The atherogenic index of plasma predicts long-term outcomes in patients with severe coronary artery calcification undergoing rotational atherectomy: a machine learning-based cohort study
Authors' conclusion
AIP was independently associated with adverse outcomes in patients with severe CAC undergoing RA. The integration of this low-cost, readily available biomarker into explainable machine learning frameworks offers a promising avenue for enhancing risk prediction models.
Study context
- Population
- Patients with severe coronary artery calcification undergoing rotational atherectomy, January 2017–October 2024; 58.28% male
- Participants
- 513 patients
- Exposure
- Highest vs lowest atherogenic index of plasma (AIP) tertile
- Outcome
- Cardiovascular death or all-cause death
- Follow-up
- Median 40.55 months
Methodological detailsmachine learning-based, AIP tertiles, multivariable Cox regression
- Machine learning-based
- AIP tertiles
- Multivariable Cox regression
Main finding
As reported in the abstract
Cardiovascular death; all-cause death
Highest AIP tertile vs Lowest AIP tertileHazard ratio (95% CI)
Cardiovascular death
2.40(1.40–4.13)
All-cause death
1.68(1.13–2.51)
Observational data: associations, not effects.
Secondary outcomes
| Outcome | HR (95% CI) | |
|---|---|---|
| Non-fatal MI, target vessel revascularization, strokeHighest AIP tertilevs Lowest AIP tertileNon-fatal MI · multivariable Cox analysis | 2.27 (1.20–4.29) | |
| Non-fatal MI, target vessel revascularization, strokeHighest AIP tertilevs Lowest AIP tertileTarget vessel revascularization · multivariable Cox analysis | 1.74 (1.04–2.90) | |
| Non-fatal MI, target vessel revascularization, strokeHighest AIP tertilevs Lowest AIP tertileStroke · multivariable Cox analysis | 1.92 (1.00–3.69) | |
| 0.20.5125Hazard ratio (log scale) |
Dose-response and subgroups
- Restricted cubic splines indicated a dose-response relationship between AIP and adverse outcomes
- AIP–mortality association stronger in elderly, cardiac dysfunction or poor glycemic control
- High-AIP group had worse long-term outcomes on Kaplan-Meier curves
Machine learning prediction of cardiac death
- Random forest model, best of six algorithms: AUC
- 0.800
- SHAP identified AIP as a key feature driving predictions
Original publication
The atherogenic index of plasma predicts long-term outcomes in patients with severe coronary artery calcification undergoing rotational atherectomy: a machine learning-based cohort study. Cardiovasc Diabetol. Published 1 December 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →