Ethnic Differences in the Experience of Filial Caregiving: A Comparison of Hispanic and Non-Hispanic White Caregivers in Colorado

Transcription

1 Structured interviews were conducted with 47 Hispanic and 53 Non-Hispanic White, adult children caring for elderly parents. The selection of measures was based on a conceptual framework wherein caregiving outcomes (both positive and negative) were viewed as a product of variables in five domains: background characteristics, cultural context, primary stressors, secondary stressors, and mediators. Using Multiple Classification Analysis wherein group differences in sociodemographic factors were controlled by treating them as covariates, Hispanic caregivers reported significantly greater adherence to norms of reciprocity, higher levels of caregiving involvement, larger family support networks, higher levels of caregiving satisfaction, and more gains derived from caregiving. Different "world views" and structural differences related ethnicity help explain these findings. Ethnic Differences in the Experience of Filial Caregiving: A Comparison of Hispanic and Non-Hispanic White Caregivers in Colorado Clifton E. Barber and L. Diego Vega Introduction Purpose and Premises The purpose of this study was to compare the caregiving experiences of Hispanic and non-hispanic White families caring for elderly parents. The study was based on two premises: first, that family care of the frail and dependent elderly does not take place in a cultural vaccum, and second, that cross cultural comparisons can enhance understanding of family caregivng. Regarding the latter premise, Pearlin and Zarit (1993) have written: One of the best ways to discern the social and cultural circumstances that undergird patterns of caregiving is to compare caregiving in different national, cultural, and ethnic contexts... such comparisons alert us both to the extent of the variability of caregiving arrangements and to the factors that help explain it.... These comparisons... do not necessarily involve different societies but may also involve the comparison of different cultural groups within a society. (p. 158) Rationale The rationale for the study rested on two points: First, in the absence of comparative data, caregiving patterns observed in one group can erroneously be assumed to apply to other groups; and second, comparative studies of caregiving can be of help not only in identifying the limits of generalizing findings, but also in identifying conditions that explain variations in patterns of family caregiving. To the extent that such The Southwest Jounlal on Aging: Vol. 19,

2 variations are due to ethnicity, the investigators felt that the findings of the study would be useful in developing caregiver interventions and services that would be sensitive to unique ethnic characteristics and influences. Orienting Conceptual Framework The initial task in the study was to construct a conceptual framework depicting the key components of the caregiving experience. The intended use of the framework was to guide the selection of key measures in constructing the interview questionnaire. The resulting framework was an integration of previous caregiving ltiodels developed by Barber (1989), Lawton, Rojagopal, Brody and Kleban (1992), Pearlin, Mullan, Semple and Skaff (1990), and Whitlatch and Noelker (1996). As portrayed in Figure 1, the framework is comprised of six domains: (a) background characteristics, (b) cultural context, (c) primary stressors, (d) secondary stressors, (e) mediators of stress, and (f) outcomes. Boxes drawn with a solid line indicate domains-labeled in CAPS. Nested within domains are categories of variableswith underlined labels-indicated by boxes drawn with a dotted line. Within domains and categories, examples of specific variables/measures in indicated in italic font. The following sections offer brief descriptions of.. each domain and category of variables. Background characteristics. Background characteristics include such factors as caregiver age, gender, income, occupation, employment status, marital status, living arrangements, and level of education. These factors exert their influence throughout the caregiving process. Almost every facet of caregiving is potentially influenced by caregiver characteristics, and to some extent by the background characteristics of the care recipient. Cultural context. Cultural context includes such factors as ethnicity, religiosity, norms of reciprocity, and traditional caregiver ideology. Level of acculturation is also included in this domain, but is not reported in this study since it was assessed only for Hispanic caregivers. Primary stressors. Primary stressors are categorized as being both objective and subjective in nature. Objecti ve stressors center on the care receiver's level of impairment. This includes both the ability to perform activities of daily living as well as cognitive status. The range and difficulty of caregiving activities and the ability of caregivers to manage their relationships with their impaired relatives are directly impacted by a loved one's ADL capacity, memory loss, communication deficits, and recognition failures, and problem behaviors such as wandering. Subjective indicators of stress include such factors as role overload, role captivity (feelings of entrapment and confinement), and relational deprivation (the loss caregivers feel as their relationship with the care recipient becomes more distant and unfamiliar). Secondary stressors. Secondary stressors represent the concept of "secondary appraisal" reported by Lawton, Moss, Kleban, Glicksman, and Rovine (1991). These stressors are viewed as secondary, not because they are less significant than primary stressors, but because they are an outgrowth of the ongoing process of caregi ving (Whitlatch & Noelker, 1996). Secondary stressors are of two types: role strains and intrapsychic strains. Role strains are found in the roles and activities that occur external to caregiving, but that are still influenced by the carcgiving situation. Conflict between family and work roles as well as conflict among family members regarding caregiving are examples of secondary role strains. Intrapsychic strains include challenges to a caregiver's sense of mastery/control, and the erosion of feelings of competency and self-esteem. Outcomes. The experience of caring for an elderly parent produces both positive and negative outcomes. The negative outcomes of caregiving, however, have received far more attention in the literature than have positive outcomes. Common measures of negative outcomes include depression, anxiety, and declines in physical health. Since family caregiving is performed for an individual with whom the caregiver has had a form of emotional attachment, some of the outcomes of caregiving are likely to be positive. In a study of Black and White families caring for frail and elderly loved ones, Lawton et al., (1992) reportedproportionate to the amount of caregi ving Tfze Southwest Journal on Aging: Vol. 19,

