LAUSR

OBJECTIVE Psychological stress (PS) has always been a hot topic for researchers to investigate the potential determinants of childhood overweight/obesity. Up to now, cohort studies investigating the relationship between PS and childhood obesity have used different ways to assess PS, different indicators to measure obesity, different analysis methods, and the results were also inconsistent. METHODS Data were obtained from the second to eighth follow-up of an ongoing cohort of school-aged children in Chongqing, China (June 2015 - June 2018) with seven waves (W1-W7, NW1 = 1419). The latent growth curve model (LGCM) was used to estimate the co-developmental trends between PS and obesity (BMI, WHtR). Random intercept cross-lagged panel models (RI-CLPM) were constructed to examine their longitudinal bidirectional associations. RESULTS The changes in PS and obesity (BMI, WHtR) were co-developing (rBMI = -1.105, P = 0.003; rWHtR = -0.991, P = 0.004). Longitudinal models showed significant negative correlations between PS and obesity at the between-person level (rBMI = -4.993; rWHtR = -1.591). BMI at W3 could negatively predict PS six months later (β = -1.508, P = 0.027), and WHtR at W1 could negatively predict PS at W3 (β = -2.809, P = 0.014). Different aspects of PS had different associations with obesity. Notably, there was a significant reciprocal relationship between peer interaction PS and obesity. CONCLUSIONS Different aspects of PS were differentially associated with obesity. Notably, there may be a clear reciprocal relationship between peer interaction PS and obesity. These findings provide new directions for protecting children's mental health to prevent or control childhood overweight/obesity.