Following is a response from DHS staff member, Tom Pullum:
These variables are test results, not responses to questions. I agree with you that it is confusing to have the same label for more than one variable. In this case, the two variables differ in that sb267 is for women and sb367 is for men. Here is how I confirmed my suspicion that that was the case (hv104=1 is males and hv104=2 is females):
summarize sb*67 if hv104==1
Variable | Obs Mean Std. Dev. Min Max
-------------+---------------------------------------------------------
sb267 | 0
sb367 | 6,553 143.1547 188.5172 48 996
. summarize sb*67 if hv104==2
Variable | Obs Mean Std. Dev. Min Max
-------------+---------------------------------------------------------
sb267 | 7,998 133.1017 164.0002 35 996
This distinction extends to all of the sb2* and sb3* variables in this survey.
I don't know of a manual that covers this kind of a pattern, but it's common for biomarkers. Please ask if there are others that you have questions about.