3 Figure I. Conceptual Framework of Filial Caregiving BACKGROUND CHARACTERISTICS Age. gender, income. occupation. employment status. marital status, level of education, relationship history. living arrangements CULTURAL CONTEXT Ethnicity. religiosity. norms of reciprocity, traditional caregiver ideology Objective Indicators CR s level of impairment Ability to perform activities of daily living. cognitive status/memory problems; problem behaviors CG s involvement in caregiving Length of time in caregiving role; time spent per week in caregiving PRIMARY STRESSORS Subjective Indicators Relational deprivation Reduction in intimacy, decrease in shared activities Role captivity Restrictions in social activity, degree of vigijance required in supervising CR MEDIATORS Caregiver Coping Skills and Strategies Management of meaning. management of stressful situations in caregiving,i Role Strains Work-family conflict. conflict between or among family members SECONDARY STRESSORS + Intrapsychic Strains Loss of mastery, challenges to a sense of competency Caregiver Financial Resources Amount of income to meet expenses associated with caregiving Positive Satisfactions from caregiving involvement, Personal gains derived from the caregiving role Negative Decline in Health Perceived health. comorbidity. health behaviors and diet, stress symptoms I... ~ Negative Affect Depression, anxiety, subjective burden Caregiver Social Support Formal Support... Informal Support Family network, Friendship network CG = Caregiver (an adult son or dauighter); CR = Care Repient (an elderly mother or father) 35 The Southwest Journal on Aging: Vol. 19,

4 behavior-an increase both in satisfaction and burden. This finding provided support for what Lawton et ai., (1992) termed the "two-factor moder' of caregiving. Mediators. Over time, as the care recipient's condition worsens, primary and secondary stressors proliferate. This can often produce negative caregiver outcomes such as depression, anger, anxiety, and compromised physical health. Mediating factors (coping skills, resources, and social support) have the potential to limit or buffer this proliferation by lessening the negative effects of the multiple domains of stress. Methods Recruiting participants Participants were recruited by contacting a variety of sites located primarily in northern Colorado. These included caregiver support groups, nursing homes, assisted living facilities, adult day care programs, and congregate meal sites. The key-screening criterion was that caregivers needed to be providing at least monthly assistance (for two or more activities of daily living) to an elderly parent. A total of 113 caregivers met the screening criterion and consented to be interviewed. However, nine caregivers did not provide information on key variables (ethnicity, age, gender, income, etc.) and were excluded from the analyses reported in this article. Excluded, too, were four caregivers who reported an ethnic affliliation that was neither Hispanic nor Non-Hispanic White. Hence, the analyses reported in this study consists of 100 filial caregivers: 47 Hispanics and 53 non-hispanic Whites. Measurement of Variables Most of the measures in this study were derived from the REACH project. Additional measures were derived from a large NIMH study of caregivers in California (Pearlin et al., 1990), from the National Study of Households and Families (Sweet, Bumpass, & Call, 1988), and from a study of work-fami1y conflict among employed caregivers (Neal, Ingersoll-Dayton, & Starrels, 1997). The conceptual framework presented previously guided the selection of variables. The following sections describe key measures. These are presented in the order they appear in the conceptual framework even though the primary domains and subcategories are not mentioned. Sociodemographic information. Background information was collected for caregivers and care recipients. This information included age, gender, marital status, income, level of education, employment, living arrangements, and number of persons residing in the caregiver's household. Ethnicity. Caregivers were asked to indicate the primary racial or ethnic group with which they identified: White/Caucasian, Black! African American, Native AmericanlEskimo/Aleut, Asian or Pacific Islander, HispaniclLatino, No Primary Group, and Other. Filial caregivers who identified themselves as "HispaniclLatino" were further asked to indicate whether they viewed themselves as members of one of the following groups: Mexican! Mexican American, Chicano, Cuban or Cuban American, Puerto Rican, Dominican, or Other. Reciprocity. The measure of reciprocity was derived from a 4-item scale used in the National Survey of Households and Families (Kulis, 1992). A sample item is: "Children ought to Jet aging parents live with them when the parents can no longer live by themselves." All items involve a 5-pointresponse set (Strongly Disagree-Strongly Agree). A reciprocity score is derived by summing the four items. The higher the score, the greater the adherence to norms of reciprocity. Traditional caregiver ideology. Traditional caregiving ideology was measured using a 4-item index of attitudes toward caregiving (Lawton, Kleban, Moss, Ravine, & Glicksman, 1989). A sample item focused on how often the caregiver feels that taking care of an elderly parent is important because it represents a means of being true to family traditions. The response set is a 5-point scale: Nearly always, Quite frequently, Sometimes, Rarely, and Never. Reported alphas are. 71 for a minority [Black] sample and.67 for a White sample. Scale items are coded and summed so that a higher score is indicative Th~ Southwest Journal on Aging: VoL 19,

