Home > Journals > Minerva Pediatrica > Past Issues > Articles online first > Minerva Pediatrica 2020 Jun 05



To subscribe
Submit an article
Recommend to your librarian


Publication history
Cite this article as



Minerva Pediatrica 2020 Jun 05

DOI: 10.23736/S0026-4946.20.05745-X


language: English

Establishing body mass index growth charts for Pakistani children and adolescents using the Lambda-Mu-Sigma (LMS) and quantile regression method

Muhammad ASIF 1 , Muhammad ASLAM 2, Justyna WYSZYŃSKA 3, Saima ALTAF 2

1 Department of Statistics, Govt. Degree College, Multan, Pakistan; 2 Department of Statistics, Bahauddin Zakariya University, Multan, Pakistan; 3 Medical College of Rzeszów University, Rzeszów, Poland


BACKGROUND: Assessment of growth and nutritional status is an essential part of clinical evaluation and care in pediatrics. We therefore, aimed to establish age and gender specific smoothed BMI growth reference charts of Pakistani children and adolescents aged 2-18 years and to compare our LMS median percentile values with WHO 2007 international references and with references from other foreign studies.
METHODS: A representative cross-sectional sample of 10,668 healthy Pakistani subjects aged 2 to 18 years was studied. For calculation of BMI (kg/m2), height (cm) and weight (kg) were measured using standard procedures. Age and gender specific smoothed BMI growth reference values and associated charts were obtained using lambda-mu-sigma (LMS) as well as quantile regression (QR) statistical methods.
RESULTS: In the sample studied, the mean (± SD) BMI of all subjects was 16.50 (± 2.83 kg/m2). Smoothed BMI percentile curves (5th, 10th, 25th, 50th, 75th, 90th and 95th) obtained by using LMS and QR methods, showed that BMI increased with age in both sexes. During pubertal age, girls had larger percentiles than boys. Centile values estimated by the LMS and QR procedure had small variability from the empirical centiles. Comparison of LMS BMI 50th percentile with references from WHO 2007 and data from other countries demonstrated that Pakistani children had substantially lower BMI percentiles than their counterparts in the reference population.
CONCLUSIONS: This comprehensive study suggests that WHO 2007 references are not suitable for Pakistani children. The QR method should be considered as an alternative method to develop growth charts.

KEY WORDS: Body mass index; Growth charts; Lambda-mu-sigma and quantile regression method; Pakistani children

top of page