Height-specific Blood Pressure Cutoffs for Screening Elevated and High Blood Pressure in Children and Adolescents: an International Study

Yaping Hou,Pascal Bovet,Roya Kelishadi,Mieczysław Litwin,Anuradha Khadilkar,Young Mi Hong,Tadeusz Nawarycz,Barbara Stawińska-Witoszyńska,Hajer Aounallah-Skhiri,Mohammad Esmaeil Motlagh,Hae Soon Kim,Vaman Khadilkar,Alicja Krzyżaniak,Habiba Ben Romdhane,Ramin Heshmat,Shashi Chiplonkar,Małgorzata Krzywińska-Wiewiorowska,Jalila El Ati,Mostafa Qorbani,Neha Kajale,Pierre Traissac,Lidia Ostrowska-Nawarycz,Gelayol Ardalan,Lavanya Parthasarathy,Liu Yang,Min Zhao,Arnaud Chiolero,Bo Xi
DOI: https://doi.org/10.1038/s41440-018-0178-2
2018-01-01
Hypertension Research
Abstract:Pediatric blood pressure (BP) reference tables are generally based on sex, age, and height and tend to be cumbersome to use in routine clinical practice. In this study, we aimed to develop a new, height-specific simple BP table according to the international child BP reference table based on sex, age and height and to evaluate its performance using international data. We validated the simple table in a derivation cohort that included 58,899 children and adolescents aged 6–17 years from surveys in 7 countries (China, India, Iran, Korea, Poland, Tunisia, and the United States) and in a validation cohort that included 70,072 participants from three other surveys (China, Poland and Seychelles). The BP cutoff values for the simple table were calculated for eight height categories for both the 90th (“elevated BP”) and 95th (“high BP”) percentiles of BP. The simple table had a high performance to predict high BP compared to the reference table, with high values (boys/girls) of area under the curve (0.94/0.91), sensitivity (88.5%/82.9%), specificity (99.3%/99.7%), positive predictive values (93.9%/97.3%), and negative predictive values (98.5%/97.8%) in the pooled data from 10 studies. The simple table performed similarly well for predicting elevated BP. A simple table based on height only predicts elevated BP and high BP in children and adolescents nearly as well as the international table based on sex, age, and height. This has important implications for simplifying the detection of pediatric high BP in clinical practice.
What problem does this paper attempt to address?