The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Kenya » Height, Weight and BMI measurements
Height, Weight and BMI measurements [message #20520] Fri, 13 November 2020 09:03 Go to next message is currently offline
Messages: 3
Registered: November 2020

I'm using variables for women's Weight(v437), Height(v438) and BMI(v445).

1) It is claimed that these variables prior to Phase IV only included mothers of children under three/five years of age, while in Phase IV forward all interviewed women were included. However, the recode manual for all phases classifies these measurements under "Maternal Anthropometry".
So in which phases are all women included, and in which ones only mothers?

2) The recode manual states Weight and Height must be divided by 10, whilst BMI must be divided by 100. What is the explanation for having unrealistic values for these measures after dividing? For example, how should we interpret a weight of 8kg and 573kg for a woman or a BMI of 70? Are these data errors?


[Updated on: Fri, 13 November 2020 09:04]

Report message to a moderator

Re: Height, Weight and BMI measurements [message #20521 is a reply to message #20520] Fri, 13 November 2020 09:50 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2861
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

Prior to Phase IV, measurements of height and weight for children and women were part of the interviews of women. Afterward, and now, these measurements are part of the household survey and include all children under five and all women 15-49, regardless of any relationship between the children and the women other than that they are in the same household.

The KR file includes children born in the past five years to women who are eligible for the women's interview and are interviewed. The mother's information I s attached to the child's information. Therefore, in this file, you have mother-child pairs. You can look at the correspondence between the anthropometry for a living child in the household and a living mother in the household. In this context, the label "maternal anthropometry" would be accurate. This is just because the KR file links the living mothers and children in the household.

If you want anthropometry for all children, regardless of whether their mother is alive and in the household, you use the PR file. If you want anthropometry for all women age 15-49, regardless of whether they have any children or those children are alive and in the household, you use the PR file (or the IR file). To repeat, what's special about the KR file is that it links children and mothers.

Some heights and weights in some files are clearly out-of-range and some are implausible but not impossible. The most extreme values will lead to codes of 9999 (or similar) for the Z scores. The formula for the BMI may be applied without exclusion rules--you can check the Guide to Statistics. If users want to adjust or exclude extreme values, they are free to do so. It is DHS policy to do a minimum of adjustment. A major effort to avoid measurement or recording error for height and weight has been made during the past few years.
Previous Topic: Causes of maternal mortality
Next Topic: Missing wealth index in Kenya DHS
Goto Forum:

Current Time: Thu Sep 21 22:21:33 Coordinated Universal Time 2023