Development and validation of a model to predict the risk of hypertension using anthropometric indicators in the Chinese population: a retrospective cohort study

Background: The prevalence of hypertension and obesity in China has sharply increased in recent decades. We aimed to develop and validate a novel model for predicting the risk of hypertension based on anthropometric indicators relating to obesity in the general population of China.

Methods: In this retrospective study, 6196 participants from the China Health and Nutrition Survey (CHNS) during the 2009-2015 waves were included. Risk factors for hypertension were assessed by LASSO regression combined with multivariate logistic regression analysis. A nomogram was developed as a predictive model based on the screening prediction factors. The discrimination and calibration of the model were evaluated by receiver operating curve (ROC) and calibration plots, respectively. Decision curve analysis (DCA) was used to evaluate the clinical application value of the model.

Results: A total of 6196 participants were divided into two sets at a ratio of 7:3, using computer-generated random numbers: 4337 individuals were assigned to the training set and 1859 to the validation set. The training set was divided into a hypertension group (n = 1016) and a non-hypertension group (n = 3321) based on the follow-up outcomes for hypertension. Predictive factors of hypertension included age, drinking, body mass index (BMI), systolic blood pressure (SBP), diastolic blood pressure (DBP), waist-to-hip ratio (WHR), waist-to-height ratio (WHtR) and arm-to-height ratio (AHtR) at baseline as predictors. The area under the ROC curve (AUC) for the training and validation sets was 0.906 (95% CI: 0.897-0.915) and 0.905 (95% CI: 0.887-0.922), respectively. In bootstrap validation, the C-index was 0.905 (95% CI: 0.888-0.921). The model also had good predictive accuracy according to the calibration plot. DCA demonstrated that people would benefit more when the threshold probability was between 5% and 80%.

Conclusion: A nomogram model was successfully established to effectively predict the risk of hypertension based on anthropometric indicators. The model could be a feasible tool for hypertension screening in the general population of China.