Table 3.

Regression models for the effect of an agriculture–nutrition intervention on anthropometric indices (z‐scores) of Ghanaian rural children, unadjusted and adjusted for covariatesa

	Length‐for‐age/Height‐for‐age z‐score			Weight‐for‐age		Weight‐for‐length/Weight‐for‐height
	Model 1		Model 2	Model 1	Model 2	Model 1	Model 2
Group assignment
Control (reference)
Intervention	0.14 (0.11)		0.22 (0.06)**	0.09 (0.12)	0.15 (0.07)*	0.04 (0.13)	0.07 (0.08)
Baseline value of outcome			0.59 (0.03)***		0.61 (0.03)***		0.55 (0.04)***
Baseline age, month			0.02 (0.01)*		0.03 (0.01)*		0.05 (0.01)**
Sex
Male (reference)
Female					−0.19 (0.04)**		−0.21 (0.06)**
Maternal ethnicityb
Krobo (reference)
Non‐Krobo			−0.18 (0.05)**		−0.15 (0.06)*
Household food securityc
Food secure (reference)
Mild insecurity			−0.02 (0.09)
Moderate insecurity			−0.01 (0.09)
Severe insecurity			0.12 (0.14)
Reared chickens at baseline							0.14 (0.08)†
Constant	−1.41 (0.06)***		−1.16 (0.11)***	−0.99 (0.08)***	−0.83 (0.15) ***	−0.36 (0.10)**	−0.64 (0.15)**
R 2	0.005		0.48	0.002	0.51	0.0005	0.44
Sample n	415		408	416	411	415	408

Note. Values are beta coefficients (standard errors).

^aThis is an intention‐to‐treat analysis with logistic regression models with standard errors adjusted, accounting for clustering. For Model 2, models initially included phase of enrolment and covariates for child (baseline age, sex, baseline value of the outcome, and time elapsed between measurements), mother (education, marital status, and ethnicity), and household (food security, wealth, and raised poultry prior to project). Backward elimination stepwise covariate selection procedure was used; the models retained covariates with a P value of <0.10 for the overall significance for the variable (not individual categories). No interaction terms with intervention were significant.

^bKrobo: the local ethnic group; others: Akan, Ewe, Ga, among others.

^cFood security: classification based on the 15‐item Food Insecurity Experience Scale (Food and Agriculture Organization, 2012).

^† P < 0.10.

^* P < 0.05.

^** P < 0.01.

^*** P < 0.001.