Contextual and individual factors associated with sexual abuse in girls: the WHO multi-country study on women's health and domestic violence against women

• Grant number: 13/05024-5
• Support Opportunities: Scholarships abroad - Research
• Start date: July 15, 2013
• End date: July 14, 2014
• Field of knowledge: Health Sciences - Collective Health - Public Health
• Principal Investigator: Ivan França Junior
• Grantee: Ivan França Junior
• Host Investigator: Charlotte Watts
• Host Institution: Faculdade de Saúde Pública (FSP). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Institution abroad: London School of Hygiene and Tropical Medicine, England

Abstract

Sexual abuse of girls during childhood is a highly prevalent social phenomenon. However, its determinants are largely unknown. The WHO-VAW study - a worldwide effort to document many forms of violence against women using standardized concepts, methods and procedures - is an opportunity to expand our knowledge regarding the causes of childhood sexual abuse. To date, the study has been performed at 19 sites in 13 countries, ensuring representative samples of women from 15 to 49 years old. Here, our objective is to estimate the contextual and individual factors that are associated with childhood sexual abuse in girls. The World Health Organization has granted access to data from the 13 countries. The model to be tested will include 4 levels of causation: ontogenic, microsystem (family), exosystem (community), and macrosystem (societal). As part of this approach, a multi-level analysis will be performed, with three operational levels: individual, site and country. Individual variables will include respondent's age; age of abuse; nobody to trust in the community; being socially isolated; attitudes towards gender roles, and family characteristics such as age difference between the child and the offender; being exposed to conflicts between parents; religion; family violence history; socioeconomic status. Contextual variables will include at site level community violence indicators; urban/rural-urban site; services related to CSA: religious beliefs: women's movement activities, and, at country level, gender inequalities; human development; history of war (civil or between countries); adherence to human rights treaties; legal restrictions of CSA; country violence indicators; child labor, and women's movement development. At the individual level, multiple Poisson regression models will be used to determine whether the individual factors influence the occurrence of CSA. For the multivariate analysis, factors with p<0.20 identified in univariate analyses will be included in a Poisson regression model. A variable will be kept in the individual final model when associated with the outcome or when having a significant confounding effect on other variables (i.e., a change in adjusted risk measure over 10%). After performing individual and partial analyses, we will implement multiple multilevel models. Initially, we will analyze the variables at the individual level (level 1), followed by the contextual variables (levels 2 and 3). (AU)