Skip to main content
Open Access Publications from the University of California


UCLA Electronic Theses and Dissertations bannerUCLA

The Role of Health Behaviors and Socioeconomic Status in Explaining the Relationship Between Child Abuse and Cancer


Recent research has linked experiences of child abuse to cancer later in life. However much of the available research has failed to look at the independent effects of child abuse types (i.e. physical, sexual and emotional abuse) and does not attempt to test potential reasons for this association. To address these shortcomings, the purpose of this study is to investigate the contribution of three health behaviors (i.e. smoking, overweight or obesity and alcohol drinking) and two measures of socioeconomic status (i.e. educational attainment and household income) that are hypothesized to mediate the association between abuse type experienced in childhood and cancer risk in adulthood. Overall, evidence exists supporting some of the connections. However, not all evidence examines all abuse types and not all evidence is consistent. Data from the 2011 Behavioral Risk Factor Surveillance Survey (BRFSS), a representative telephone survey of adults 18 years of age and over living in the United States was used in this investigation.

The following four aims were examined: 1) Assess the representativeness of data from the BRFSS; 2) Determine if each abuse type was independently associated with cancer, after controlling for other abuse types, other adverse childhood experiences and sociodemographic characteristics; 3) Determine if the relationship between child abuse and cancer is mediated by health behaviors 4) Determine if the relationship between child abuse and cancer is mediated by socioeconomic status. Aim 1 was examined using 1 proportion z-tests, comparing BRFSS estimates by state to comparison data obtained from national sources. Aim 2 was examined using logistic regression. Aims 3 and 4 were examined using logistic regression, with Karlson, Holm and Breen’s KHB method used to analyze mediating effects of health behaviors and socioeconomic status. Analyses were conducted using Stata 14.0 and using probability weights.

Results for Aim 1 revealed that data was not entirely representative. For Aim 2, models showed that physical, sexual and emotional abuse were each associated with increased odds of cancer. For Aim 3, mediation analyses revealed that smoking partially mediated the relationship between physical abuse and cancer and between sexual abuse and cancer. Smoking fully mediated the association between emotional abuse and cancer. For Aim 4, mediation analyses revealed that household income partially mediated the relationship between sexual abuse and cancer.

Overall, this study highlighted that abuse is different from other adverse childhood experiences when it comes to its association with cancer. Furthermore, abuse types have unique mediators, suggesting that there are unique causal pathways connecting each abuse type and cancer.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View