Prevalence ratios of stunting and severe stunting estimated from multivariable Poisson regression models using contraceptive use
| Prevalence ratio of stunting (95% CI) | Prevalence ratio of severe stunting (95% CI) | |
| Current contraceptive use | * | *** |
| No use | (Reference) | (Reference) |
| Traditional method | 0.93 (0.86 to 1.02) | 0.84 (0.69 to 1.03) |
| Modern method, <15 months | 1.01 (0.94 to 1.10) | 0.93 (0.79 to 1.09) |
| Modern method, ≥15 months | 0.87*** (0.81 to 0.94) | 0.61*** (0.50 to 0.73) |
| Prior contraceptive use | ** | *** |
| No use | (Reference) | (Reference) |
| Traditional method | 0.89* (0.81 to 0.98) | 0.71** (0.59 to 0.87) |
| Modern method | 0.93* (0.87 to 0.98) | 0.79** (0.69 to 0.92) |
Asterisks not associated with estimates reflect the overall p value of the variable. The same prespecified confounding variables were included in all models: age of child, age of mother, wealth index, child sex, area of residence, maternal and partner education attainment, maternal literacy, maternal marital status, region of country, ethnic group, language, and presence of diarrhoea in the last 2 weeks. Estimates account for sampling design.
*P<0.05, **P<0.01, ***P<0.001.