Development of growth chart for Malaysian children

Yusoff, Ahmad Faudzi and HuiQi, Pan and Mustafa, Amal Nasir and Cheong, Kee Chee and Md Iderus, Nuur Hafizah and Mohd Ghazali, Sumarni and Zainuddin, Ahmad Ali and Abd Samad, Hazizi and Ab Rahman, Jamalludin and Raib, Junidah and Zailani, Mohd Hanif and Poh, Bee Koon and Selamat, Rusidah and Shahar, Suzana and Wai Sew, Teh (2015) Development of growth chart for Malaysian children. Technical Report. Institute of Medical Research, Ministry of Health Malaysia, Kuala Lumpur.

PDF (Technical report) - Published Version Restricted to Repository staff only Download (1MB) | Request a copy

Abstract

Growth charts are internationally used as a tool for assessment of physical growth which reflects the nutritional status of children. It is used in public health for screening for malnutrition, and for monitoring children’s growth patterns. Poor physical growth is closely related to poor health status. Currently most countries are using WHO Child Growth Standards 2006 for children less than 5 years old and WHO Growth Reference 2007 for school-aged children and adolescents. Some countries have developed growth references for their population as an additional reference besides the WHO’s using nationally-representative data and revising them periodically. Differences in the socio-economic and health environment of each country may result in differences in the growth potential of its children. Therefore there is a need to develop current, country-specific growth references for children that can be used in public health screening for malnutrition. This report present the first growth chart developed for Malaysian children (length/height-for-age, weight-for-age, body mass index-for-age) and describes the methodological processes involved.The Malaysian Children Growth Chart (MyGC) was developed by using the nationally representative data from the Third National Health and Morbidity (NHMS III) conducted in 2006. The NHMS III was population based cross-sectional study using two-stage stratified sampling proportionate to population size throughout Malaysia. The weight and length/height measurements of all apparently healthy children (11,177 boys and 10,855 girls) age 0 to 18 years in selected households were taken.The data were screened for extreme values and outliers (biological implausible) and any extreme values and outliers were removed based on the recommendations of the WHO leaving a final sample of 10,454 boys and 10,259 girls. The LMS ChartMaker Pro software was used to derive age-related reference centiles and z-scores for the anthropometric data. This method is based on the assumption that anthropometric data can be converted to a standard normal distribution by a Box-Cox transformation for any given age. It summarizes the age-changing distribution by 3 curves, namely L (Box-Cox Power) which measures skewness (λ); M, the median at each age (μ); and S, the coefficient of variation by age (σ). Using penalised likelihood, the three curves are fitted as cubic splines by non-linear regression, and the extent of smoothing required is expressed in terms of smoothing parameters or equivalent degrees of freedom (edf). The optimal model was obtained by balancing smoothness of the curves (e.d.f) and the model goodness of fit (Q test of fit and detrended Q-Q plot). The sex-specific percentiles and z-score curves for, weight-for-age, length/height-for-age and BMI-for-age were generated. Weight-for-age. The weight observations of 5722 boys and 5550 girls aged 0 to 10 years were used in the final construction of the growth charts. We produced sex-specific weight-for-age percentile chart which comprised of 3rd, 15th, 50th, 85th and 97th percentile curves. For z-score charts, -3SD, - 2SD, -1SD, Median, 1SD and 2SD curves (+3SD z-scores for age 9 to 10 years could not be produced by the software therefore +3SD curve were not shown). Length/height-for-age. For both boys and girls birth to 2 years, recumbent length measurements were used to construct length-for-age percentile and z-score curves while standing height measurements were used for age 2 to 18 years old. The length measurements of 1018 boys and 981 girls aged 0 to 24 month were used to construct length-for-age percentile and z-score curves. For 24-216 months, height data of 9124 boys and 9083 girls were used. We present percentile and z-score curves for age ranges of birth to 2 years, 2 to 5 years and 5 to 18 years. BMI-for-age. BMI curves for birth to 2 years were constructed using length measurements, for 2-18 years using height measurements. For birth to 2 years, there were 1022 boys and 995 girls records with both weight and length and BMI observations. After data cleaning, BMI for 1018 boys and 995 girls records were used to generate BMI-for-age percentile and z-score curves. For 2 to 18 years, there were 9415 boys and 9225 girls records with both weight and length and BMI observations. After data cleaning, BMI for 9234 boys and 9070 girls records were used to generate BMI-for-age percentile and z-score curves. We present percentile and z-score curves for age ranges of birth to 2 years, 2 to 5 years and 5 to 18 years. Comparison of weight-for-age percentile between MyGC and WHO. From birth to 5 years, MyGC curves for boys and girls aged 0 to 5years were lower than WHO Growth Standard for all percentiles. From 5 to 10 years, the curves for 3rd, 15th, 50th and 85th percentiles for the MyGC were lower than the WHO corresponding percentiles. MyGC 97th percentile crossed the WHO 97th percentile at the age of between 5 to 6 years for boys and 7 years for girls. Therefore above these ages, if the MyGC cut-off point for obesity was used, children will be considerably less likely to be classified as obese. Comparison of length/height-for-age percentile between MyGC and WHO. All MyGC percentile curves were below their respective WHO curves except for median boys and girls below 1 year and the 97th percentile for boys age 2 to 5 years. Estimates of prevalence stunting (<3rd percentile) will be lower if MyGC is used compared to using WHO references. Comparison of BMI-for-age percentile between MyGC and WHO. All MyGC percentile curves were below their respective WHO curves except for the 97th percentile for boys and girls all ages and 85th percentile for boys and girls aged 5 to 18 years, MyGC percentile curve was above WHO growth reference. There is no apparent difference between MyGC 85th percentile curve for boys and girls 2 to 5 years and the corresponding WHO curve. In conclusion, there are differences between MyGC and WHO Child Growth Standards and WHO Growth Reference for children. WHO Child Growth Standards/Reference are likely to overdiagnose obesity, thinness/underweight and stunting for most age groups as compared to MyGC. Thus health practitioners who are using WHO Child Growth Standards/Reference should be aware of this possibility and exercise caution when assessing the children physical growth. MyGC is representative of the existing growth pattern of Malaysia children; therefore it can be used by nutritionists, dieticians, nurses and paediatricians and public health practitioners as an additional reference for screening and early management of malnutrition and for research purposes.

Item Type:	Monograph (Technical Report)
Additional Information:	3297/49395
Subjects:	R Medicine > RA Public aspects of medicine > RA0421 Public health. Hygiene. Preventive Medicine
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button):	Kulliyyah of Medicine > Department of Community Medicine (Effective: 1st January 2011) Kulliyyah of Medicine
Depositing User:	Dr Jamalludin Ab Rahman
Date Deposited:	18 May 2020 13:17
Last Modified:	18 May 2020 13:17
URI:	http://irep.iium.edu.my/id/eprint/49395

Actions (login required)

View Item