5 of greater adherence to traditional caregiver ideology. Religiosity. The religiosity measure consisted of four items~ with several follow-up items conditional on responses. The first item solicited information on the caregiver's religious preference or affiliation. Eighteen options included the major religions of the world and their largest sects (e.g.~ Catholic, Lutheran, Methodist, Baptist~ Islamic, Hindu, Buddhist, Jewish) as well as "none~" "spiritual, not religious," and "other" categories. The remaining religiosity items attempted to assess the general areas of participation in religious activities, importance of spirituality, and the function of religion as a coping resource for caregiving. The higher the score, the higher the frequency of religioius participation or the greater the importance of religious faith andlor spirituality. Ability to peiform activities of daily living. Descriptive data characterizing the care recipient's level of functioning/physical impairment and overall status were collected using adaptations of widely used measures of the functional capacity to perform activities of daily living-adl (for reviews of these measures, see Guralnik & Simon sick, 1993; Spector, 1990). The degree of care recipient impairment was measured using items from the Katz Activities of Daily Living scale (Katz, Ford, Moskowitz, Jackson~ & Jaffe, 1963), and the Lawton Instrumental Activities of Daily Living scale (Lawton & Brody, 1969). The ADL/IADL sections assessed care recipient's ability to perform tasks of day-to-day importance (Activities of Daily Living: ADLs), and to live independently in the community (Instrumental Activities of Daily Living: IADLs). The ADL section consisted of seven items. These items were based on the list of activities originally used by Katz et a1.( 1963). Response options were "yes/no," and the follow-up items were asked only in the event of a "yes" response. The caregiver was asked if he or she provided help with the activity, how many times per week the caregiver helped with the activity, and how much time on average does helping require. The IADL section consisted of eight items, based on the list of activities originally used by Lawton and Brody (1969). A "yes/no" response was used to indicate whether the care recipient requires help with a given activity over the past week. The caregiver was then asked if he/she provided help with the activity (yes/no response). An item assessing the amount of time spent providing help (number of hours) with these activities followed all IADL items. ADL and IADL items are summed to derive an overall measure of impairment or functional capacity. The higher the score, the greater the impairment. Other problems. Depression, disruptive behavior, memol)'. In addition to measuring functional impairment, the Revised Memory and Behavior Problem Checklist (RMBPC) developed by Teri et al. (1992) was used to measure the extent to which care recipient demonstrated problems that might produce stress for the caregiver. This assessment was The RMBPC consists of a list of potential problems related to depression, disruptive behavior, and memory that the care recipient might have exhibited over the past week. Standard format for item phrasing is: "Within the past week, has (care recipient) done (problem behavior)?" The response options for the initial items are "yes/no." Each item was accompanied by a follow-up item, which is contingent upon an indication of "yes" for the problem behavior. The follow-up items are: "How bothered or upset were you by this?" Response options range from "not at all" to "extremely" on a 5-point scale. With respect to reliability, Teri, et al. (1992) reported alphas of.84 and.90, and subscale alphas ranging from.67 to.89, using a sample of 201 geriatric patients.the severity of elderly parent's depressive symptoms~ disruptive behaviors, and memory-related problems was calculated by summing subscale items; the higher the score, the greater the severity of the problem. Relational deprivation. Relational deprivation was measured using a scale developed by Pearlin et a1. (1990). The items comprising this scale are divided into two subscales. The first centers on the exchange of intimacy; the second pertains more to goals and social activities that were once shared with the care recipient, but that are no longer attainable. Items were preceded by the following: "Caregivers The Southwest Journal on Aging: Vol. 19,

