Domestic abuse is common in India, but varies widely by region. In a study conducted in five districts of Uttar Pradesh, 18-45% of husbands reported physically abusing their wives.1 Of those who acknowledged being physically abusive, more than four in 10 reported an episode of violence during the prior year and more than six in 10 admitted repeated abuse. Men who had little education, those who had more than one child and those who were extremely poor were more likely than other men to have physically abused their wives. In separate analyses of these data examining relationships between wife abuse and male reproductive health, the prevalence of abuse was significantly higher among men who had had extramarital or premarital sex, those who had ever had a sexually transmitted disease (STD) and those whose wives had had an unplanned pregnancy than it was among other men.2

Data on spousal abuse were collected in 1995-1996 as part of a survey on male reproductive health that included 6,695 married men aged 15-65 in five districts of Uttar Pradesh, one of the least developed states in India. Men who reported abusing their wives were asked about the type, frequency and extent of the abuse. Those who reported that they had sexually abused their wives were asked whether they had had nonconsensual sex and whether they had physically forced their wives to have sex. They were also asked whether they had ever abused their wives while the women were pregnant, what kinds of behaviors their wives exhibited while being abused, and whether their wives had ever sought medical treatment for injuries resulting from abuse.

The interview also included questions about men's reproductive health and behavior. The men were asked if they had had any symptoms of STDs at three points in their lives (before marriage, at any time after marriage and at the time of the interview). Men were classified as having symptoms of an STD if they reported having any of eight indicators, such as discharge from the penis, genital or anal sores or painful urination. The men were also asked if they had had premarital or extramarital sex, whether they and their wives were practicing contraception at the time of the study, and whether they had ever experienced an unplanned pregnancy.

Finally, the interviewers collected data about social and demographic factors such as age, caste, educational level, the age at which a husband first started living with his wife, the number of children a couple had, household composition, urban-rural residence, the duration of the current marriage, and poverty level. Poverty was measured by the number of modern possessions a family owned; a family owning fewer than two was considered extremely poor.

Overall, 1,990 men (30%) reported physically abusing their wives. The level and type of wife abuse fluctuated across the five districts (Aligarh, Bandha, Gonda, Kanpur Nagar and Nainital). The proportion of men reporting physical abuse ranged from 18% in Nainital to 45% in Bandha. Sexual abuse followed a similar pattern: Men in Nainital were least likely to say they had had nonconsensual sex with their wives (18%), while those in Bandha were most likely to do so (40%). Men were much less likely to say they had physically forced their wives to have sex (4-9% across the districts). In all five districts, men who reported physically abusing their wives were significantly more likely to report nonconsensual sex (odd ratios of 2.1-3.0) and forced sex (2.3-5.8).

Most of the men who said that they had physically abused their wives reported multiple episodes (63-91% across districts), and large proportions said that they had physically abused their wives within the past year (47-74%). Smaller percentages (5-13%) reported abusive behavior while their wives were pregnant. Men were more likely to report shouting or yelling at their wives (33-94%) and slapping or pushing their wives (47-77%) than punching or kicking them (8-32%) or using a weapon or object against their wives (5-10%).

The men were most likely to say that their wives had responded to the last episode of abuse by crying (49-90% across districts) or by shouting and yelling back (7-42%). Smaller proportions of men said that their wives had run away from home (4-10%) or had physically retaliated (0-6%). No more than 3% reported that their wives had sought medical treatment after being abused.

Bivariate analysis showed that in four of the five districts, men with five years of education or less were significantly more likely to abuse their wives than were men with more education (odds ratios of 2.1-3.3), while extreme poverty was associated with abuse in three districts (1.5-1.8). In two districts, men who had been married for more than five years were more likely than other men to physically abuse their wives (1.5-3.1), as were those who had more than one child (1.5-2.1). Living with one's wife before the age of 20 was associated with abuse in one district (1.8).

In the examination of relationships between domestic violence and reproductive and behavioral health variables, the men were divided into four groups--no abuse, physical abuse only, sexual abuse without physical force (nonconsensual sex) and sexual abuse with physical force.* Significant bivariate associations were found between abuse and reproductive health and behavior variables. For example, men who had had premarital sex were more likely than those who had not to have abused their wives in some way (67% vs. 41%), as were those who had had extramarital sex (78% vs. 44%). Men who reported STD symptoms were also more likely than those who did not to report abuse, whether they were asked about symptoms before marriage (60% vs. 43%), since marriage (62% vs. 43%) or at the time they were interviewed (64% vs. 43%). Abuse was more common among men who did not practice contraception than among those who did (49% vs. 42%), and was reported more frequently by men who had experienced an unplanned pregnancy than by those who had not (52% vs. 44%). For each variable, the strongest association was with nonconsensual sex, followed by physical abuse and then by forced sex. In addition, significant differences in the prevalence of abuse varied significantly by district of residence, rural-urban residence, caste, socioeconomic status, education and age.

Logistic regression analysis generally confirmed the bivariate findings. In four districts, men with five years of education or less were more likely than other men to abuse their wives when the effects of all other variables were accounted for (odds ratios of 1.9-2.8). Extreme poverty and wife abuse were significantly associated in two districts (1.4-1.5). Men in one district had significantly elevated odds of wife abuse if they had lived with their wives at a young age (1.6), had more than one child (1.3) or had been married for more than five years (2.3). Measures of privacy, such as a household size of fewer than five people, the husband's parents not being present in the household and urban residence, were not significantly associated with domestic violence in any district.

The significant associations between wife abuse and the reproductive health and behavior variables persisted when the effects of social and demographic factors were controlled for in a logistic regression analysis. Elevated odds of all types of abuse were associated with premarital sex, extramarital sex, STD symptoms and unplanned pregnancy (Table 1). The associations were weakest for physical abuse alone, intermediate for nonconsensual sex and strongest for forced sex. There was a small but significant positive association between contraceptive use and sexual abuse without physical force.



Table 1. Odds ratios from logistic regression analyses predicting the risk of domestic abuse, by reproductive health variable and type of abuse
Variable Physical abuse Nonconsensual sex Forced sex
Sex outside marriage
Premarital 2.21* 3.03* 3.63*
Extramarital 2.72* 4.31* 6.22*
STD symptoms
Before marriage 2.03* 2.32* 2.19*
Since marriage 2.16* 2.23* 3.08*
Current 1.77* 2.17* 2.43*
Current contraceptive use 1.08 1.21* 0.91
Unplanned pregnancy 1.43* 1.67* 2.62*
*p<.05. Note: For each category, the reference group is men who reported no abuse.


According to the investigators, the extent of wife abuse reported in this study is consistent with the results of previous international research. They conclude that "the findings that family stressors, especially low education levels and poverty, are strong risk factors for wife abuse suggest that broad social changes aimed at bettering the Indian population's access to advanced education and employment opportunities could lead to improvements in many aspects of family life, includingthe prevention or reduction of family violence."--L. Gerstein


1. Martin SL et al., Domestic violence in northern India, American Journal of Epidemiology, 1999, 150(4):417-426.

2. Martin SL et al., Sexual behaviors and reproductive health outcomes: associations with wife abuse in India, Journal of the American Medical Association, 1999, 282(20):1967-1972.

*About 40-50% of the last group also reported physical abuse.