What is the clinical evidence from this study?

Study design: Cross-Sectional. Population: Hypertension (n=29700). Intervention: Random Forest machine learning model vs. Logistic regression, CatBoost, and MLP neural network models. Primary outcome: Area Under the Curve (AUC) for predicting hypertension risk.

September 24, 2021Open Access

Predicting the Risk of Hypertension Based on Several Easy-to-Collect Risk Factors: A Machine Learning Method

HZHuanhuan ZhaoNanjing Agricultural University XZXiaoyu ZhangHuaihua University YXYang XuChinese Academy of Sciences

Key Result

A Random Forest machine learning model based on easy-to-collect risk factors accurately predicted hypertension risk, achieving an AUC of 0.92 and outperforming CatBoost, MLP neural network, and logistic regression models.

Study Design

Type

Cross-Sectional (n=29,700)

Multicenter

Structured PICO

Can machine learning algorithms accurately predict the risk of hypertension using only easy-to-collect, non-invasive risk factors without clinical or genetic data?

Population

29,700 adults aged 20 to 70 years who underwent physical examinations in Beijing, China, were included to develop and validate machine learning models for predicting hypertension risk.

Exposure

Machine learning algorithms (Random Forest, CatBoost, MLP neural network, and Logistic Regression) using 10 easy-to-collect, non-invasive risk factors (age, gender, BMI, waist circumference, family history, occupation, smoke, drink, healthy diet, and physical activity) to predict hypertension risk.

Comparator

Comparison of performance among the four different machine learning models.

Outcome

Model performance evaluated by Area Under the Curve (AUC), accuracy, sensitivity, and specificity on the test set.

A Random Forest machine learning model using 10 easy-to-collect, non-invasive risk factors can accurately predict hypertension risk with an AUC of 0.92, offering a practical and economical tool for large-scale population screening.

Main Result

Absolute Event Rate: 0.92% vs 0.77%

Limitations

Data derived from cross-sectional physical examinations cannot predict absolute risk over time.
Data collected from a single local hospital in Beijing limits generalizability to other regions.
Did not evaluate the effect of all possible lifestyle variables because they were not included in the health examination.

Abstract

Hypertension is a widespread chronic disease. Risk prediction of hypertension is an intervention that contributes to the early prevention and management of hypertension. The implementation of such intervention requires an effective and easy-to-implement hypertension risk prediction model. This study evaluated and compared the performance of four machine learning algorithms on predicting the risk of hypertension based on easy-to-collect risk factors. A dataset of 29,700 samples collected through a physical examination was used for model training and testing. Firstly, we identified easy-to-collect risk factors of hypertension, through univariate logistic regression analysis. Then, based on the selected features, 10-fold cross-validation was utilized to optimize four models, random forest (RF), CatBoost, MLP neural network and logistic regression (LR), to find the best hyper-parameters on the training set. Finally, the performance of models was evaluated by AUC, accuracy, sensitivity and specificity on the test set. The experimental results showed that the RF model outperformed the other three models, and achieved an AUC of 0.92, an accuracy of 0.82, a sensitivity of 0.83 and a specificity of 0.81. In addition, Body Mass Index (BMI), age, family history and waist circumference (WC) are the four primary risk factors of hypertension. These findings reveal that it is feasible to use machine learning algorithms, especially RF, to predict hypertension risk without clinical or genetic data. The technique can provide a non-invasive and economical way for the prevention and management of hypertension in a large population.

Ask AI

Helpful

Bookmark

View Full Paper