Objective:

To identify cardio-metabolic risk profiles of individuals with general and central obesity among middle-aged and older Chinese, highlighting the significance for clinical practice using unsupervised machine learning.

Key Findings:

• Unsupervised machine learning effectively identifies subtypes of obesity based on cardio-metabolic profiles, which can inform targeted interventions.
• Central obesity is more strongly associated with non-communicable diseases compared to general obesity, underscoring the need for focused clinical attention.
• The study provides insights for precise risk stratification and diagnosis in clinical settings, potentially improving patient outcomes.

Interpretation:

The findings suggest that unsupervised machine learning can enhance understanding of obesity subtypes, leading to better clinical management and prevention strategies, particularly in high-risk populations.

Limitations:

• The study relies on observational data, which may limit causal inferences; biases in self-reported health conditions and lifestyle factors could skew results.

Conclusion:

This research highlights the utility of unsupervised machine learning in analyzing complex obesity profiles, paving the way for improved clinical approaches to obesity management and addressing a significant public health issue.