Developmental trajectories of physical fitness and health and their influencing factors among primary school students in Suzhou, China (2015–2023): an accelerated longitudinal study

Background This study characterized developmental trajectories of physical fitness and health among primary school students, identified latent trajectory subgroups, and examined associations with sex, baseline weight status and the COVID-19 pandemic. Methods We analyzed mandatory annual National Student Physical Fitness Standard test records of students in grades 1–6 at five primary schools in Suzhou, China (2015–2023). After linking records across student-ID changes, 8,280 records from 2,238 students formed an accelerated longitudinal design. A three-level growth model with a piecewise term at 2020, using grade as the developmental axis and calendar year as a period covariate, described the average trajectory of the total score. Group-based trajectory modeling (GBTM) identified latent classes, and multinomial logistic regression examined their predictors. Sensitivity analyses examined cohort splice-ability, attrition, breakpoints, classification uncertainty and the scored body mass index (BMI) item. Results The total score declined across grades (−0.73 per grade, P < 0.001), but within-student scores were stable before 2020 (+0.29 per year, P = 0.14) because a positive period trend offset this decline; cohort splice-ability was only partly supported. From 2020, the period trend changed from +1.02 to −1.35 points per year ( P < 0.001); breakpoints in 2021–2022 fitted better. Girls scored higher than boys on this sex-specific composite (+2.80, P < 0.001). Students with obesity scored 6.99 points lower than normal-weight peers ( P < 0.001), about two-thirds of which reflected the scored BMI item. GBTM identified four classes: persistently low (11.4%), below-average declining (33.4%), good-but-declining (35.7%) and high-then-declining (19.6%) (average posterior probabilities 0.80–0.90). Relative to the good-but-declining class, girls were less likely to belong to the lower-level classes, whereas baseline overweight/obesity was associated with the persistently low (OR = 5.25) and below-average declining (OR = 2.42) classes. Conclusions Scores were stable within students before 2020 and declined thereafter, coinciding with the pandemic period; the decline across grades depends on how developmental and period effects are separated. Trajectories were heterogeneous, and baseline overweight/obesity was associated with weaker trajectories beyond the scored BMI item. Early identification and targeted intervention should prioritize students in the persistently low and high-then-declining classes.