# Development and validation of a nomogram based on LASSO-logistic regression for predicting carotid atherosclerosis in patients with hypertension

## TL;DR

This study creates a tool to predict carotid atherosclerosis risk in hypertension patients using factors like age, diabetes, and cholesterol.

## Contribution

A novel nomogram based on LASSO-logistic regression for predicting carotid atherosclerosis in hypertensive patients.

## Key findings

- Eight risk factors including diabetes were identified as significant predictors of carotid atherosclerosis.
- The nomogram showed strong predictive accuracy with AUC values of 0.858 in the development cohort and 0.808 in the validation cohort.
- Calibration curves and decision curve analysis confirmed the model's reliability and clinical utility.

## Abstract

Carotid atherosclerosis (CAS) is increasingly prevalent among hypertensive patients. This study aims to develop a predictive nomogram for CAS in hypertensive population.

A total of 930 patients with hypertension were hospitalized in the Department of Cardiology of the Affiliated Hospital of Changzhou, Nanjing University of Chinese Medicine (August 2018–August 2024) formed the development cohort, categorized into CAS (156 individuals) and non-CAS (774 individuals) groups. Additionally, 398 hypertensive patients from the Department of Cardiology of the Second Affiliated Hospital of Soochow University served as the validation cohort (ratio 7:3), with 72 CAS individuals and 326 non-CAS individuals. LASSO regression initially identified key risk factors, followed by logistic regression for further analysis. The nomogram, constructed using the “rms” package in R 4.2.6, underwent internal validation via the 1,000 iterations of Bootstrap resampling. Model performance was evaluated through ROC curves, calibration curves, and decision curve analysis.

Eight significant risk factors—Age, history of smoking (Smoke), history of diabetes mellitus (DM), course of hypertension (Course), physical activity (PA), body mass index (BMI), low-density lipoprotein (LDL), and uric acid (UA)—were identified (P < 0.05), among which DM was the most important influencing factor. The nomogram demonstrated strong predictive accuracy, with AUC values of 0.858 [95% CI (0.798, 0.918)] in the development cohort and 0.808 [95% CI (0.740, 0.876)] in the validation cohort. Calibration curves closely aligned with the ideal model, and decision curve analysis indicated optimal predictive performance within a probability threshold range of 0.050–0.960.

This study presents a robust nomogram for assessing CAS risk in hypertensive patients, offering a valuable tool for clinical risk evaluation.

Nomogram predicting carotid atherosclerosis risk in hypertension patients. Includes 930 hypertension participants, dividing into 156 with carotid atherosclerosis and 774 without. Risk factors: age, diabetes, smoking, low-density lipoprotein, body mass index, and uric acid. Logistic regression analysis is used.