6 sometimes feel that they lose important things in life because of their relative's illness. To what extent to you feel that you have personally lost the following?" A sample item is: "How much have you lost being able to confide in (care recipient)?" There is a 4-point response set: Not at all, Somewhat, Quite a bit, and Completely. Reported alphas are.77 for the first subscale and.67 for the second subscale (Pearlin et ai., 1990). No subscale scores were used in this study; only a total relational deprivation score derived by summing scale items; the higher the score, the greater the caregiver's feelings of relational deprivation. Role captivity/restrictions on social activities. Role captivity is based on the perception of being restricted in social activities due to caregiving involvement (Aneshensel, Pearlin, & Schuler, 1993). The scale used to measure role captivity was derived from the Social Activities measure developed for the REACH project (Rubert & Ory, 1998). The scale consisted of six items designed to assess how satisfied the caregiver is - in the context of caregiving responsibilities-with the frequency that they have engaged in activities such as spending quiet time by oneself and taking part in hobbies over the past month.. There was a seventh summary item which is worded "Over the past month, how satisfied are you with the amount of time you have been able to spend in activities that you enjoy?" Response options to this item matched the others in the scale: Not at all~ A little; A lot. A score for this scale is derived by summing the 7 items. The higher the score, the fewer the restrictions (i.e., the less the role captivity) in the caregiver's social activities as a result of caregiving. Vigilance. The Vigilance Scale developed for the REACH project and consisted of three items probing the amount of time that the caregiver is required to spend with the care recipient in order to supervise him or her. The first item asked if the care recipient can be left alone in the event of a family emergency (response options: "yes," "no," "refused," "unknown"). Three follow-up items assessed how long the recipient can be left alone, if the recipient can be left alone in a room, and how long the recipient can be left alone in a room. The second item inquired about the number of hours the caregiver feels he/she must "be there" to care for the care recipient, and the last item asked the caregiver for an estimate of the number of hours he/she is actually doing something to provide care for the care recipient. A "vigilance" score was derived from the caregiver's response to the second item. The higher the score, the greater the degree of viligence or supervision required. Work-family conflict. Work-family conflict was measured using the Job-caregiving Conflict Scale developed by Pearlin et ai., (1990). This is a 5-item scale with a reported alpha of.75. Given a time frame of the preceding two months, caregivers are asked, "From your own personal experience, how much do you agree or disagree with the following statements about your present work situation?" A sample item is: "You worry about your (elderly parent receiving care) while you are at work." The amount of work-family conflict stemming from caregving is calculated by summing the responses to the five items. Higher scores are indicative of higher levels of work-family conflict due to caregiving. Family conflict. The amount of family conflict spawned by caregiving was measured using three subscales developed by Pearlin et a1. (1990): (a) conflict regarding seriousness of parent's condition and safety (alpha =.80), (b) conflict regarding attitudes and actions toward the parent (alpha =.86), and (c) conflict regarding attitudes and actions toward the caregiver (alpha =.84). The repsonse set for all items in these subscales was as follows: quite a bit of disagreement; some disagreement; just a little disagreement; and no disagreement. For each subscale, scores were summed across items; higher scores are indicative of greater conflict. Sense of mastery/control (or lack thereof). Caregiving mastery was measured using 7 items developed by Pearlin and his colleagues (Pearlin & Schooler, 1978; Skaff, Pearlin, & Mullan, 1996). This scale measures the extent to which a caregiver feels she/he can exercise some degree of control over caregiving experiences and circumstances. Response categories are set on a 4-point likert scale (strongly disagree to strongly agree). Responses to scale items are summed to derive a total score; higher scores are indicative of greater perceived mastery and control The Southwest Journal on Aging: Vol. 19,

7 Competency in careglvlng (or lack thereof). Perceived care giving competency was measured using a 4-item scale requesting participants to rate the perceived adequacy of their performance as caregivers (alpha =.74). The Caregiving Competence Scale was created by Marilyn M. Skaff (Pearl in et ai., 1990). Sample items include: "How much do you believe that you've learned how to deal with a very difficult situation?" and "How much do you feel that all in all, you are a good caregiver?" Response set categories are: Very much; Somewhat; Just a little; Not at all. A total scale score is obtained my summing items. Higher scores are indicative of greater feelings of competency in fulfilling the caregiving role. Managen1ent of lneaning. The Management of Meaning measure consisted of three subscales; each designed to measure a cognitive strategy that might be used to manage the meaning of a stressful caregiving situation: reduction of expectations (subscale alpha =.48), use of positive comparisons (subscale alpha =.63), and the a search for a larger sense of the illness (subscale alpha =.49). Each subscale consists of three items. Sample items include: "How often do you try to accept your (care recipient) as (he/she) is, not as you wish (care recipient) could be?" (reduction of expectations), "How often do you remind yourself that others are worse off?" (using of positive comparisons), and "How often do you try to make sense of the illness?" (searching for a larger sense in which to frame the care recipient's illness and frail condition). All items use a 4-point response set designed to assess how often the caregiver employs the strategy: Very often, Fairly often, Once in a while, and Never. A higher score in each of the scales is indicative of better management of the meaning of the caregiving experience. Subscale scores are derived by summing responses for the three items; higher scores indicate higher frequency of using the strategy to manage. Management of caregiving situation. Previous research (Pearlin et ai., 1990) involving family caregivers of Alzheimer's patients has reported four strategies that may be to more easily manage the demands involved in caring for a loved with behavior and/or memory problems: (a) being firm in directing (care recipeint's) behavior, (b) doing only those things that really need to be done (and letting other things "slide"), (c) trying to find ways to keep (care recipient) busy, and (d) trying to learn as much as possible about (care recipient's) illness(es) and condition(s). The response set for each item is intended to measure the frequency with which the strategy is used by the caregiver: never, once in a while, fairly often, very often. Higher scores are indicative of more frequent use. Financial resources. In addition to soliciting information on the caregiver's household income, an assessment was also made of the extent to which the economic situation of caregiver has been affected by the caregiving responsibilities. This was used by the REACH project as a measure of a secondary stressor. However, in this study it was used as a mediator of because it was viewed as a caregiver resource. This measure includes 3 indicators: (a) reduction in household income due to caregiving, (b) increased expenditures related to the care and treatment of the care recipient, and (c) whether there is enough money to make ends meet month to month (Pearlin et ai., 1990). The first two questions ask the participant to compare the financial situation as it was just before the onset of providing care for the care recipient, and the financial situation at present. Specifically, the wording of these two items was as follows: "Compared to that time (before caregiver began caring for care recipient), how would you describe your total household income from all sources?" and "Compared to that time (before the onset of caregiving), how would you describe your monthly expenses 7" Response categories were: Much less now; Somewhat less now; About the same; Somewhat more now; Much more now. The third question was: "In general how do your family finances workout at the end of the month?" Response categories for this latter question were: Not enough to make ends meet; just enough to make ends meet; and, Some money left over. Utilization of formal support services. Formal support was measured with the Formal Care and Services questionnaire from the REACH project. This questionnaire consists of 22 "yes/no" items that The Southwest Journal on Aging: Vol. 19,

