Factors associated with stunting among children according to the level of food insecurity in the household: a cross-sectional study in a rural community of Southeastern Kenya

Table 3 Adjusted odd ratios for child stunting among separate child groups by household food insecurity level using multiple logistic regression

1) Severe food insecure group (N = 252)			2) Non-severe food insecure group (N = 108)
Variables	Adjusted OR	95% CI	Variables	Adjusted OR	95% CI
Tea/porridge with milk in previous 24 h			Animal rearing
No	Ref.		No	Ref.
Yes	3.22	(1.43 - 7.25)	Yes	3.24	(1.04 - 10.07)
Non-milk liquids			Caregiver’s education status
No	Ref.		Not educated/Preschool	Ref.
Yes	0.50	(0.22 - 1.16)	Primary/Secondary	0.44	(0.16 - 1.26)
Jigger infection			Socioeconomic status (SES)
No	Ref.		Poorest	Ref.
Yes	1.84	(0.88 - 3.84)	Second	0.14	(0.03 - 0.73)
			Third	0.22	(0.04 - 1.08)
Child age (mo)			Fourth	0.13	(0.03 - 0.55)
0-5	Ref.
6-11	2.38	(0.48 - 11.92)	Siblings of preschool age
12-23	1.89	(0.47 - 7.62)	0 or 1	Ref.
24-35	4.04	(1.01 - 16.14)	2 or more	2.81	(0.92 - 8.58)
36-47	3.16	(0.85 - 11.79)
48-59	1.59	(0.36 - 6.94)
Socioeconomic status (SES)
Poorest	Ref.
Second	0.71	(0.33 - 1.53)
Third	0.80	(0.35 - 1.85)
Fourth	0.33	(0.13 - 0.85)

Note:
OR, Odds Ratio; 95% CI, 95% Confidence Interval; HFIAS, Household Food Insecurity Access Scale.
For multivariate logistic regression, all variables listed in Table 1 were used and selected by backward stepwise selection with 0.2 of significant level of removal from the model. All the selected variables were used for calculating adjusted odds ratios (adjusted ORs).
For non-severe-food insecure group, 33 children aged less than 12 months were removed from the analysis because only 1 stunting condition was found in these categories.

ISSN: 1471-2458