Skip to main content
An official website of the United States government
Grant Details

Grant Number: 5R03CA184478-02 Interpret this number
Primary Investigator: Colabianchi, Natalie
Organization: University Of Michigan At Ann Arbor
Project Title: Impact of Public Housing Assistance on Modifiable Cancer Risk Factors in Adults
Fiscal Year: 2015


Abstract

DESCRIPTION (provided by applicant): Despite improvements in cancer mortality in the past two decades, significant disparities persist among racial, ethnic, and socioeconomic status (SES) subpopulations. Housing has been identified as an important social determinant for health disparities in general and cancer-related disparities specifically. Public housing (PH) assistance aims to improve housing affordability and quality for the lowest-SES households. Thus, PH represents an intervention to improve housing as a social determinant of health. However, few studies have examined the effects of living in PH on health-related outcomes. Additionally, several researchers have proposed pathways through which PH influences health but have not tested them empirically. The proposed study will use data collected from 1999-2013 in the Panel Study of Income Dynamics (PSID) to estimate the effect of living in PH on four leading modifiable risk factors for cancer (smoking, alcohol use, physical inactivity, and overweight/obesity) in adults <62 years of age. The PSID offers a unique opportunity to examine the effects of PH because it is a nationally representative, longitudinal survey with a wide range of demographic, household, housing, economic, employment, social, and health variables; neighborhood-level variables will also be linked using geocodes. The treatment group will include only respondents who move into PH during the study period (expected n=550). The comparison group will be limited to renters who are income-eligible for PH (expected n=1650). We will apply propensity score methods to the rich set of baseline characteristics available in the PSID to help control for selection bias inherent in observational data. Inverse probability of treatment weights will be calculated from estimated propensity scores and used in logistic regression models to estimate average treatment effects for each risk factor at 2, 4, and 6 years after baseline. Multiple mediation models will be developed to examine the extent to which PH influences modifiable cancer risk factors through individual-level mediators (e.g., non-housing income, stress, residential stability) and neighborhood-level mediators (e.g., neighborhood SES, access to supermarkets, walkability). Study results will be used to help clarify whether living in PH improves or worsens modifiable risk factors for cancer. Importantly, information obtained about pathways through which these effects occur can inform future housing and public health research, practice, and policy.



Publications


None

Back to Top