8 assess whether a service was used during the past month. Each item is accompanied by a follow-up item, which is contingent on a ''yes'' response, asking the number of times the service was utilized during the past month. This form consists of a fairly exhaustive list of possible in-home, community, physician, and hospital services that either the care recipient, caregiver, or both may have utilized within the past month. Each "yes" response was counted as one, and the total of different services used during the past month was tallied. Higher scores are indicative of higher rates of service utlization in terms of the number of different services used. Infonnal support networks. Caregivers' informal social networks were measured with the Lubben Social Network Index (Lubben, 1988). This scale includes separate subscales for family (3 items), ftiends (3 items) and confidants (2 items), and captures overall size of family and friends portions of the network. Alpha =.70 for the total scale. Confirmatory factor analysis suggests multiple factors, good fit and high loadings for family and friend factors (Newsom & Schulz, 1996). All items are on a 5-point scale, with frequency ratings or categories for the number of network members. Received support was measured using the Received Support Scale, which contained items derived from the Inventory of Socially Supportive Behaviors (Barrera, Sandler, & Ramsey, 1981) and modified by Krause (1995). Scale consists of 11 items (with 3 for tangible, 4 for emotional, and 4 for informational subscales). Response set is based on a 4-point rating of frequency ("never" to "very often"). Previously reported confirmatory factor analysis indicates good fit and high loadings ranging from.678 to.883. Caregivers' satisfaction with support was measured using a scale developed by Krause (1995). The scale consists of three items; one each concerning the overall satisfaction with tangible, emotional, and informational support received. The original measure used a yes/no response format, but in order to increase sensitivity these items were changed to a 4-point scale, "not at all" to "very." Previously reported confirmatory factor analysis reports a good fit and high factor loadings, ranging from.771 to.994 (Krause, 1995). An additional item was added at the beginning of the form which captures all of the types of support, which is worded "Overall, how satisfied have you been in the last month with the help you have received from friends, neighbors, or family members?" Positive aspects of caregiving. This scale consisted of eleven items, phrased as statements about the caregiver's mental/affective state in relation to the caregiving experience. Response options are on a 5-point agree/disagree scale. This scale was adapted from the Caregiver Health Effects Study (Schulz et ai., 1997), and differed from the latter in three ways: (a) response options were changed from the yes/no format used on the original in order to increase variability of responses and improve reliability; (b) questions were rephrased as statements to accommodate the agree/disagree response option; and (c) overall instructions were modified to avoid negative reactions from caregi verso Overall scale score was calculated by adding the 11 items. Higher scores are indicative of more perceived positive aspects of caregiving. One of the ways in which caregivers may cope with the stress and negative affect associated with caregiving is by emphasizing the positive aspects of providing such care. This scale, which probes agreement with statements such as "Providing help to (care recipient) has made me feel more useful," provides a means of measuring these compensatory mechanisms. This topic has received considerable attention in recent years as a counterpoint to the negative aspects of caregiving emphasized in the majority of the literature (e.g., Lawton et ai., 1991). Personal gains froln caregiving. Personal gains were measured a 4 item scale wherein are listed some benefits of caregiving (e.g., "How much have you become more aware of your inner strengths from taking care of (parent receiving care)?"). Reported alpha for this scale is.76 (Pearlin et ai., 1990). Response categories were: Very much; Somewhat; Just a little; Not at all. The higher the score, the greater the personal gains perceived by the caregiver. Health. Four elements of self-reported health were assessed: perceived health, comorbidity, health behaviors and diet, and stress-related symptoms. The The Southwest Journal on Aging: Vol. 19,

