Table 4.
Sample size calculations to assess intervention impact on primary and secondary outcomes
| Variable | Detectable difference | βa | Unadjusted sample size per treatment group | Attrition | Adjusted sample size per treatment groupb | Total sample size, four treatment groups |
|---|---|---|---|---|---|---|
| Primary outcomes | ||||||
| Length and length-for-age-Z-score | 0.20 SD | 0.10 | 710 | 15% | 850 | 3400 |
| Weight and weight-for-age-Z-score | 0.20 SD | 0.10 | 710 | 15% | 850 | 3400 |
| Incidence of diarrhea | 0.20 SD | 0.10 | 710 | 15% | 850 | 3400 |
| Biochemical indicators of micronutrient statusc | 0.40 SD | 0.20 | 133 | 30% | 190 | 760 |
| Indicators of innate and adaptive immune response | 0.50 SD | 0.20 | 88 | 30% | 125 | 500 |
| Secondary outcomes | ||||||
| Intestinal parasite burden | 0.30 SD | 0.20 | 244 | 20% | 305 | 1220 |
| Indicators of environmental enteric dysfunction | 0.35 SD | 0.20 | 179 | 20% | 225 | 900 |
| Hair cortisol | 0.40 SD | 0.20 | 133 | 30% | 190 | 760 |
aPower used for sample size calculation with a significance of α = 0.05
bAdjusted for attrition and rounded up
cHemoglobin concentration was used as a screening criteria for study enrollment; thus the hemoglobin concentration of all children enrolled in the study was assessed and sample size calculations were not completed for this primary outcome