RESUMO
BACKGROUND: Child marriage is a breach of the basic fundamental rights of a child and is a major concern for policy makers in many countries. OBJECTIVE: The present study intended to identify the direct and indirect socio-demographic factors associated with child marriage among 15-17 years old girls in India. PARTICIPANTS AND SETTING: The study used National Family Health Survey-4 data. The sample comprised of 1459 married girls and an equal number of unmarried girls aged 15-17 years at the time of the survey. METHODS: Multivariable logistic regression was performed to identify the factors associated with child marriage. A hypothetical path model was constructed and tested. The path models were refitted after removing the insignificant path coefficients, and finally, the significance of direct as well as indirect effects were obtained. RESULTS: Path analysis revealed that mother's education was the only factor that had both significant direct and indirect effects on child marriage (p < 0.001). Religion had only significant indirect effect (p = 0.047) through girls' education. Besides mother's education, the factors that showed significant direct effects on child marriage were wealth index (p = 0.012), caste (p = 0.009) and education of the girl (p = 0.009). CONCLUSIONS: Education of both mothers and the girls, wealth index and caste had significant direct effects on child marriage whereas significant indirect effect was exerted by mothers' education and religion. Therefore, these factors should be given primary focus while formulating strategies to curb child marriage in countries like India.