Development and validation of a nomogram for predicting all-cause mortality in American adult hypertensive populations

Long Yang; Xia Shen; Zulihuma Seyiti; Jing Tang; Abudushalamu Kasimujiang; Tuohutasheng Dejite; Ling Zhao; Xue-Feng Shan; Xiao-Ming Gao

doi:10.3389/fphar.2023.1266870

Development and validation of a nomogram for predicting all-cause mortality in American adult hypertensive populations

Front Pharmacol. 2023 Nov 22:14:1266870. doi: 10.3389/fphar.2023.1266870. eCollection 2023.

Authors

Long Yang^#¹, Xia Shen^#², Zulihuma Seyiti¹, Jing Tang^{1

3

4}, Abudushalamu Kasimujiang¹, Tuohutasheng Dejite¹, Ling Zhao^{4

5}, Xue-Feng Shan^{1

4

5

6}, Xiao-Ming Gao^{1

4

5}

Affiliations

¹ State Key Laboratory of Pathogenesis, Prevention and Treatment of High Incidence Diseases in Central Asia, College of Pediatrics, Xinjiang Medical University, Clinical Medical Research Institute of Xinjiang Medical University, Urumqi, China.
² Department of Nursing, Wuxi Medical College of Jiangnan University, Wuxi, China.
³ Clinical Laboratory, First Affiliated Hospital of Xinjiang Medical University, Urumqi, China.
⁴ Xinjiang Key Laboratory of Medical Animal Model Research, Urumqi, China.
⁵ Department of Cardiology of the First Affiliated Hospital of Xinjiang Medical University, Urumqi, China.
⁶ Pediatric Cardiothoracic Surgery, First Affiliated Hospital of Xinjiang Medical University, Urumqi, China.

^# Contributed equally.

Abstract

Backgrounds: Hypertension stands as the predominant global cause of mortality. A notable deficiency exists in terms of predictive models for mortality among individuals with hypertension. We aim to devise an effective nomogram model that possesses the capability to forecast all-cause mortality within hypertensive populations. Methods: The data for this study were drawn from nine successive cycles of the National Health and Nutrition Examination Survey (NHANES) spanning the years from 1999 to 2016. The dataset was partitioned into training and validation sets at a 7:3 ratio. We opted for clinical practice-relevant indicators, applied the least absolute shrinkage and selection operator (LASSO) regression to identify the most pertinent variables, and subsequently built a nomogram model. We also employed concordance index, receiver operating characteristic (ROC) curves, calibration curves and decision curve analysis (DCA) to assess the model's validity. Results: A total of 17,125 hypertensive participants were included in this study with a division into a training set (11,993 individuals) and a validation set (5,132 individuals). LASSO regression was applied for the training set to obtain nine variables including age, monocytes, neutrophils, serum albumin, serum potassium, cardiovascular disease, diabetes, serum creatinine and glycated hemoglobin (HbA1C), and constructed a nomogram prediction model. To validate this model, data from the training and validation sets were used for validation separately. The concordance index of the nomogram model was 0.800 (95% CI, 0.792-0.808, p < 0.001) based on the training set and 0.793 (95% CI, 0.781-0.805, p < 0.001) based on the validation set. The ROC curves, calibration curves, and DCA curves all showed good predictive performance. Conclusion: We have developed a nomogram that effectively forecasts the risk of all-cause mortality among American adults in hypertensive populations. Clinicians may use this nomogram to assess patient's prognosis and choose a proper intervention in a timely manner.

Keywords: LASSO; NHANES; hypertension; machine learning; mortality; nomogram.

Grants and funding

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This work was supported by the grants of the Natural Science Foundation of China (No. U1903212 and 82360051), the key project of the Natural Science Foundation of Xinjiang Uygur Autonomous Region, Special projects on Cardiovascular Disease from the State Key Laboratory of Pathogenesis, Prevention and Treatment of Central Asia High Incidence Diseases (SKL-HIDCA-2022-XXG1, SKL-HIDCA-2021-XXG1), an opening project of the Xinjiang Key Laboratory (2021D04020), and the Xinjiang Medical University Postgraduate Scientific Research innovatian Project (XJ2023G156).