Research That Matters (January 17 - 20, 2008)

Background and Purpose Although the relationships between IPV and women's health have been extensively studied in the United States, studies in countries other than the United States remain limited. In addition, few studies in the U.S. accounted for exposure to childhood abuse or abuse perpetrated by non-intimates or relevant socio-demographic characteristics. Accounting for these variables, this study in Japan investigated the relationship between the experience of partners' emotional, physical and sexual abuse and health status among women in Japan.

Methods As part of a cross-national research effort coordinated by the World Health Organization, data were collected through face-to-face interviews, augmented by paper-and-pencil questionnaires. The latter was included to avoid vocalization of sensitive questions and respondents' answers due to small housing quarters in most Japanese homes. The sample was a stratified cluster sample of 1,371 women aged 18 to 49 in Yokohama, Japan. The mean age of the respondents was 35.2 years, and the majority were married and had one or more children.

We compared the health status across three groups of respondents: those who had not experienced intimate partner violence (IPV), those who had experienced partners' emotional abuse only (EA only), and those who had experienced partners' emotional abuse as well as physical and/or sexual violence (EA+PSV). Using logistic or negative binomial regression models, we estimated odds ratio, as well as adjusted odds ratio, controlling for relevant characteristics known to be associated with health status.

Results In most indicators examined, the odds of experiencing health-related problems were significantly higher for those who experienced EA+PSV compared to those with no IPV, controlling for age, relationship status, number of children, education, childhood abuse, adulthood abuse by non-intimates and socioeconomic well-being at interview. In most areas, there were no statistically significant differences between the EA only group and the EA+PSV group. Emotional abuse alone was associated with poorer self-rated health at interview, greater difficulty walking, greater difficulty performing usual activities, greater use of health services in the past month, lifetime suicidal ideation, and greater distress symptoms during the past month. Areas where the EA+PSV group fared significantly worse than the EA only group were difficulty with memory or concentration, use of sleeping pills/tranquilizer, the number of distress symptoms, and suicidal ideation.

Conclusions Given the wide range of health consequences of intimate partner violence that this study highlighted, in addition to emergency departments and surgery clinics, other units such as primary care have opportunities to reach out abused women. The increased reporting of functional limitations and distress symptoms among abused women, especially those who have experienced EA only, is notable. Typically, these types of symptoms are difficult to pair with a medical diagnosis and are categorized as “non-specific symptoms.” As a result, health care providers may dismiss distress that women are experiencing, or worse, label women as hypochondriac. Professional training and health care organizations' protocols must address how to respond when female patients present non-specific symptoms. In addition, the current Japanese legal definition that does not include emotional abuse must be expanded.