Distinct relationships of compartment-specific fat distribution profiles with cardiovascular ageing and future cardiovascular events
- UK Biobank study
- machine learning-estimated biological heart age
- K-means clustering
- mediation analysis
- sex-stratified analyses
+4 more
- machine learning-estimated biological heart age
- K-means clustering
- mediation analysis
- sex-stratified analyses
Authors' conclusion
Among the adiposity compartments studied, VAT showed the strongest relationships with cardiovascular ageing and incident CVD. While PAT showed initial associations, these were not independent of VAT. Imaging-defined adiposity phenotypes can allow more precise definition of cardiovascular risk and may enhance mechanistic understanding of obesity-related CVD, including sex-specific susceptibilities.
Study context
- Population
- UK Biobank participants without cardiovascular disease (CVD), with cardiac and abdominal MRI and linked health records55%, n=18,978 females, average age 63.5 years
- Participants
- 34,496 participants
- Comparison
- Imaging-defined adiposity phenotypes compared with anthropometric measures
Main finding
As reported in the abstract
Body mass index
Median body mass index (BMI)
25.7 kg/m 2
Overweight or obese
58.4%
Further findings
Biological heart age by adiposity cluster
- High-risk phenotype associated with greater heart ageing
- Lower-risk phenotype linked with reduced ageing
- Imaging-defined adiposity clusters outperformed BMI categories in their association with biological heart age
Mediation of association with incident CVD through accelerated heart ageing
- Visceral adipose tissue (VAT), proportion mediated (strongest across all fat depots)
- 14%
- Pericardial adipose tissue (PAT), proportion mediated
- 10.7%
- PAT, proportion mediated after adjustment for VAT
- <1%
Sex differences
- Associations were more pronounced among males
Original publication
Distinct relationships of compartment-specific fat distribution profiles with cardiovascular ageing and future cardiovascular events. Heart. Published 25 August 2026. doi:10.1136/heartjnl-2026-327946
Summarised from the published abstract; every number is checked against it