converting BMI, age & sex into BMISDS

[Kevin Marks]

My dataset includes an ID number, observation date, weight, height, BMI, sex, and age variables for children adolescents and young adults. For my analysis, I must use a variable for BMISDS (or BMI standard deviation score or a BMI z score) for those under 18 years old. The BMISDS also incorporates information about the child's age and sex and is only used for those under 18 years old. In my dataset, n= 14,910 and the age range is 2.9 to 30 years with n = 6325 under 18.0 years old. I have been searching for downloadable software to convert BMI into BMISDS for those under 18 years old. In Stata, is there a way to converts the variables of BMI, age and sex into BMISDS?

[Chen Samulsion]

Jef Leroy's zscore06 can do some work, but it only caculates BMI z-score for children age from 0 to 5. (If you examine the code in zscore06.ado, you will find it's very complicated.)

search zscore06

zscore06 calculates anthropometric z-scores using the 2006 WHO child growth standards. Lenght/heigh-for-age, weight-for-height, BMI-for-age and weight-for-age Z-scores are calculated for children 0 to 5 years of age. If the age, height or weight values are outside the range of reference values, a value of 99 is given for the corresponding z-score(s). Note: in contrast with the WHO's Igrowup program, none of the z-scores are calculated if child age is missing.

Another choice is igrowup developed by the Department of Nutrition and Food Safety (there're also igrowup macro in SPSS and SAS form), see details in github:
https://github.com/unicef-drp/igrowup_update, and below is its description:

The macro (igrowup_standard.ado) calculates z-scores for the eight anthropometric indicators, weight-for-age, length/height-for-age, weight-forlength/height, body mass index (BMI)-for-age, head circumference-for-age, arm circumference-for-age, triceps skinfold-for-age and subscapular skinfold-for-age based on the WHO Child Growth Standards. In this macro, all available (nonmissing and non-flagged) z-score values are used for each indicator-specific prevalence estimation (standard analysis).

And here's some caution:https://www.statalist.org/forums/for...468066-igrowup