Objective  Investigate the dynamic transition patterns among normal weight, overweight, and obesity in school-aged children aged 6-12 years in Chengdu, estimate the mean sojourn time and transition probabilities for different weight statuses; and analyze the effects of sex, baseline body mass index (BMI), annual height increment (AHI), and annual weight increment (AWI) on status transitions.

Methods Based on health examination data form primary and secondary school students in Chengdu, a total of 2510 school-aged children aged 6-12 years with at least three follow-up visits were enrolled, yielding 9543 person-time records. Weight status was classified according to the Screening for Overweight and Obesity among School-Age Children and Adolescents (WS/T 586—2018) criteria. A three-state homogeneous Markov model was constructed using chronological age as the time scale to estimate transition intensities, 1-year and 5-year transition probabilities, and mean sojourn times. Sex, baseline BMI, AHI, AWI, and baseline age were respectively included as covariates for univariate and multivariable analyses using the likelihood ratio test.

Results Among the 2510 participants, there were 1370 boys (54.58%) and 1140 girls (45.42%), with a mean baseline age of (7.60 ± 1.98) years. The baseline prevalence of normal weight, overweight, and obesity was 73.31% (n = 1840), 14.22% (n = 357), and 12.47% (n = 313), respectively. The null model showed that the mean sojourn time for overweight was the shortest (1.41 years 95% CI: 1.29-1.55), followed by obesity (3.42 years 95%CI: 2.94-3.98), while normal weight had the longest mean sojourn time (9.70 years 95% CI: 8.69-10.83). Within 1 year, 30.487% of overweight children reverted to normal weight, whereas 16.649% progressed to obesity; 77.216% of obese children remained obese. Subgroup analyses showed that both 1-year and 5-year follow-up results indicated shorter sojourn and mean sojourn times for overweight and obesity in girls compared with boys, while the mean sojourn time for normal weight was longer in girls (11.80 years) than in boys (8.34 years). Children with younger baseline age had lower probabilities of maintaining normal weight and overweight status, as well as shorter mean sojourn times. Multivariable analysis revealed that each 1 kg/m² increase in baseline BMI was associated with a 77% increase in the instantaneous transition intensity from normal weight to overweight (hazard ratio HR = 1.77, 95% CI: 1.56-2.02, P < 0.001), whereas it reduced the intensity of reversion from obesity to overweight by 27% (HR = 0.73, 95% CI: 0.66-0.80, P < 0.001). Additionally, each 1 kg/year increase in the time-varying AWI from the preceding year was associated with a 14% increase in the instantaneous transition intensity from normal weight to overweight (HR = 1.14, 95% CI: 1.05-1.24, P = 0.002).

Conclusion Overweight status is highly unstable, whereas obesity is relatively stable among school-aged children in Chengdu. Overweight represents a critical window for weight intervention. Priority for intervention should be given to overweight children and those in early school age, especially boys and children with higher baseline body weight.