Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Purpose A large body of literature has examined the relationship between health and socioeconomic status (SES) measured by income and wealth (Alwin & Wray, 2005; Lyons & Yilmazer, 2005; Ram, 2005; Wu, 2003). While these studies find a strong positive correlation between health and SES, there is little consensus on the direction of causality so that health can be both cause and effect of SES. This study examines whether health is cause or effect of financial affordability as a measure of SES, using a latent growth curve modeling (LGCM). Considering low-income households are more likely to experience problems of health and financial affordability, this study assesses how health and financial affordability are interrelated among low-income households.

Methods This research is based on the data (n=730) collected at the Tulsa, Oklahoma Individual Development Accounts (IDA) program whose clients are low-income residents of the Tulsa metropolitan area. The data consist of three waves collected in 1999, 2001, and 2003. This study used demographic variables measured at the first wave. Demographic variables include household size, age, race, marital status, education, and homeownership. This study used self-rated health. Financial affordability construct was measured by three factors such as income-to-needs ratio, liquid assets to income ratio, and self-rated composite measures of living affordability such as home, furniture, medical care, car, food, clothing, and bill. Three wave measures of health and financial affordability were used in this study to construct theoretical models in the LGCM. LISREL 8.52 was used to analyze the LGCM.

Results Three theoretical models were constructed to investigate the relationship between health and financial affordability. First, the three measures of financial affordability were modeled as dependent variables. In the second model, health was considered as a dependent variable. Last, feedback relationships between health and financial affordability were constructed in the final model. Based on the criteria of model fit indices in the structural equation modeling (SEM), the last model with the feedback loops has the better model fits than the other models (RMSEA=0.038; CFI=0.967; GFI=0.970). Among the demographic variables, education and homeownership significantly affect changing rate of health status. Controlling for the demographic variables, health was found to be significantly related to both levels and rates of financial affordability. Healthy householders are likely to have higher levels of income-to-needs ratio, assets-to-income ratio, and living affordability. In addition, the healthier are likely to have positive changing rates of income-to-needs ratio and assets-to-income ratio. However, there were no significant effects of financial affordability on health status.

Discussion Noting that there could be feedback effects between health and financial affordability, this study tested the feedback effects with the LGCM. A key finding is that there is a causal link from health to financial affordability suggesting that it is unlikely that financial affordability accelerates health status. Future study is needed to examine to what point health affects financial affordability and to what degree that point varies for different members of households.