Comparing Frequentist and Bayesian Quantile Regression Models for Child Hypertension in South Africa
DOI:
https://doi.org/10.6000/1929-6029.2025.14.66Keywords:
Bayesian quantile regression, frequentist quantile regression, paediatric hypertension, South Africa, extreme valuesAbstract
Background: Traditional approaches to modelling paediatric hypertension in South Africa have relied on descriptive or mean regression methods, which inadequately capture risk factors driving the distributional extremes of blood pressure. Quantile regression provides a flexible alternative, and Bayesian methods offer advantages in precision and uncertainty estimation, yet their comparative performance has not been assessed in this context.
Methods: Nationally representative cross-sectional data from 1,812 adolescents (15–17 years) in the South African National Income Dynamics Study (NIDS) Wave 5 (2017–2018) was analysed. Frequentist and Bayesian quantile regression models were fitted for systolic (SBP) and diastolic blood pressure (DBP) at the 75th and 95th percentiles. Model performance was compared in terms of parameter estimates, interval precision, and convergence diagnostics.
Results: BMI and gender were consistent predictors of both SBP and DBP across models. Bayesian quantile regression additionally identified age, race, and pulse rate as significant risk factors for upper quantiles. Bayesian credible intervals were consistently narrower than frequentist confidence intervals, indicating improved precision. Convergence diagnostics confirmed robust posterior inference.
Conclusion: Bayesian quantile regression provides more efficient inference than the frequentist alternative when modelling health outcomes concentrated in distributional extremes. This is the first study to apply Bayesian quantile regression to paediatric hypertension in South Africa, demonstrating both methodological value and empirical insights into adolescent health risks.
References
Gomwe H, Seekoe E, Lyoka P, Marange CS. Blood pressure profile of primary school children in Eastern Cape province, South Africa: prevalence and risk factors. BMC Pediatr 2022; 22: 207. DOI: https://doi.org/10.1186/s12887-022-03221-5
Nkeh-Chungag BN, Sekokotla AM, Sewani-Rusike C, Namugowa A, Iputo JE. Prevalence of Hypertension and Pre-hypertension in 13-17 Year Old Adolescents Living in Mthatha - South Africa: a Cross-Sectional Study. Cent Eur J Public Health 2015; 23: 59-64. DOI: https://doi.org/10.21101/cejph.a3922
Kagura J, Adair LS, Musa MG, Pettifor JM, Norris SA. Blood pressure tracking in urban black South African children: birth to twenty cohort. BMC Pediatr 2015; 15: 78. DOI: https://doi.org/10.1186/s12887-015-0402-z
Joubert N, Walter C, Du Randt R, Aerts A, Adams L, Degen J, et al. Hypertension among South African children in disadvantaged areas and associations with physical activity, fitness, and cardiovascular risk markers: A cross-sectional study. J Sports Sci 2021; 39: 2454-67. DOI: https://doi.org/10.1080/02640414.2021.1939964
Koenker R, Hallock KF. Quantile Regression An Introduction. J Econ Perspect 2001; 15: 143-56. DOI: https://doi.org/10.1257/jep.15.4.143
Benoit DF, Van den Poel D. A Bayesian Approach to Quantile Regression. J Stat Softw 2017; 76. DOI: https://doi.org/10.18637/jss.v076.i07
Yu K, Moyeed RA. Bayesian quantile regression. Stat Probab Lett 2001; 54: 437-47. DOI: https://doi.org/10.1016/S0167-7152(01)00124-9
National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and Adolescents. The Fourth Report on the Diagnosis, Evaluation, and Treatment of High Blood Pressure in Children and Adolescents. Pediatrics 2004; 114: 555-76. DOI: https://doi.org/10.1542/peds.114.2.S2.555
Hu Y, Wang HJ, He X, Guo J. Bayesian joint-quantile regression. Comput Stat 2021; 36: 2033-53. DOI: https://doi.org/10.1007/s00180-020-00998-w
Soomro S, Yu K. Bayesian fractional polynomial approach to quantile regression and variable selection with application in the analysis of blood pressure among US adults. J Appl Stat 2024: 1-22. DOI: https://doi.org/10.1080/02664763.2024.2359526
Barrodale I, Roberts FDK. An Improved Algorithm for Discrete $l_1 $ Linear Approximation. SIAM J Numer Anal 1973; 10: 839-48. DOI: https://doi.org/10.1137/0710069
Yu K, Zhang J. A Three-Parameter Asymmetric Laplace Distribution and Its Extension. Commun Stat - Theory Methods 2005; 34: 1867-79. DOI: https://doi.org/10.1080/03610920500199018
Koenker R, Chernozhukov V, He X, Peng L, editors. Handbook of Quantile Regression. 1st ed. Chapman and Hall/CRC; 2017. DOI: https://doi.org/10.1201/9781315120256
Martin AD, Quinn KM, Park JH. MCMCpack: Markov Chain Monte Carlo in R. J Stat Softw 2011; 42. DOI: https://doi.org/10.18637/jss.v042.i09
Hamra G, MacLehose R, Richardson D. Markov Chain Monte Carlo: An introduction for epidemiologists. Int J Epidemiol 2013; 42: 627-34. DOI: https://doi.org/10.1093/ije/dyt043
Plummer M, Best N, Cowles K, Vines K, Sarkar D, Bates D, et al. Output Analysis and Diagnostics for MCMC 2016.
