Review of Literature
The study aims to examine the impact of caring for a family member with I/DD Latina and African American women in later life. Informal caregiving of family members with chronic conditions has increased rapidly, and as a result, there is a need for more research on caregiving across different disability groups, chronic conditions, and ethnic minority populations. There is some evidence that the health and well-being of older caregivers of color are severely compromised at the intersection of low socio-economic status (SES), older age, minority status and caregiving challenges. Studies comparing African American caregivers of adults with mental disabilities to White caregivers found that they were less depressed, less burdened and had higher self-esteem than White caregivers; however when controlling for social support and SES, these differences were often not significant (Picket et al., 1993; Pruchno et al., 1997; Steuve et al., 1997). Research on Latino/a caregivers of an adult family member with a mental disability has found that these caregivers experience burden and depression at equal or higher levels than their White counterparts (Magaña et al., in press; Stueve et al., 1997). Both African American and Latina/o caregivers in these studies have been found to be severely disadvantaged due to low education, income and poor health.
Studies comparing African American caregivers of adults with mental disabilities to White caregivers found that they were less depressed, less burdened and had higher self-esteem than White caregivers; however when controlling for social support and SES, these differences were often not significant (Picket et al., 1993; Pruchno et al., 1997; Steuve et al., 1997). Research on Latino/a caregivers of an adult family member with a mental disability has found that these caregivers experience burden and depression at equal or higher levels than their White counterparts (Magaña et al., in press; Stueve et al., 1997). Both African American and Latina/o caregivers in these studies have been found to be severely disadvantaged due to low education, income and poor health.
A growing literature has criticized the effort to understand the experience of persons of minority populations through comparisons with White samples (Betancort & Lopez, 1993; Garcia Coll et al. 1996). Garcia Coll et al. (1996) argue that people of color experience unique environmental contexts that are not shared by Whites, making the comparability of groups questionable. Research comparing families of color to those in the majority culture may under or overestimate the effects on families of color because culture may influence how distress is expressed (Betancort & Lopez, 1993). Other researchers have reported that some minority groups express distress through somatization (Calderón & Tennstedt, 1998; Vega & Rumbaut, 1991). African American and Latino caregivers tend to somatize the effects of cargiver stress, expressing it through physical ailments (Calderón & Tennstedt, 1998). Understanding the effects of caring for an adult child with I/DD on African American and Latino caregivers requires studies that compare these families to African American and Latino families who do not have the atypical parenting responsibilities, but who may experience societal racism, poverty, and difficulty accessing resources. It is important to investigate how pathways to well-being may vary across different ethnic groups. Previous studies that have examined caregivers of color used convenience samples; therefore, representative studies are needed to address limitations of generalizability in previous research.
The theoretical perspective of this project incorporates two bodies of research; caregiving literature in which the stress process model has been prominent, and minority health literature. . While stress process models are useful in conceptualizing the pathways to psychological functioning among caregivers (Lazaurus & Folkman, 1984; Pearlin et al. 1990), few studies have examined physical well-being measures as outcomes. Aging African American and Latina women, have poorer physical health, lower functional status, and higher prevalence rates of most chronic health conditions than aging White women (Collins & Winkleby, 2002;).
In a preliminary study we found that among African Americans mothers of adults with schizophrenia, had more chronic health conditions than women without such caregiving responsibilities (Magaña & Greenberg, 2003). However, the pathways from caregiving to chronic health conditions are unknown. In addition to ethnicity/race, low SES, older age, poor health related behaviors and low access to and utilization of health services are risk factors for poor health are (Collins & Winkleby, 2002; National Center for Health Statistics, 2002. To understand the relationship of caregiving to health, this study will examine risk factors as potential mediators of health conditions.
Sample Population and Methodology
The sample will be drawn from the National Health Interview Survey (NHIS), which collects a representative sample of households obtained through a complex sample design involving stratification, clustering, and multistage sampling. It is an annual multi-purpose face to face health survey conducted by the National Center for Health Statistics (NCHS), CDC. In a preliminary run of the 2000 data, there were an estimated75 Latina women age 40 or older and 98 African American women age 40 and over who were caring for a person with a mental or developmental disability in their household. Because the number of households with people who have DD is small, I will use combined data from the years 1999, 2000, and 2001 during which time, NHIS over-sampled Latino and African Americans. Assuming the numbers are similar across the 3 years, the N for households with a member with DD should be approximately 225 for Latinas and approximately 294 for African Americans.
Data Collection and Measures
Measures used in this analysis include: demographic variables, caregiving or non-caregiving, access to health care, service utilization, health behaviors, and health outcomes. Demographic variables include age, marital status, education level and family income of the women. Access to health care and service utilization includes health insurance, health screenings, usual source of care, delay accessing care, and could not afford prescriptions. Health behaviors include a series of questions about the frequency of drinking, smoking, and exercise. Height and weight variables will be used to calculate BMI. Health status includes self-reported health status for chronic health conditions such as high blood pressure, heart disease, stroke, asthma, cancer, diabetes, and arthritis.
Bivariate analyses (e.g. t-tests, chi squares) will be conducted on all of the domains (demographics, service access, health behaviors, and health outcomes) and outcome variables, comparing caregivers and non-caregivers within each ethnicity (African American and Latino). Demographic variables will be examined for comparability of the caregiver and non-caregiver groups on income, education, age and marital status. If there are significant differences in any of these variables, analyses of covariance will be conducted using the variables in which there are differences as covariates.
Correlations will assess relationships among variables. Logistic regression models will be estimated when using each chronic condition as a dependent variable, and linear regression models will be estimated for the variable of Activities of Daily Living. The initial regression will test for direct effects of explanatory, control, and mediating variables on health outcomes. Statistical techniques outlined by Baron and Kenny (1986) will be used to test mediating effects (e.g., caregiver role leads to poor health behaviors which leads to poor health outcomes) using hierarchical regression models.