square = 982.92 (p<0.001), CFI = 0.91, RMSEA = 0.08, SRMR= 0.06 was observed. With respect to test-retest reliability, the internal consistency for 7 subscales was rated as good (Cronbach's alpha =0.74 to 0.95) among female adolescents, likewise the internal consistency of the seven subscales was also rated as good (Cronbach's alpha =0.70 to 0.96) among male participants. Good convergent validity was observed, reflected by associations of the subscales of the gender-specific SATAQ-4R with muscularity-related attitude, body image-acceptance, body appearance, perceived stress level, symptoms of eating disorder and self-esteem.


Discussion
For women and men, the original 7-factor structure was validated among Chinese adolescents, internal reliability coefficients for the seven subscale scores were good and test-retest reliability was acceptable. Our results also confirmed the convergent validity of the two different gender-appropriate scales.