Fenske N, Kneib T, Hothorn T. Identifying Risk Factors for Severe Childhood Malnutrition by Boosting Additive Quantile Regression. J Am Stat Assoc 2011; 106: 494-510. DOI: https://doi.org/10.1198/jasa.2011.ap09272
Sinharay S. Assessing Convergence of the Markov Chain Monte Carlo Algorithms: A Review. Princeton, NJ: Educational Testing Service; 2003. DOI: https://doi.org/10.1002/j.2333-8504.2003.tb01899.x
Rea LM, Parker RA. Designing and conducting survey research: a comprehensive guide. Fourth edition. San Francisco, CA: Jossey-Bass; 2014.
Lesaffre E, Lawson A. Bayesian Biostatistics. Chichester: Wiley; 2012. DOI: https://doi.org/10.1002/9781119942412
Gelman A, Rubin DB. Inference from Iterative Simulation Using Multiple Sequences. Stat Sci 1992; 7: 457-72. DOI: https://doi.org/10.1214/ss/1177011136
Matizirofa L, Kuhudzai AG. Predictors of High Blood Pressure in South African Children: Quantile Regression Approach. Int J Stat Med Res 2017; 6: 84-91. DOI: https://doi.org/10.6000/1929-6029.2017.06.02.4
Raphadu TT, Staden MV, Dibakwane WM, Monyeki KD. A Non-Invasive Investigation into the Prevalence of Higher than Normal Blood Pressure, Hypertension and the Association between Blood Pressure and Body Weight in Male and Female Adolescents in the Polokwane Local Municipality, Limpopo-South Africa: A Cross-Sectional Study. Children 2020; 7: 18. DOI: https://doi.org/10.3390/children7030018
Christofaro DGD, Casonatto J, Vanderlei LCM, Cucato GG, Dias RMR. Relationship between Resting Heart Rate, Blood Pressure and Pulse Pressure in Adolescents. Arq Bras Cardiol 2017. DOI: https://doi.org/10.5935/abc.20170050
Muntner P. Trends in Blood Pressure Among Children and Adolescents. JAMA 2004; 291: 2107. DOI: https://doi.org/10.1001/jama.291.17.2107
Chen J, Wang Y, Li W, Zhang Y, Cao R, Peng X, et al. Physical activity and eating behaviors patterns associated with high blood pressure among Chinese children and adolescents. BMC Public Health 2023; 23: 1516. DOI: https://doi.org/10.1186/s12889-023-16331-1
Levy RV, Brathwaite KE, Sarathy H, Reidy K, Kaskel FJ, Melamed ML. Analysis of Active and Passive Tobacco Exposures and Blood Pressure in US Children and Adolescents. JAMA Netw Open 2021; 4: e2037936. DOI: https://doi.org/10.1001/jamanetworkopen.2020.37936
Olive LS, Abhayaratna WP, Byrne D, Telford RM, Berk M, Telford RD. Depression, stress and vascular function from childhood to adolescence: A longitudinal investigation. Gen Hosp Psychiatry 2020; 62: 6-12. DOI: https://doi.org/10.1016/j.genhosppsych.2019.10.001
Juhan N, Zubairi YZ, Mohd Khalid Z, Mahmood Zuhdi AS. A Comparison Between Bayesian and Frequentist Approach in the Analysis of Risk Factors for Female Cardiovascular Disease Patients in Malaysia. ASM Sci J 2020: 1-7. DOI: https://doi.org/10.32802/asmscj.2020.sm26(1.1)
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Policy for Journals/Articles with Open Access
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are permitted and encouraged to post links to their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work
Policy for Journals / Manuscript with Paid Access
Authors who publish with this journal agree to the following terms:
- Publisher retain copyright .
- Authors are permitted and encouraged to post links to their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work .
