Severely malnourished children with a low weight-for-height have a higher mortality than those with a low mid-upper-arm-circumference: I. Empirical data demonstrates Simpson’s paradox

Table 3 Significance levels of comparisons between diagnostic groups using Marascuilo post-hoc analysis procedure

Children with marasmus
Comparisons	All	IPF	OTP	SFC
0–60 m	p	p	p	p
muac v whz	0.000	ns	0.001	ns
muac v both	0.000	0.000	0.000	0.000
whz v both	0.000	0.000	0.000	0.000
6 - < 18 m
muac v whz	0.000	ns	ns	ns
muac v both	0.000	0.000	0.000	0.003
whz v both	0.000	0.017	0.000	ns
18 - < 36 m
muac v whz	0.000	ns	0.000	ns
muac v both	0.000	0.001	0.000	ns
whz v both	0.000	0.000	0.000	ns
36–60 m
muac v whz	ns	ns	ns	ns
muac v both	0.000	0.000	ns	ns
whz v both	0.000	0.017	0.019	ns
Children with kwashiorKor/marasmic-kwashiorKor
Comparisons	All	IPF	OTP
0–60 m	p	p	p
K-muac v Kwash	ns	ns	ns
K-whz v Kwash	0.000	0.000	0.000
K-both v Kwash	0.000	0.000	0.000
K-muac v K-whz	0.000	0.001	0.000
K-muac v K-both	0.000	0.000	0.000
K-whz v K-both	ns	ns	0.041
6 - < 18 m
K-muac v Kwash	ns	ns	ns
K-whz v Kwash	0.014	ns	ns
K-both v Kwash	0.000	0.000	ns
K-muac v K-whz	0.007	ns	0.006
K-muac v K-both	0.000	0.002	0.000
K-whz v K-both	ns	ns	ns
18 - < 36 m
K-muac v Kwash	ns	ns	ns
K-whz v Kwash	0.000	0.003	0.000
K-both v Kwash	0.000	0.000	0.000
K-muac v K-whz	0.000	0.012	0.001
K-muac v K-both	0.000	0.000	0.002
K-whz v K-both	ns	ns	ns
36–60 m
K-muac v Kwash	0.047	ns	ns
K-whz v Kwash	0.000	0.011	0.000
K-both v Kwash	0.000	0.000	0.000
K-muac v K-whz	0.007	ns	0.002
K-muac v K-both	0.000	0.002	ns
K-whz v K-both	ns	ns	ns

ISSN: 1475-2891