If a would-be mother is subjected to violence by an intimate partner — spouse, boyfriend or girlfriend — the chances of her baby dying as an infant or a child are higher than women with violence-free pregnancies.

A population-based survey on pregnant Indian women and their exposure to intimate partner violence (IPV) concludes that the relationship between exposure to household IPV and child death1 could be attributable to the mother's inability to care for her child, psychological stress associated with witnessing violence, and the use of maternal violence victimisation as a proxy for child violence victimisation.

Researchers Leland Ackerson from University of Massachusetts and S. V. Subramanian at the School of Public Health, Harvard University, Boston collected information on 39, 096 children from the National Family Health Survey of India, 2005-06.

They analysed maternal reports for any evidence of any kind of violence — physical, sexual, psychological and IPV. They keenly studied the deaths of infants (0 to 12 months), older children (12 to 60 months) and other children (0 to 60 months) in this population.

"Maternal experience of physical IPV was associated with increased mortality rates among all these category of children," co-researcher Subramanian says.

The stress from witnessing IPV tends to produce potentially harmful physiologic reactions, such as atypical cortisol production patterns, in children. Research from developing countries has found that child exposure to family violence is associated with increased levels of poor nutritional outcomes, respiratory infections, diarrhea and asthma.

Sexual and psychological IPV were less strongly associated with child death. Also, this association had nothing to do with the child's gender, he says. Close to two million Indian children die each year — an annual mortality rate of 76 deaths per 1000. Intimate partner violence (IPV) could affect the child's health by affecting the mother's physical or psychological health. A physically or psychologically unstable mother is unable to provide proper care to the child.

Earlier research among community-specific samples in India had shown that IPV was related to neonatal and infant deaths. The nonphysical types of abuse had not been investigated vis-à-vis IPV. "Pathways through which physical and non-physical types of abuse harm health may be substantially different and the relationship between abuse and child health may vary according to the age of the child," Subramanian says.

For a better understanding of these relationships, the duo took up the study to see whether physical, psychological, and sexual forms of IPV were associated with infant and child death in India.

The researchers suggest that comes workers in the fields of medicine, public health, public policy and public security should act to eliminate IPV to improve health outcomes and to reduce social health disparities.