9 first four items probe perceived physical health (Stewart & Archbold, 1986). Response options for the first item (general health) range from "poor" to "excellent" (5-point scale). This item is commonly used by epidemiologists and is included in such scales as the MOS SF-36 (Ware et al., 1995). The next three items are statements, with response options ranging from "definitely true" to "defmitely false" (5-point scale). The next seven items assess comorbidity. These items ask the caregiver whether helshe has been diagnosed as having or has any of several chronic health problems, ranging from arthritis to stroke. Response options are yes/no. The format of these items has been changed from the original so that the interviewer does not have to repeat the initial question stem, which now reads as follows: "Do you have, or has a doctor told you that you have, any of the following health problems?" Three items, previously used by Stewart and Archbold (1986), assess stress-related health symptoms. These items probe the extent to which the caregiver has experienced dizziness, headaches, and stomach or bowel problems over the past two weeks. Response options are: Not at all; A little; A lot. The last five items are from the Nutrition Screening Initiative (NSI), sponsored by the American Academy of Family Physicians and the American Dietetic Assosiation (Posner, Jette, Smith, & Miller, 1993). The items generally address the caregiver's diet and daily routine, with regard to eating, drinking, weight gainlloss, exercise, and smoking. All responses are yes/no. In all instances, higher scores are indicative of poorer health, higher comorbidity, and a higher frequency of unhealthy behaviors! stress-related symptoms. Subjective burden. The amount of subjective burden experienced by the caregiver was measured using 10 items expressing the general sense of the personal costs of caregiving (Lawton et al., 1989). An example is, "I feel very tired as a result of caring for (name of person receiving care)." Reported alpha is.87. Higher scores are indicative of greater subjective burden. Depression. Caregiver depression was measured using the Center for Epidemiological Studies Depression (CES-D) Scale (Radloff, 1977). The CES-D is a 20-item measure that asks about the frequency with which respondents have experienced depressive symptoms within the last week. Response options range from 0 to 3 for each item (0= rarely or none of the time, 1= some or little of the time, 2= moderately or much of the time, 3= most or almost all the time). Higher scores are indicative of more depression. The CES-D was chosen as one of the measures of depression because it is brief, can be administered by interviewers with little special training, and is widely used in medical, epidemiological, and psychological studies. The CES-D also provides cutoff scores (e.g., 16 or greater) that aid in identifying individuals at risk for clinical depression. The CES-D has shown relatively high internal reliability and high predictive validity (Baker, Velli, Friedman, & Wiley, 1995). Anxiety. Anxiety as an indicator of negative affect as an outcome of caregiving was measured using 10 item-scale, taken from Spielberger's State-Trait Personality Inventory (STPI: 1979). The anxiety subscale of the STPI is comprised of items taken from the larger State-Trait Anxiety Inventory (Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983). Respondents are asked how much they agree with statements such as "I feel nervous." Response options range on a 4-point scale of intensity, from "not at all" to "very much" ("refused" & "unknown" also available). Four of the items are scaled negatively (a "very much" response would indicate a low level of anxiety, e.g., "I am relaxed"). The higher the score, the more anxiety experienced by the caregiver. Results and Discussion Background Characteristics Table 1 compares Hispanic and Non-Hispanic White caregivers on selected background characteristics. Significant differences between the two groups included caregiver age, education level, employment status, religion, number of persons living in 'caregiver's household, and co-residence with care recipient. There were no differences with regard to gender, income or marital status. Non-Hispanic White filial caregivers were on average eight years older than were Hispanic (The Southwest Journal on Aging: Vo]

10 Table 1. Differences in Background Characteristics by Ethnicity Non-Hispanic Hispanic b Characteristic White" (n = 53) (n = 47) t Chi square.---~~-- Caregiver age (mean) *** Care receil'er age (mean) *** Caregiver gender (% women) Care receiver gender (% women) Caregiver marital status (%) Never married Married Widowed Separatedld i vorced Care receiver marital status (%) 3.9 Never married Married Widowed Separatedldi vorced Caregiver education (%) 24.3*** Less than High School High School Some College College Graduate Post-graduate Work Care receiver education (%) 28.6*** Less than High School High School Some College College Graduate Post-Graduate Work Caregiver Income 4.2 Less than $ $ $ $ $ More than $50, Caregiver Employment Status (%) 14.0** Full time Part time Homemaker Retired Unemployed The Southwest Journal on Aging: Vol. 19,

11 Non-Hispanic Hispanic b Characteristic White- (n = 53) (n = 47) Chi square Number or people residing in caregiver's house (%) Care receiver's living arrangements (%) With caregiver 30.2 With another relative 0.0 In an assisted living facility 18.6 In an nursing home 23.3 In own home 27.9 Other 0.0 Caregiver relationship with care receiver (%) Daughter 63.0 Daughter-in-law 17.4 Son 17.4 Son-in-law 2.2 Caregiver religion (%) Protestant 49.1 Catholic 9.4 Other-Christian 22.6 Spiritual. not religious 15.1 None 3.8 a Il 53 p Il = 47 ~'*p<.oi. ':'**p<.ooi. caregivers (54 years versus 46 years). As will be discussed shortly, this difference was also mirrored in the fact that care recipients in Non-Hispanic White families were older than those in Hispanic families. The finding that the two groups did not differ significantly in terms of annual household income merits further discussion because it should not be interpreted to mean that these two groups were equal in terms of monetary resources. Differences in caregiver employment status and number of dependent persons residing in the caregiver's household provided contextual evidence that Hispanic caregivers did not share the same financial advantage. Whereas 60% of the Hispanic caregivers were employed full time, such was true for only 40% of the Non-Hispanic White caregivers. The ** 17.1 ** *** percentage of Non-Hispanic White caregivers who were retired was almost five times higher than that of Hispanic caregivers (19% versus 40/0). Further, 32% of the Hispanic caregivers indicated that there household size was 5-7 persons (mostly dependents), compared to 6% of Non-Hispanic White caregivers reporting a household size of 5-7 persons. There were also significant differences in level of education, with a higher percentage of Hispanic caregi vers reporting 12 years or less of formal education (47% of Hispanics compared to 13% of Non-Hispanic Whites). Despite this difference, both groups of caregivers were likely not representative of the general population of caregivers, with an over-representation of those who had higher levels of 43 The Southwest Journal on Aging: Vol. 19,

