Prevention & lipidsDiagnostic or prognostic study
Automated Abdominal Aortic Calcification Scores and Atherosclerotic Cardiovascular Disease in the UK Biobank Imaging Study
- external validation
- machine learning-derived abdominal aortic calcification (AAC) 24-point algorithm (ML-AAC24)
- age- and sex-adjusted analysis
Authors' conclusion
Assessing ML-AAC24 on lateral spine images offers a new and promising screening method to identify people with higher risk of incident ASVD events.
Study context
- Population
- Middle-aged individuals from the UK Biobank Imaging Study with lateral spine images, obtained via dual-energy x-ray absorptiometry52% female, mean age 65 years, 50,923 after excluding people with prevalent atherosclerotic cardiovascular disease (ASCVD) or missing data
- Participants
- 53,611 individuals
- Comparison
- compared to those with low ML-AAC24, those with moderate and high ML-AAC24
Performance
As reported in the abstract
ML-AAC24 distribution
Further findings
| Outcome | HR (95% CI) | |
|---|---|---|
| Incident ASCVD event · median follow-up of 4.1 [3.0-5.5] years — moderate ML-AAC24 vs low ML-AAC24 · age- and sex-adjusted · After excluding people with prevalent ASCVD or missing data | 1.80 (1.57–2.08) | |
| Incident ASCVD event · median follow-up of 4.1 [3.0-5.5] years — high ML-AAC24 vs low ML-AAC24 · age- and sex-adjusted · After excluding people with prevalent ASCVD or missing data | 2.87 (2.39–3.44) | |
| 0.20.5125Hazard ratio (log scale) |
- Incident ASCVD event — incident ASCVD event, of 50,923 people
- 1,163 (2.3%)
Incident ASCVD event · median follow-up of 4.1 [3.0-5.5] years
- Results remained comparable after adjustment for established ASCVD risk factors
Incident coronary artery disease, myocardial infarction, and stroke
- Consistent patterns were observed
Original publication
Automated Abdominal Aortic Calcification Scores and Atherosclerotic Cardiovascular Disease in the UK Biobank Imaging Study. JACC Adv. Published 29 January 2026. doi:10.1016/j.jacadv.2025.102570
Summarised from the published abstract; every number is checked against it