BMI distribution and sex- and grade-specific trends in primary school health surveillance records in Xi’an, China, 2014–2024: a repeated cross-sectional study

Background Conventional childhood obesity surveillance relies mainly on mean body mass index (BMI) and overweight/obesity prevalence, which may mask unequal changes across the BMI distribution and population subgroups. Whether long-term BMI increases are concentrated in the upper tail and how high-BMI prevalence trends differ jointly by sex and school grade remain insufficiently characterized in consecutive school surveillance data. Methods We analyzed 7,187,779 records from students aged 6.0–12.99 years across 1,394 school codes and 13 districts in Xi’an, China, during 2014–2024 using a repeated cross-sectional design. High BMI comprised overweight or obesity under national sex- and grade-specific cutoffs. Primary analyses used logistic regression to estimate high-BMI prevalence and quantile regression to estimate covariate-adjusted changes in conditional BMI quantiles at the 50th, 75th, 90th, and 95th percentiles (P50, P75, P90, P95). Models adjusted for age in months, sex, grade, and district, with school-clustered inference. Secondary analyses assessed sex-by-grade differences in prevalence change; exploratory analyses examined P5, P10, and P25. Results Between 2014 and 2024, observed high-BMI prevalence increased from 17.01 to 21.11%. Adjusted prevalence peaked at 22.75% in 2020; its 2014–2024 increase was 3.35 percentage points (95% confidence interval [CI] 2.69–4.01). Covariate-adjusted changes in conditional BMI quantiles were 0.134 kg/m2 at P50 and 0.735 kg/m2 at P95; the P95 − P50 difference in 2014–2024 change was 0.601 kg/m2 (95% CI 0.468–0.734; p < 0.001). Exploratory estimates were 0.097 kg/m2 at P5 (pointwise 95% CI 0.027–0.167), 0.047 at P10 (−0.008–0.103), and 0.048 at P25 (−0.002–0.097). The increase in adjusted high-BMI prevalence was 1.65 percentage points greater among boys than girls (95% CI 1.17–2.14); the difference within Grades 4–6 was 2.59 percentage points (95% CI 1.95–3.23). Conclusion Among participating-school surveillance records, adjusted high-BMI prevalence increased between 2014 and 2024, with greater increases among boys than girls overall and within Grades 4–6. BMI increases were greater at the upper conditional quantiles than at the median. Joint reporting of high-BMI prevalence, upper-tail BMI changes, and sex- and grade-specific estimates may provide a more complete picture of BMI trends.