In a nationwide population-based psychometric investigation published in Eating and Weight Disorders – Studies on Anorexia, Bulimia and Obesity, researchers evaluating the Japanese version of the Sociocultural Attitudes Towards Appearance Questionnaire-4-Revised identified distinct, sex-specific factor structures among 1,368 Japanese adolescents aged 11 to 17 years. The cross-sectional study examined appearance ideal internalization and sociocultural pressure in a non-Western cohort where young people report high body dissatisfaction despite comparatively low body mass index values. Led by Takumi Saito, Kazue Ishitsuka, Aurélie Piedvache, Satoshi Yamaguchi, Syudo Yamasaki, Satoshi Usami, Akira Ishiguro, and Naho Morisaki, the team evaluated factor structures, reliability, and convergent validity against eating pathology inventories. Exploratory factor analysis indicated a five-factor, 23-item structure for girls and a seven-factor, 23-item structure for boys. Confirmatory factor analysis demonstrated suboptimal fit for the female model and comparatively better fit for the male model, alongside inadequate reliability in one female pressure subscale.

The theoretical rationale for the study rests on the tripartite influence model of body image and eating disturbances, which posits that appearance pressures originate from three primary sociocultural sources: peers, family, and media. These influences foster the internalization of appearance ideals, referring to the extent to which individuals cognitively endorse socially prescribed beauty standards, express desires to attain them, and modify their behaviors accordingly. Adolescents are particularly susceptible to these pressures during this developmental window, when body image concerns and eating disorders frequently emerge. The Sociocultural Attitudes Towards Appearance Questionnaire has evolved since its initial 1995 release to capture evolving aesthetic standards, with the revised fourth version designed to measure muscular ideal internalization, general attractiveness, and significant-other pressures more precisely. Although validated among American college and adolescent cohorts, international adaptations in Italy, Turkey, Brazil, and China reported factor merging, item deletions, and varying fit, demonstrating that the scale constructs are culturally sensitive and warrant direct evaluation in Japan.

Data for the investigation were gathered through the Japan Adolescent and Youth Cohort Study, which commenced as a pilot study in 2020. The investigators utilized a stratified two-stage cluster sampling design based on Japan’s statutory resident registration system under the Residential Basic Book Act, ensuring a representative national household sample. Stratification divided Japan into eight geographical regions: Hokkaido, Tohoku, Kanto, Hokuriku-Koushinetsu, Chubu, Kinki, Chugoku-Shikoku, and Kyushu-Okinawa. Initially, researchers randomly selected six to seven municipalities from each prefecture across fifty municipalities, subsequently selecting thirty households each with fifth-grade elementary and second-grade junior high school students. In 2021, the frame expanded to include thirty households across all remaining junior high school grades up to age fifteen, continuing annually with thirty newly recruited fifth-grade households per municipality. The analytical sample included 1,368 adolescents from grades six through eleven who completed mail surveys in autumn 2023 with full questionnaire responses, comprising 694 girls and 674 boys with matching mean ages of 14.1 years and standard deviations of 1.8 years.

The adaptation of the questionnaire followed standard forward and backward translation procedures conducted after obtaining permission from the original instrument developers. Original English items were forward-translated into Japanese and independently back-translated into English, with semantic equivalence, cultural appropriateness, and linguistic comprehensibility reviewed by experts and informally checked with an age-appropriate child. The researchers deliberately avoided referencing earlier Japanese translations of the fourth edition during this process. In their original English formats, the male instrument contains 28 items while the female instrument contains 31 items, each distributed across seven theoretical dimensions: Internalization: Thin/Low Body Fat, Internalization: Muscular, Internalization: General Attractiveness, Pressure: Peers, Pressure: Family, Pressure: Significant Others, and Pressure: Media. Every item is scored using a five-point Likert scale ranging from 1, defined as definitely disagree, to 5, defined as definitely agree, capturing both the personal valuation of specific body types and perceived societal pressures to conform to those standards.

To assess convergent validity, the researchers administered the Children’s Eating Attitudes Test and the eight-item short form of the Eating Disorder Examination-Questionnaire. The Children’s Eating Attitudes Test comprises 26 self-report items rated on a six-point scale from never to always, evaluating dieting behaviors, food preoccupation, bulimia, and concerns about being overweight, with validated psychometric properties in Japanese pediatric cohorts. The eight-item eating disorder examination assesses cognitive and behavioral eating pathology over the preceding fourteen days, including restraint, food avoidance, food preoccupation, feelings of fatness, weight-loss desires, eating guilt, weight dissatisfaction, and discomfort viewing one’s body. Items are scored on a seven-point scale from 0, indicating no days or not at all, to 6, indicating every day or markedly. Because the Japanese translation of this eight-item inventory was prepared for the cohort without formal prior validation in adolescents, its correlations were treated as exploratory, although its sample internal consistency was high, with Cronbach’s alpha values of 0.90 for girls and 0.85 for boys.

Statistical analyses began with exploratory factor analysis using principal axis factoring with promax oblique rotation to accommodate correlated dimensions. Suitability was confirmed through statistically significant Bartlett’s tests of sphericity and Kaiser-Meyer-Olkin values exceeding the 0.60 threshold. Factor retention incorporated Kaiser-Guttman criteria, scree plots, parallel analysis, and theoretical considerations, with candidate item deletion applied to primary factor loadings at or below 0.40 or secondary cross-loadings at or above 0.30. Confirmatory factor analysis treated responses as approximately continuous using maximum likelihood estimation in Stata, supplemented by ordinal sensitivity analyses conducted in R with the lavaan package. Model evaluation utilized the comparative fit index, Tucker-Lewis index, root mean square error of approximation, and standardized root mean square residual, defining good fit by comparative indices of 0.95 or higher and approximation errors at or below 0.08. Reliability was evaluated using Cronbach’s alpha and McDonald’s omega for dimensions with three or more items, alongside inter-item correlations and Spearman-Brown coefficients for two-item factors.

