JMIR Publications

ABSTRACT

Background:

Hypertension (HTN) is a major global health issue and a significant modifiable risk factor for cardiovascular diseases (CVD), contributing to a substantial socioeconomic burden due to its high prevalence. In China, particularly among populations living near desert regions, hypertension is even more prevalent due to unique environmental and lifestyle conditions, exacerbating the disease burden in these areas.

Objective:

This study aims to develop and validate a machine learning-based model for predicting the risk of developing hypertension within a two-year period.

Methods:

The study utilized health examination data collected between 2019 and 2023 from populations in four regions surrounding the Taklamakan Desert of China. A retrospective cohort (n=1,038,170) and a prospective cohort (n=961,519) were constructed. Four machine learning algorithms and two deep learning methods were employed, with Bayesian optimization used to fine-tune the hyperparameters of those models.

Results:

The two-year incidence rate of hypertension was found to be 15.20% (n=157,766). Among the models developed, the CatBoost model demonstrated the best performance, achieving AUC values of 0.888 (95% CI: 0.886-0.889) in the retrospective cohort and 0.803 (95% CI: 0.801-0.804) in the prospective cohort. Additionally, a web-based prediction tool based on this model was developed to facilitate its application in clinical practice.

Conclusions:

The machine learning model developed in this study effectively predicts the two-year risk of hypertension, making it particularly suitable for preventive health management in high-risk populations in desert regions. The model exhibits excellent predictive performance and has potential for clinical application. With the development of a web-based application, this model can be widely implemented in public health and personalized medicine, supporting early intervention and prevention of hypertension.