The DHS Program User Forum: Biomarkers » NFHS 5 BMI in PR File of India

Home » Topics » Biomarkers » NFHS 5 BMI in PR File of India

Show: Today's Messages :: Show Polls :: Message Navigator

NFHS 5 BMI in PR File of India [message #27421]

Thu, 10 August 2023 12:28

Bandana
Messages: 3
Registered: July 2023

Member

Dear DHS Forum,
I want to know to the meaning of '.a' in the score of BMI of household members in the PR file of NFHS 5 data of India. I have seen the DHS recode manual but it was not there. Please let me know the answer.

Thanking you,
Regards
Bandana Mondal

Report message to a moderator

Re: NFHS 5 BMI in PR File of India [message #27448 is a reply to message #27421]

Tue, 15 August 2023 12:01

Janet-DHS
Messages: 888
Registered: April 2022

Senior Member

Following is a response from DHS staff member, Tom Pullum:

I am looking at this file (IAPR7EFL.dta) and don't see what you are referring to. The only BMI variable is hc73, which refers to children. The label (HC73) includes the following special codes:

. label list HC73
HC73:
9996 height out of plausible limits
9997 age in days out of plausible limits
9998 flagged cases

I don't see ".a". Please be more specific.

Report message to a moderator

Re: NFHS 5 BMI in PR File of India [message #27453 is a reply to message #27448]

Tue, 15 August 2023 12:28

Bandana
Messages: 3
Registered: July 2023

Member

In the PR file the body mass index of women are ha40 and body mass index of men are hb40. The observations of these two variables are lying between 1200 to around 5000. But some observations are given as '.a'. So my question is what the implication of '.a' for body mass index?

Report message to a moderator

Re: NFHS 5 BMI in PR File of India [message #27483 is a reply to message #27453]

Mon, 21 August 2023 12:58

Janet-DHS
Messages: 888
Registered: April 2022

Senior Member

Following is a response from DHS staff member, Tom Pullum:

I now see ha40 and hb40 but I do not see any non-numeric values such as ".a". Non-numeric values should not be possible, given the format of these variables. You can generate the bmi score for women using this Stata command: "gen bmi=(10^7)*ha2/(ha3^2) if ha2<9994 & ha3<9994". Similarly for men, using hb2 and hb3.

Report message to a moderator

Re: NFHS 5 BMI in PR File of India [message #27498 is a reply to message #27483]

Tue, 22 August 2023 23:03

Bandana
Messages: 3
Registered: July 2023

Member

Thank you for your suggestion. I shall definitely try it out.

Report message to a moderator

Re: NFHS 5 BMI in PR File of India [message #27712 is a reply to message #27421]

Mon, 25 September 2023 03:01

Riesterscare
Messages: 1
Registered: September 2023

Member

Bandana wrote on Thu, 10 August 2023 12:28

Dear DHS Forum,
I want to know to the meaning of '.a' in the score of BMI of household members in the PR file of NFHS 5 data of India. I have seen the DHS recode manual but it was not there. Please let me know the answer.

Thanking you,
Regards
Bandana Mondal

;)

Report message to a moderator

Re: NFHS 5 BMI in PR File of India [message #29120 is a reply to message #27483]

Thu, 25 April 2024 13:49

AkhilK28
Messages: 19
Registered: February 2022

Member

Hi Tom,

When working with male and female BMI data in the PR data for NFHS-5, there are observations coded with ".a"

Running this code in the PR file: br ha40 if ha40==.a results in 54,874 observations

I was wondering the difference between ".a", ".", and "flagged" observations in BMI for men and women.

All three types are classified as missing (collectively being ~7% for Women and ~15% for Men) but are there any specific meanings behind these?

Best,

Report message to a moderator

Re: NFHS 5 BMI in PR File of India [message #29151 is a reply to message #29120]

Tue, 30 April 2024 16:42

Janet-DHS
Messages: 888
Registered: April 2022

Senior Member

Following is a response from DHS staff member, Tom Pullum:

Yes, there are cases, for some variables, not just ha40, with the code ".a" . This is just some kind of coding error and can be treated as equivalent to "." or NA. You can identify them, recode them to NA, and then confirm the recode, with the following three lines in Stata:

tab ha40 if ha40>=.,m

replace ha40=. if ha40>.

tab ha40 if ha40>=.,m

Report message to a moderator

Previous Topic:	BMI of males
Next Topic:	New calculation of the percentage of children with anemia

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Fri Nov 22 14:42:07 Coordinated Universal Time 2024