Development and validation of a risk predictive model in young patients with hypertension

Hypertension is increasingly prevalent in young people and is associated with poor long-term outcomes and an elevated risk of cardiovascular events. Therefore, it is important to identify reliable early predictors of hypertension formation in this population. Between January 2016 and December 2024, we collected data from 1242 young adults (33.86 ± 0.16) with essential hypertension and 536 healthy volunteers (33.90 ± 0.26). We split the hypertensive patients and healthy population in a 7:3 ratio between the training and validation sets. Logistic regression analysis was used to develop predictive models, and the receiver operating characteristic (ROC) curve was used to evaluate the model's effectiveness. The study included 1778 participants. Logistic regression analysis identified γ-glutamyl transpeptidase and urea as independent risk factors for hypertension, while superoxide dismutase was an independent protective factor. The ROC curve showed a sensitivity of 80.2% and specificity of 82.6%. We developed a simplified scoring system for each index, and the resulting ROC curve had an area under the curve of 0.87. The calibration analysis curve showed a mean absolute error of 0.01, and the clinical decision analysis curve indicated a positive benefit when the threshold probability was between 0.04 and 1.00. In conclusion, γ-glutamyl transpeptidase, superoxide dismutase, and urea are important predictors of hypertension in young people, and our findings suggest potential new diagnostic and treatment directions for hypertension prevention and treatment in this population.