A large-scale longitudinal study of Chinese university students has revealed that physical fitness does not decline uniformly across the student population, but follows three distinct trajectories — and one subgroup is deteriorating far faster than the rest.

According to research published in Frontiers in Public Health on 19 November 2026, researchers at Fujian Medical University analysed routine annual physical fitness test records from 24,111 undergraduate students at a medical university in China, collected between 2021 and 2025. Of those students, 18,131 (75.2%) had been tested in at least two of those years, forming the core longitudinal dataset of 46,082 records.

The study constructed a composite Physical Health Index (PHI) covering six domains: body composition (BMI), pulmonary function (vital capacity), speed (50-metre dash), lower-body explosive power (standing long jump), flexibility (sit-and-reach), and sex-specific strength and endurance tests. Scores were standardised within sex and averaged with equal weights.

Overall, the index declined for both sexes across the five-year period. Male students' average score fell from 0.160 to −0.205 Z-scores, while female students' average dropped from 0.171 to −0.220 Z-scores. Male students also showed sharp increases in BMI and in rates of overweight and obesity.

Using a statistical method called latent class growth analysis, which identifies subgroups with similar patterns of change without predefining them, the researchers found three distinct trajectories:

  • A low-baseline rapid decline group (11.5% of students), who entered university with the lowest fitness scores (baseline index −0.722) and declined most steeply (a change of 0.391 Z-scores across the study window). This group had the highest average baseline BMI (23.3 kg/m²) and scored lowest on vital capacity, standing long jump, and sit-and-reach.
  • A stable low-moderate group (52.8%), with a baseline index of −0.029 and a moderate decline of 0.181 Z-scores.
  • A high-baseline slow decline group (35.7%), who started with the highest fitness (baseline 0.579) and declined most slowly (0.089 Z-scores).

Importantly, the study found that BMI trajectories and composite fitness trajectories showed almost no concordance (adjusted Rand index 0.036; kappa −0.055 after aligning class labels). This means that tracking weight alone would miss the students whose overall fitness is declining fastest, the researchers said.

The three-class structure remained stable across sensitivity analyses that tested alternative time axes, standardisation references, and weighting schemes.

What this means for students

For university students worldwide, the findings suggest that a one-size-fits-all approach to fitness — such as a single gym requirement or a general wellness campaign — is unlikely to reach those who need support most. The study's authors argue that health management on campus should move beyond single indicators like BMI and implement stratified, longitudinal monitoring from the first year of study.

Students who enter university with higher BMI and lower scores across multiple fitness tests may be at particular risk of rapid decline, and could benefit from early, targeted interventions. The study also highlights that weight change and overall fitness are different dimensions of health: a student whose weight stays stable could still be losing cardiorespiratory or muscular fitness.

Limitations

The study has several limitations. It is based on data from a single medical university in China, so the findings may not generalise to other institutions, countries, or student populations. The research is retrospective, using data collected for administrative purposes rather than a controlled experiment. The composite index was standardised within the study sample itself, which means the absolute Z-score values are relative to this specific cohort, not to an external national or international reference. The study also did not examine behavioural or environmental factors — such as diet, sleep, screen time, or access to sports facilities — that might explain why some students decline faster than others.

Because the research was published in a single peer-reviewed journal and has not yet been independently replicated, the findings should be treated as preliminary evidence rather than settled fact.

The bottom line

About one in nine university students in this cohort followed a trajectory of rapid fitness decline, entering with lower baseline fitness and higher BMI and deteriorating faster than their peers. The study provides a proof of concept for using routine fitness test data to identify at-risk students early, but broader and more diverse studies would be needed before such an approach could be recommended for campuses outside China.