Perceived weight status, body image, and depressive symptoms in Korean adolescents: a three-wave longitudinal study and within-person analysis

Body dissatisfaction is a well-established risk factor for adolescent depression, but two questions remain open. First, it is unclear whether perceived weight status—a subjective cognitive appraisal of one’s body—relates to body image and depression differently from objective body size. Second, most longitudinal evidence relies on between-person models that cannot isolate within-person change. Drawing on a cognitive-behavioral model of body image, we distinguish perceived weight status (a perceptual-categorical appraisal) from body image (the attitudinal-evaluative dimension) and examine both between- and within-person processes. We analyzed three-wave longitudinal data from the Panel Study on Korean Children (ages 13–15; analytic N = 1,148–1,304). A temporally ordered mediation model (perceived weight status at Wave 14 → body image at Wave 15 → depressive symptoms at Wave 16) was estimated using structural equation modeling, controlling for prior depressive symptoms and sex-standardized BMI. Indirect effects were estimated using 10,000 bootstrap resamples with bias-corrected and accelerated (BCa) confidence intervals. We examined within-person dynamics using two- and three-variable random-intercept cross-lagged panel models (RI-CLPMs). Body image was assessed using a refined 4-item scale (α = .735, ω = .77). Perceiving oneself as heavier at W14 predicted a less positive body image at W15 (β = −0.126, p < .001), which in turn predicted higher depressive symptoms at W16 (β = −0.101, p < .001), yielding a small but significant indirect effect (0.013, 95% BCa CI [0.005, 0.025]). Sex-standardized BMI did not significantly predict body image (β = −0.064, p = .091), but the coefficients for perceived weight status and BMI did not differ significantly (Δ = −0.062, p = .368). The indirect effect did not differ by sex (Δ = −0.003, p = .776), although the component paths did. In the RI-CLPM, perceived weight status showed no prospective within-person association with body image ( p = .135) or depressive symptoms ( p = .594). In contrast, body image and depressive symptoms were reciprocally associated within persons (both ps < .01), independent of stable between-person differences. Perceived weight status was prospectively associated with body image, while body image and depressive symptoms showed reciprocal within-person associations. However, the mediational pathway was not supported at the within-person level, suggesting that the observed indirect association may primarily reflect between-person differences. Prevention and intervention efforts may therefore benefit from targeting the reciprocal body image–depression cycle rather than assuming a unidirectional, perception-driven pathway.