12 Table 2. Caregiving Characteristics by Ethnicity (adjusted mean deviations) Grand Non-Hispanic DOMAIN High Score Code Mean White [cul TURAL CONTEXT: Variables ill this domain were adjusted f~r care-g-h-'e-r-a-g-e Adherence to norms of reciprocity Greater adherence Traditional caregiving ideology Greater adherence to traditional ideology Religiosity Frequency of religion practices Higher frequency * Importance of religion Greater importance * PRIMARY STRESSORS: CR'sfimcliollal ability was adjusted/or CR age alld CG educatio1l; CG's involvemelll was adjustedfor CG gender, and er's level ofimpairme11l: and subjective indicators o.{primary stress were adjustedfor CR's level ofimpairmem. Objective Indicators-CR's Functional Ability Overall level of impainnent (ADUIADL) Greater impairment Depressive problems Greater impairment Disruptive problems Greater impairment Memory-related problems Greater impairment Objective Indicators-CG's involvement Number of ADLIIADL activities for * More involvement which CG provides assistance to CR Hours per day CG cares for CR More involvement *** Subjective Indicators Relational deprivation Greater deprivation Role captivity -Restriction in social activities Less restriction Vigilance (hours "on duty" per day) More hours SECONDARY STRESSORS: Variables ill this domain were adjustedfor CR age. Role Strains Amt. work-family conllict More conflict Amt. family connict regarding safety More conflict Amt. family connict-attitudes toward CR More conflict Aml. family conllict-attitudes toward CG More contl ict Intrapsychic Str,lins Caregiving mastery Higher mastery Hispanic F 14.65*':'::: 2.52

13 DOMAIN High Score Code Grand Mean Non-Hispanic White Hispanic F Caregiving competence Higher competence MEDIATORS: Measures of caregivillg coping skills (e.g. mallagemelll o/meallillg. mallagement of caregivillg situalion) were atijtlsledjor CG emp/oymem : Social Slipporl was adjusted for CG age, CG ii/come, alld CR's level of impairmelli Caregiver Coping Skills and Strategies Religious observancclritual as a source of help More helpful Management of meaning (strategy) I. Reduction of Expetactions More frequently used 2. Positive comparisons More frequently used 3. Search for a larger sense of the illness More frequently used Management of caregiving situation (strategy) 1. By trying to be finn in directing (er's) M f I d behavior ore requent y use 2. By doing things you really have to do 3. By trying to find ways to keep (CR) husy 4. By learning more about (CR's) illness Economic Resources More frequently used More frequently used More frequently used Compared to the time just before you began to I-Less now take care of (CR),?, how would you describe your total household income, from all sources'? 5-More now Compared to that time. how would you describe I-More now your monthly expenses? 5-Less now Social Support Utilization of formal support services CG's family network CG's friends network CG's tangible support received CG's emotional support received Higher utilization More support More support More support More support CG's infonnational support received More support * ** Satisfaction with received infonnal support Greater satisfaction OUTCOMES: Variables in this domain were adjusted for CG employment and income. Subjective burden was adjusted for CG employment. income and CR's level of impairment. Positive

14 DOMAIN Positive aspects of caregiving Personal gains in caregiving Negative-Decline in CG Health Perceived health status Co-morbidity CG's unhea1lhy behaviors CG's stress-related symptoms ~ Negative- Affect ~ ~ Subjective burden ~ Depression (CES-D) ;:s- ~ Anxiety ~... *12 <.05. **n <.01. ***12 <.OOL ~ ;: ~ ~ - Grand High Score Code Mean More aspects Greater gains Poorer health 9.21 Higher co-morbidity 0.84 More behaviors 2.54 More symptoms 1.31 Higher burden More depression Higher anxiety Non-Hispanic White Hispanic F * *

