Dear Colleagues,
I am attempting to analyze historical MUAC data using the https://www.who.int/tools/child-growth-standards/software , but the software is prompting me with the message: “You need to provide a mapping for variable 'Weight (kg)' or 'Length or height (cm)'.”
I have been looking for the WHO MUAC-for-age standards and found the relevant tables for girls acfa_girls_3_5_zscores.pdf (who.int) and for boys acfa_boys_3_5_zscores.pdf (who.int) However, I wasn’t able to locate the specific information I need from these tables:
- WHO Median MUAC: The median MUAC from the WHO reference table for each given age.
- WHO SD: The standard deviation for MUAC for each given age.
My main queries are:
- How to read median MUAC and standard deviation for each given age from these tables.
- How to analyze MUAC-for-age data using WHO softwares (WHO anthro survey analyser and/or WHO anthro software).
Thank you for your help
Dear Selam,
There are a couple of previous posts on en.net that may be able to provide some guidance on your queries (although I cannot find direct reference to the points you raise).
https://www.en-net.org/forum/question/3549 (Discussion re MUAC for age calculator)
https://www.en-net.org/forum/question/5120 and https://www.en-net.org/forum/question/3841 (Discussion re using excel for z-score calculations)
Hope these might help,
Lucy
Answered:
30 days agoDear Salem:
Thank you for the very good questions. I will attempt to address the several issues in the order in which the are raised:
- Generally, I try to avoid one-size-fits-all cookbook survey data analysing toolkits. These toolkits include WHO Anthro Survey Analyser. Analysing nutrition survey data requires epidemiologic and statistical skills in order to prevent mistakes leading to erroneous conclusions. These toolkits make it too easy for unskilled and inexperienced persons to push one button and get standardized output which may be incorrect or not fit the specific research questions or population addressed in a particular survey. Therefore, I have no idea why the WHO Anthro Survey Analyser is asking for weight and length when trying to analyse MUAC data.
- Regarding the MUAC for age tables, column 9 is labelled 'median'. This gives the age- and sex-specific median MUAC.
Columns 8 and 10 give the median minus 1 standard deviation and the median plus 1 standard deviation. Just subtract the number in column 9 from the number in column 10 to get the standard deviation of MUAC measurements for children of the indicated age in the standard population. However, note that sometimes the standard deviation derived from column 10 minus column 9 is different from the standard deviation derived from column 8 minus column 9. This is one of the idiosyncrasies of the WHO standard data. If I need a rough estimate of the standard deviation for an individual age and sex group, I use the average of these 2 numbers. However, because the LMS method of calculating anthropometric z-scores is quite complicated, I always use one of the routines written for standard software packages which are offered by WHO. These can be downloaded at: https://www.who.int/tools/child-growth-standards/software.
I hope this answers your questions.
Answered:
22 days ago