Comparison of nomograms designed to predict hypertension with a complex sample

고혈압 예측을 위한 노모그램 구축 및 비교

  • Received : 2020.06.03
  • Accepted : 2020.08.06
  • Published : 2020.10.31


Hypertension has a steadily increasing incidence rate as well as represents a risk factors for secondary diseases such as cardiovascular disease. Therefore, it is important to predict the incidence rate of the disease. In this study, we constructed nomograms that can predict the incidence rate of hypertension. We use data from the Korean National Health and Nutrition Examination Survey (KNHANES) for 2013-2016. The complex sampling data required the use of a Rao-Scott chi-squared test to identify 10 risk factors for hypertension. Smoking and exercise variables were not statistically significant in the Logistic regression; therefore, eight effects were selected as risk factors for hypertension. Logistic and Bayesian nomograms constructed from the selected risk factors were proposed and compared. The constructed nomograms were then verified using a receiver operating characteristics curve and calibration plot.

고혈압은 발병률이 꾸준히 증가하고 있을 뿐 아니라, 심혈관 질환과 같은 2차 질병의 주된 위험 요인이 되었다. 게다가 고혈압은 뇌졸중, 혈관성 치매와 같은 다른 합병증을 유발하는 질병이다. 따라서 고혈압 발병률을 예측하는 것은 중요한 일이다. 본 연구에서, 고혈압 발병률을 예측할 수 있는 노모그램을 구축하였다. 데이터는 2013년부터 2016년까지의 국민건강영양조사로부터 얻어졌다. 복합 표본의 특성을 고려하여 Rao-Scott chi-squared test를 통해 고혈압에 영향을 미치는 10가지 요인을 규명하였다. 하지만 로지스틱 회귀분석 시, 흡연 상태와, 운동 유무는 유의하지 않았다. 따라서 8개의 주 효과를 고혈압의 위험요인으로 최종 선별하였다. 그리고 최종 선별된 위험 요인들로 로지스틱 노모그램과 베이지안 노모그램을 제시 및 비교하였다. 마지막으로 ROC curve 그래프와 calibration plot을 통해 노모그램을 검증하였다.



  1. Akobeng, A. K. (2007). Understanding diagnostic tests 3: receiver operating characteristic curves, Acta Paediatrica, 96, 644-647.
  2. Cook, N. R. (2008). Statistical evaluation of prognostic versus diagnostic models: beyond the ROC curve, Clinical Chemistry, 54, 17-23.
  3. D'Agostino Sr, R. B., Grundy, S., Sullivan, L. M., Wilson, P., and CHD Risk Prediction Group (2001). Validation of the Framingham coronary heart disease prediction scores: results of a multiple ethnic groups investigation, Jama, 286, 180-187.
  4. Iasonos, A., Schrag, D., Raj, G. V., and Panageas, K. S. (2008). How to build and interpret a nomogram for cancer prognosis, Journal of Clinical Oncology, 26, 1364-1370.
  5. Kim, M. H. (2020). Nomogram model for predicting the incidence of hypertension with complex sample (Master's thesis), Yeungnam University, Gyeongsan.
  6. Kim, M. H. and Lee, J. Y. (2020). How to construct a nomogram for hypertension using complex sampling data from Korean adults, Communications in Statistics-Theory and Methods, Published online: 07 June 2020.
  7. Kim, M. H., Seo, J. H., and Lee, J. Y. (2019). Nomogram building to predict dyslipidemia using a naive Bayesian classifier model, The Korean Journal of Applied Statistics, 32, 619-630.
  8. Korea Centers for Disease Control and Prevention (2016). The Seventh Korea National Health and Nutrition Examination Survey (KNHANES VII-1).
  9. Korean Statistical Information Service (KOSIS). Census, Statistic Korea, Republic of Korea. Accessed December 2018. Available from:
  10. Kshirsagar, A. V., Chiu, Y. L., Bomback, A. S., August, P. A., Viera, A. J., Colindres, R. E., and Bang, H. (2010). A hypertension risk score for middle-aged and older adults, The Journal of Clinical Hypertension, 12, 800-808.
  11. Lee, K. M., Kim, W. J., and Yun, S. J. (2009). A clinical nomogram construction method using genetic algorithm and naive Bayesian technique, Journal of Korean Institute of Intelligent Systems, 19, 796-801.
  12. Mozina, M., Demsar, J., Kattan, M., and Zupan, B. (2004). Nomogram for visualization of Naive Bayesian classifier, Knowledge Discovery in Databases: PKDD 2004, 337-348.
  13. Nam, H. R., Pak, S. B., Jung, S. J., Choi, I. Y., and Kim, Y. (2018). Interdependency of Risk Factors for Hypertension: the 2010-2015 Korean National Health and Nutrition Examination Survey, Korean Journal of Family Practice, 8, 372-379.
  14. Park, J. C., Kim, M. H., and Lee, J. Y. (2018). Nomogram comparison conducted by logistic regression and naive Bayesian classifier using type 2 diabetes mellitus, The Korean Journal of Applied Statistics, 31, 573-585.
  15. Rao, J. N. K. and Scott, A. J. (1981). The analysis of categorical data from complex sample surveys: chi-squared tests for goodness of fit the independence in two-way tables, Journal of the American Statistical Association, 76, 221-230.
  16. Shin, J., Park, J. B., Kim, K. I., Kim, J. H., Yang, D. H., Pyun, W. B., Kim, Y. G., Kim, G. H., and Chae, S. C. (2015). 2013 Korean Society of Hypertension guidelines for the management of hypertension: part I-epidemiology and diagnosis of hypertension, Clinical Hypertension, 21, 1.
  17. Statistics Korea (2018). Causes of death statistics 2017. Policy News. Available from: kor nw/3/index.board?bmode=read&bSeq=&aSeq=370711&pageNo=1&rowNum=10&navCount=10&currPg=&sTarget=title&sTxt=2017
  18. Sung, N. K. (2012). Sampling Methodologies (2nd ed), Freedom academy, Seoul.
  19. Van den Berg, E., Kloppenborg, R. P., Kessels, R. P., Kappelle, L. J., and Biessels, G. J. (2009). Type 2 diabetes mellitus, hypertension, dyslipidemia and obesity: a systematic comparison of their impact on cognition, Biochimica et Biophysica Acta - Molecular Basis of Disease, 1792, 470-481.