Differences in predicted stunting levels for poorest and richest quartile
| Age group (months) | Predicted absolute difference (in percentage points of stunting) | Attenuation effect relative to difference in reference model | |||
| Reference model | Nutrition-specific only | Full model | Nutrition-specific only (%) | Full model (%) | |
| 0 to less than 5 | −5.5 (−7.2 to 3.8) | −4.7 (−6.4 to 2.9) | −2.8 (−4.6 to 1.0) | 15.5 | 49.5 |
| 5 to less than 10 | −9.7 (−11.4 to 8.1) | −8.3 (−10.0 to 6.6) | −5.4 (−7.2 to 3.6) | 14.6 | 44.6 |
| 10 to less than 15 | −13.0 (−14.9 to 11.1) | −10.6 (−12.6 to 8.6) | −7.0 (−9.1 to 5.0) | 18.6 | 46.0 |
| 15 to less than 20 | −16.4 (−18.4 to 14.3) | −12.5 (−14.7 to 10.4) | −8.3 (−10.6 to 6.1) | 23.5 | 49.1 |
| 20 to less than 25 | −19.3 (−21.6 to 17.0) | −14.5 (−16.9 to 12.1) | −9.1 (−11.7 to 6.6) | 24.8 | 52.7 |
| 25 to less than 30 | −20.1 (−22.5 to 17.7) | −16.7 (−19.2 to 14.1) | −11.9 (−14.6 to 9.2) | 17.0 | 40.8 |
| 30 to less than 35 | −24.3 (−27.0 to 21.7) | −19.4 (−22.2 to 16.6) | −14.1 (−17.0 to 11.2) | 20.3 | 42.1 |
| 35 to less than 40 | −24.7 (−27.6 to 21.8) | −20.1 (−23.1 to 17.0) | −14.7 (−17.9 to 11.4) | 18.7 | 40.6 |
| 40 to less than 45 | −24.1 (−27.1 to 21.0) | −18.7 (−22.0 to 15.4) | −13.4 (−16.9 to 10.0) | 22.3 | 44.2 |
| 45 to less than 50 | −19.9 (−23.2 to 16.6) | −13.9 (−17.4 to 10.4) | −8.1 (−11.8 to 4.4) | 30.1 | 59.2 |
| 50 to less than 55 | −21.6 (−25.1 to 18.0) | −16.5 (−20.3 to 12.8) | −12.8 (−16.8 to 8.9) | 23.4 | 40.6 |
| 55 to 59 | −23.4 (−27.5 to 19.2) | −18.3 (−22.7 to 13.9) | −13.1 (−17.7 to 8.4) | 21.9 | 44.1 |
Predicted absolute differences depicted with 95% CI in parentheses. The reference model is only adjusted for survey-level fixed effects, household location as well as the basic child-level and maternal-level characteristics mentioned in the Methods section, while the full model additionally includes the full range of determinants of undernutrition (all variables interacted with child age group).