The results from this study of two large, nationally representative data sets provide evidence for a relation between domestic violence and malnutrition among women and children in India. These findings persisted, even after adjusting for geographic variation and a number of demographic characteristics, including multiple domains of socioeconomic status, that are strong confounders of domestic violence and chronic malnourishment. These findings also exhibited a dose-response relation, with frequent and recent abuse found to be associated with the highest probability of poor nutritional outcomes, particularly among women. To put our findings in context, a review of studies investigating the effect of administration of intestinal anthelmintic drugs on hemoglobin found that mass treatment of communities for intestinal parasites decreases the prevalence of anemia by 1.1–12.4 percent in adults (55
), indicating that preventing domestic violence could be just as effective as this pharmaceutical approach in combating anemia among women.
We offer two explanations for the positive association between exposure to domestic violence and malnutrition among women and children. The first relates to empowerment. Perpetrators of domestic violence often use several types of abuse, such as physical and psychological, to control behavior of their family members (56
). The withholding of food is a documented form of abuse in Indian households and is likely correlated with the perpetration of physical violence (20
). An inadequate diet resulting from this withholding of food through psychological or emotional abuse could mediate the relation between physical domestic violence and nutrient deficiencies that cause anemia and underweight. Additionally, domestic violence is strongly associated with a woman's inability to make decisions for herself and her family, including the choice of types and quantities of food that a woman prepares as she cares for herself and her children (36
). This might also explain why no significant associations were found between maternal report of domestic violence and nutritional outcomes in children less than 12 months of age, since a majority of these children in India are breastfed (57
), and thus their feeding is more closely controlled by their mothers. Although a measure of decision-making autonomy was included in our models, this measure of autonomy does not have a temporal component. That is, it may be that many women who reported participation in family decision making exercise this power at times but only at the will of their husbands, and that this participation in decision making can be curtailed as a form of abuse.
The second explanation is that the link between domestic violence and nutritional deficiencies may also involve a mediating effect of psychological stress. Women who experience domestic violence tend to have higher levels of psychological stress (58
)—a relation that has been documented among south Asian women (12
). Children who have witnessed domestic violence in their homes are also more likely to experience psychological stress (14
There are physiological reasons why psychological stress may be associated with anemia. Psychological stress is a risk factor for oxidative stress (15
), a term describing a number of chemical reactions that produce free radicals and other organic molecules capable of damaging living tissue (62
). This oxidative stress can destroy red blood cells prematurely, acting as a potential cause of hemolytic anemia (63
). Additionally, although acute stressful incidents temporarily increase blood levels of both hemoglobin and erythrocytes, chronic stress has been found to result in long-term reductions, suggesting that stress interferes with protein synthesis required to create new red blood cells (65
Psychological stress could also be linked to underweight status. Most research in food-rich environments in industrialized nations has documented a link between chronic stress and obesity mediated by metabolic changes (66
) or behavioral adjustments (67
) that promote the depositing of abdominal adipose. However, chronic stress also increases metabolic rate and energy expenditure (19
), a process that could cause weight loss in a person with a fixed and limited caloric intake. That psychological stress may be associated with underweight in a low-income country is consistent with the results of a longitudinal study from the United Kingdom finding that, under stressful conditions, lean subjects tend to lose weight while overweight subjects tend to gain weight (70
We note several important caveats to the findings presented here. First, although psychological, emotional, and sexual abuse constitute a substantial proportion of all domestic violence incidents (56
), this information was not available in the current study. Our results present the relation between malnutrition and physical domestic violence, but it could be that other forms of abuse have different associations. Another caveat involves the measurement of domestic violence with a single global question. Previous research has found that behaviorally specific questions asking, for example, whether a woman has been slapped, punched, or kicked result in higher rates of reported domestic violence than the global question used in this survey (73
). A local study from Karnataka state also found that over half of all women would keep quiet about it if abused (74
). In addition, our analysis investigated only physical domestic violence and could not address other forms of abuse. These circumstances may account for the fact that the prevalence of domestic violence reported in this sample is lower than that previously found in India (12
). However, by adjusting our models for a number of important socioeconomic and demographic variables including wealth, education, occupation, religion, and caste, we worked to limit the influence of bias introduced by social patterning of reliability of domestic violence reports on our results. A third caveat relates to the cross-sectional nature of the data. Although we attempted to address the idea of causal direction as best we could with the temporal reporting of domestic violence, it still was not possible to distinguish whether any reported abuse preceded attainment of the current nutritional status. Longitudinal research is called for to untangle this question.
In conclusion, domestic violence adversely impacts numerous health outcomes in India, including gynecologic morbidity (75
), contraceptive use (77
), asthma (78
), tobacco use (79
), attempted suicide (13
), and early child mortality (80
). To our knowledge, this is the first large, nationally representative study using objective measures of malnutrition and measurement of multiple covariates to show a robust association between domestic violence and the likelihood of malnutrition among women and children. Additional adverse health effects of domestic violence suggest that public policy interventions to reduce domestic violence not only are fundamentally important from a moral and intrinsic perspective (81
) but that there is a clear case to focus on reducing domestic violence given the instrumental health benefits that are likely to accrue.