Vest Joshua R, Catlin Tegan K, Chen John J, Brownson Ross C
Saint Louis University School of Public Health, Department of Community Health, St. Louis, Missouri 63104-1314, USA.
Am J Prev Med. 2002 Apr;22(3):156-64. doi: 10.1016/s0749-3797(01)00431-7.
Reports on prevalence estimates and risk factors of intimate partner violence (IPV) are limited in that they (1) focus on specific subgroup populations that are not representative of all women or (2) involve long questionnaires that are not useful as surveillance tools.
To report prevalence estimates and identify demographic and lifestyle factors associated with IPV in a large population-based sample of U.S. women using surveillance data.
Behavioral Risk Factor Surveillance System (BRFSS) data from eight U.S. states were analyzed individually and as a pooled sample (N=18,415). Multivariate logistic regression models were used to examine associations between IPV and the factors of interest.
Factors consistently associated with IPV across the majority of states and in the pooled analysis included young age (pooled adjusted odds ratio [aOR], 3.07), single marital status (pooled aOR, 2.89), divorced/separated marital status (pooled aOR, 4.67), and annual household income <$25,000 (pooled aOR, 1.89). In addition, lack of health insurance, receipt of Medicaid, cigarette smoking, presence of children in the home, self-reported fair/poor health, and frequent mental distress were associated with IPV after adjustment for covariates.
This study provides population-based estimates of IPV prevalence and factors associated with IPV using surveillance data. By pooling BRFSS data from individual states, the resulting large sample has adequate power to detect significant associations and has increased precision in the estimates of IPV risk. In addition, this study identifies high-risk populations to target for education and intervention programs and demonstrates the need for improved IPV surveillance.
关于亲密伴侣暴力（IPV）患病率估计和风险因素的报告存在局限性，因为它们（1）关注的是不具有所有女性代表性的特定亚群体，或者（2）涉及冗长的问卷，这些问卷作为监测工具并无用处。
利用监测数据报告美国大量基于人群的女性样本中IPV的患病率估计，并确定与IPV相关的人口统计学和生活方式因素。
对来自美国八个州的行为风险因素监测系统（BRFSS）数据进行单独分析，并作为汇总样本（N = 18,415）进行分析。使用多变量逻辑回归模型来检验IPV与感兴趣因素之间的关联。
在大多数州以及汇总分析中，与IPV始终相关的因素包括年轻（汇总调整优势比[aOR]，3.07）、单身婚姻状况（汇总aOR，2.89）、离婚/分居婚姻状况（汇总aOR，4.67）以及家庭年收入<$25,000（汇总aOR，1.89）。此外，在对协变量进行调整后，缺乏医疗保险、接受医疗补助、吸烟、家中有孩子、自我报告健康状况一般/较差以及频繁的精神困扰与IPV相关。
本研究利用监测数据提供了基于人群的IPV患病率估计以及与IPV相关的因素。通过汇总各个州的BRFSS数据，所得的大样本有足够的能力检测显著关联，并提高了IPV风险估计的精度。此外，本研究确定了教育和干预项目的高风险目标人群，并表明需要改进IPV监测。
