Baseline characteristics of the study participants (final sample)
| Total | QPM only | QPM+targeting | P value | |
| Mean (SD) | Mean (SD) | Mean (SD) | ||
| Household characteristics | ||||
| No of household members* | 6.2 (2.0) | 6.2 (2.0) | 6.2 (2.1) | 0.99 |
| High quality roof*† | 57.3% (49.5%) | 60.7% (48.9%) | 53.3% (50.0%) | 0.09 |
| Land owned (timad)*† | 5.6 (6.3) | 5.6 (6.6) | 5.5 (6.0) | 0.64 |
| Mazie produced (kg)*† | 1869.4 (2391.0) | 1719.3 (2342.6) | 2042.4 (2438.6) | 0.45 |
| Caregiver characteristics | ||||
| Age (years)* | 28.5 (5.8) | 28.5 (5.7) | 28.4 (6.0) | 0.82 |
| Attended school* | 33.3% (47.2%) | 29.6% (45.7%) | 37.6% (48.5%) | 0.07 |
| No of pregnancies* | 4.4 (2.3) | 4.5 (2.1) | 4.4 (2.4) | 0.79 |
| Index child characteristics | ||||
| Age (months)* | 19.9 (8.4) | 19.1 (7.9) | 20.8 (8.9) | 0.02 |
| Male* | 51.9% (50.0%) | 55.9% (49.7%) | 47.4% (50.0%) | 0.06 |
| Height-for-age (z-score)* | −1.4 (1.4) | −1.5 (1.4) | −1.2 (1.5) | 0.04 |
| Weight-for-age (z-score)* | −1.0 (1.1) | −1.1 (1.1) | −0.9 (1.2) | 0.12 |
| Health and health seeking behaviour | ||||
| Index child sick with diarrhoea in past 2 weeks | 17.4% (37.9%) | 19.1% (39.4%) | 15.4% (36.2%) | 0.32 |
| Index child sick with fever in past 2 weeks | 18.9% (39.2%) | 20.1% (40.1%) | 17.7% (38.2%) | 0.29 |
| No times caregiver sought antenatal care during pregnancy with index child | 3.2 (1.6) | 3.2 (1.7) | 3.2 (1.5) | 0.93 |
| Cooking and feeding | ||||
| Days in past week cooked specifically for young children | 1.8 (2.3) | 1.8 (2.2) | 1.8 (2.3) | 0.94 |
| Days in past week cooked something with maize | 5.7 (2.5) | 5.7 (2.5) | 5.7 (2.4) | 0.98 |
| Days in past week index child ate food with QPM | 0.1 (0.8) | 0.1 (0.7) | 0.2 (0.9) | 0.57 |
| Days in past week index child ate porridge | 0.8 (1.2) | 0.8 (1.2) | 0.8 (1.2) | 0.43 |
| Worried not enough food because not enough money, in last 3 months | 37.5% (48.5%) | 39.8% (49.0%) | 35.0% (47.8%) | 0.21 |
| Attrition | 5.0% (21.8%) | 8.3% (27.6%) | 0.16 | |
| Joint test of orthogonality F-statistic | 1.9 | 0.01 | ||
| N households | 570 | 304 | 266 | |
| N clusters | 369 | 195 | 174 |
The final sample is defined as all households where the caregiver survey was conducted at baseline, follow-up 1 and follow-up 2. P values are derived from a regression of the outcome on an indicator for the QPM+targeting group, controlling for kebele (strata) and clustered at the community health group level. Child height-for-age and weight-for-age z-scores are normalised using the 2006 WHO growth standards. The joint test of orthogonality is a test of the null hypothesis that the coefficients on all characteristics in the table are jointly equal to 0, where the outcome is an indicator for the QPM+targeting group.
*Baseline characteristics included in the regression as covariates.
†Rows are measured through the household head survey, QPM only n=303, QPM+targeting n=261.
QPM, quality protein maize.