For the female cohort of 694 participants, exploratory factor analysis indicated a five-factor structure with significant sphericity (chi-square = 15,873.99, df = 465, p < 0.001) and a Kaiser-Meyer-Olkin value of 0.92. Items 10 and 13 exhibited low factor loadings across all dimensions, while items 20, 21, 22, 24, 26, and 27 showed cross-loadings, leading to their exclusion. In the resulting 23-item female model, thin-ideal internalization merged with general attractiveness into a single construct, while peer pressure combined with significant-other pressure into another single construct, alongside muscular internalization, family pressure, and media pressure. Primary confirmatory factor analysis showed suboptimal fit (CFI = 0.88, RMSEA = 0.10, SRMR = 0.08), while ordinal sensitivity analysis showed mixed fit (CFI = 0.978, TLI = 0.975, RMSEA = 0.102, SRMR = 0.090). While most subscales demonstrated acceptable reliability, the two-item peer and significant-other pressure subscale exhibited inadequate internal consistency, with an alpha of 0.49, an inter-item correlation of 0.37, and a Spearman-Brown coefficient of 0.54.

For the male cohort of 674 participants, exploratory factor analysis supported a seven-factor structure matching the original instrument, with significant sphericity (chi-square = 15,948.55, df = 378, p < 0.001) and a Kaiser-Meyer-Olkin value of 0.93. Items 9, 10, 17, 18, and 19 were excluded due to low loadings or cross-loadings. Item 4 showed a loading of 0.39, marginally below the 0.40 cutoff, but was retained to preserve the theoretical Thin/Low Body Fat construct and prevent a single-item indicator. The resulting 23-item male confirmatory model showed comparatively good fit (CFI = 0.94, RMSEA = 0.08, SRMR = 0.04), while ordinal sensitivity analysis indicated strong fit (CFI = 0.988, TLI = 0.986, RMSEA = 0.063, SRMR = 0.038). All male subscales displayed Cronbach’s alpha values of 0.75 or higher, and McDonald’s omega ranged from 0.825 to 0.951 for multi-item factors. Inter-subscale correlations ranged from -0.04 to 0.69, showing non-significant associations between general attractiveness and muscular internalization, family pressure, and significant-other pressure.

Bivariate correlations with external eating pathology measures provided mixed evidence of convergent validity across sexes. In adolescent girls, correlations with eating scores ranged from 0.24 to 0.65, showing the strongest associations for media pressure and the combined thin-ideal and general attractiveness subscale. In boys, correlations ranged from 0.07 to 0.49, with the strongest association observed for thin-ideal internalization and notably weak associations for general attractiveness. The authors observed that the combining of peer and significant-other pressures in girls aligns with findings from American, Brazilian, and Turkish cohorts, likely reflecting developmental overlap where peer and dating relationships are not sharply differentiated in early adolescence. The merging of thin-ideal and general attractiveness in girls suggests these ideals are conceptually intertwined in Japan. Conversely, the weak associations for male general attractiveness may reflect reverse-coded items that boys interpreted as referring to facial attractiveness rather than body shape, while muscular ideal internalization correlated only modestly with disordered eating, consistent with evidence that muscularity concerns link primarily to weight-gain strategies.

The investigators highlighted several critical limitations that qualify these psychometric results. Classified as a Level III cross-sectional observational study, the design cannot determine causal relationships or track temporal changes. Because factor structures differed between sexes, multigroup measurement invariance could not be established, meaning direct score comparisons between adolescent girls and boys are unsupported. Multiple item deletions may have narrowed construct coverage, several dimensions were reduced to two items, and the female peer and significant-other subscale exhibited inadequate reliability. Additionally, the eight-item eating disorder examination lacks full Japanese pediatric validation, translation procedures omitted formal adolescent cognitive interviewing, and participants were restricted to ages 10 to 17 years. The authors concluded that while the Japanese SATAQ-4R provides preliminary psychometric utility, its modified structures, suboptimal female fit, and mixed convergent validity require cautious interpretation. Future research must evaluate test-retest reliability, discriminant validity, and cultural understandings across broader age groups before applying these models in clinical or epidemiological monitoring.

Subject of Research: Psychometric validation of the Japanese SATAQ-4R among adolescents

Article Title: Development and validation of the Japanese version of Sociocultural Attitudes Towards Appearance Questionnaire-4-Revised (SATAQ-4R) among Japanese adolescents

Article References: Saito, T., Ishitsuka, K., Piedvache, A., Yamaguchi, S., Yamasaki, S., Usami, S., Ishiguro, A., & Morisaki, N. (2026). Development and validation of the Japanese version of Sociocultural Attitudes Towards Appearance Questionnaire-4-Revised (SATAQ-4R) among Japanese adolescents. Eating and Weight Disorders – Studies on Anorexia, Bulimia and Obesity, 31(1), Article 76. https://doi.org/10.1007/s40519-026-01877-x

Image Credits: AI Generated

DOI: 10.1007/s40519-026-01877-x

Keywords: SATAQ-4R, Japanese adolescents, body image, appearance ideals, psychometric validation, eating disorders, sociocultural pressures, factor analysis, Development, validation, Japanese, version