Anthropometric metabolic subtypes and health outcomes: A data-driven cluster analysis.

Summary

Using 397,424 UK Biobank participants and replication in NHANES, six anthropometric-metabolic clusters were identified with distinct risks of all-cause, cardiovascular, and cancer mortality, MACE, and chronic renal failure. Clusters marked by low grip strength, high TG/HDL, high inflammation, or highest BMI had substantially elevated risks, whereas a high-BMI/high-grip cluster did not increase all-cause mortality.

Key Findings

• Identified six reproducible anthropometric-metabolic clusters in 397,424 UK Biobank participants and replicated associations in NHANES.
• Clusters with lowest grip strength, highest TG/HDL, highest NLR, or highest BMI showed substantially increased risks of all-cause, cardiovascular, and cancer mortality, incident MACE, and chronic renal failure.
• A high-BMI/high-grip cluster did not increase all-cause mortality but had small increases in selected outcomes (e.g., cardiovascular mortality, MACE).

Clinical Implications

Clinicians can move beyond BMI to classify patients into metabolic subtypes using simple measures (waist-to-height, grip strength, TG/HDL, NLR). This can guide tailored prevention (e.g., resistance training for low-grip phenotypes, lipid/inflammation targeting for TG/HDL or NLR-high clusters) and refine risk communication and surveillance.

Limitations

• Observational design with potential residual confounding and selection biases inherent to UK Biobank.
• Stability and transportability of clusters to diverse clinical settings require further validation.

Future Directions

Integrate cluster phenotypes into risk calculators and interventional trials (e.g., strength training, anti-inflammatory or lipid-lowering strategies) to test causality and clinical utility.