15 education (some college, a college degree, or post-graduate work). In terms of the context of caregiving, not only were Hispanic caregivers more likely to have a larger number of persons living with them, they were more likely to be co-residing with the older parent who was the recipient of care (39% co-residence with recipient for Hispanic caregivers versus 30% for Non-Hispanic Whites). Further, in Hispanic families, a higher proportion of daughters were serving as primary caregivers (82% versus 63%). For both groups, the majority of parents receiving care were mothers (83% in Hispanic families and 87% in Non-Hispanic White families). There were also no differences in marital status (about 70% of care recipients in both groups being widowed). However, Hispanic care recipients were significantly younger than Non-Hispanic White care recipients (77 years versus 85 years) and had a higher level of functioning in terms of being able to perform activities of daily living. Hispanic care recipients needed help with 7 ADL during the week of the interview, compared to 10 ADLs for Non-Hispanic White care recipients. Investigating Ethnic Differences The strategy for addressing the role of ethnicity in caregiving was based on Mutran' s work (1985) wherein she reports that ascribing ethnic effects is best achieved by identifying "residual" explanations. In this approach, the assumption is made that ethnic differences remaining after background and other sociodemographic factors have been controlled constitute evidence of a cultural or ethnic effect. In this study, multiple classification analysis (MCA) was selected as the statistical method for "controlling" for the effect of background and other sociodemographic factors. Although MCA is the mathematical equivalent of multiple regression analysis. it has the favorable feature of representing means that are adjusted for both specified covariates and that are independent of the effect of other main independent variables. Lawton et al. (1992) also used this approach in a study of Black versus White caregivers. Specifically, Lawton et al. (1992) treated caregiver age, gender, income, education and impaired-person age and gender as a covariates for all the dependent variables they examined. Unlike Lawton et al. (1992), however, this study employed what was deemed a somewhat more refined approach advocated by Stevens (2002) wherein only those covariates that are significantly correlated with the dependent variable are controlled in statistical analyses. Whereas Lawton used the same covariates for investigating racial/ethnic differences in all domains of caregiving, in this study the covariates differed across the domains of caregiving, and even across the subcategories of dependent variables within caregiving domains. For example, the age of the care recipient was the primary covariate in examining group differences in the domain of secondary stressors, but the covariates included in the analysis of ethnic differences for primary stressors included the care recipient's age and functional ability, and the caregiver's gender and educational level. Hispanic and Non-Hispanic White mean comparisons (adjusted for selected covariates) of the dependent variables are shown in Table 2. The second column in the table denotes the directionality of variable values. Generally, the higher the score, the higher the variable value. Covariates used as controls vary from domain to domain, and even from measue to measure within domains. Covariates are indicated in the narrative next to the heading for each domain.the grand mean and the adjusted mean deviations appear in the middle of the table. The F values associated with ethnic group and caregiver differences (net of the covariates and of each other) are shown in the right-hand column. Cultural Context Nonns of reciprocity and traditional caregiver ideology. Hispanic caregivers reported more frequent adherence to norms of reciprocity than did Non-Hispanic Whites. In their study of Mexican American caregivers, Clark and Huttlinger (1998) reported that reciprocity was one of the two most important emerging themes in the caregiving experience. Knowing this, the authors expected that Hispanic caregivers in this study would report higher scores on the Norms of Reciprocity scale. Indeed, The Southwest Journal on Aging: Vol. 19,

16 Hispanic caregivers did report greater adherence to the norm of reciprocity than did White caregivers, even after controlling for caregiver age. Curiously, however, no ethnic differences were found with regard to the measure of traditional caregiving ideology (Lawton et ai., 1989). Recall that the latter instrument was a 4-item index of attitudes toward caregiving that included the wish to repay the debt of being cared for as a child, continuing in the family tradition of mutual concern, fulfilling personal cultural values, and setting an example for one's children. Why were there ethnic differences on some factors related to cultural context - and particularly the measure of reciprocity-but not for the measure of traditional caregiver ideology, which included items designed to tap reciprocity and other attitudes affected by culture? Perhaps one response to this question can be derived from a closer scrutiny of specific scale items. The Norms of Reciprocity Scale consists of four items, two of which focus on intergenerational reciprocity as it pertains to co-residency ("Children ought to let aging parents live with them when parents can no longer live by themselves," and "Parents ought to let children live with them when the children are having difficulty"). The most pronounced differences between Hispanic and Non-Hispanic White caregivers were on the two co-residency items, with Hispanic Caregivers reporting significantly higher scores on both (and particularly the item focusing on adult children letting elderly parents co-reside with them). There were no ethnic differences on the other two items (reciprocity in sharing financial resources). As mentioned previously, items in the Traditional Caregiving Ideology Scale focus on caregiving as an opportunity to repay, to be true to family traditions, to live up to religious principles, and to provide a good model for the caregiver's own children. Generally, there were no significant ethnic differences on the four items in this scale. Importantly, Non-Hispanic White caregivers were more likely than their Hispanic counterparts to respond "Nearly always" or "quite frequently" to the item designed to measure reciprocity: "How often do you think of the help you give (to elderly parent) as an opportunity to repay herlhim?" As it applies to filial caregiving, a central feature of the norm of reciprocity is that adult children should help their parents in later life because their parents have previously rendered assistance to them. In a classic article on this topic, Gouldner (1960) has persuasively argued that the norm of reciprocity is universal (although not necessarily unconditional). In this study, it appears that ethnic differences in the norm of reciprocity are selective. For Hispanic caregivers, adherence to this norm most often occurs in the form of co-residence, but not for financial resources. In light of the foregoing discussion, it is not surpnslng that the percentage of caregivers co-residing with their elderly parents in this study was higher for Hispanics than for Non-Hispanic White caregivers. This finding is even more significant given the fact that the Non-Hispanic White care recipients were older and more functionally impaired than their Hispanic counterparts. It appears that intergenerational co-residence is a function of ethnicity. The comparatively higher percentage of Hispanic caregivers co-residing with their elderly parents is intertwined with the finding that Hispanic caregivers are more involved in caregiving than their Non-Hispanic White counterparts. What is unclear is whether co-residence or caregiving involvement is the most directly influenced by Hispanic culture. Are Hispanic families more involved in caring for a parent because of the higher incidence of co-residence, which is a product of a norm of reciprocity? Or does the cultural imperative to be involved in the care of one's elderly parents drive the comparatively higher incidence of co-residence? Religiosity. There were ethnic differences for several measures included in the domain of cultural context. For example, even after controlling for caregiver age (Hispanic caregivers were younger than Non-Hispanic White caregivers), the importance of religious affiliation and frequency of religious practices (e.g., prayer/meditation, attendance at religious services) was significantly higher for Hispanics caregivers The Southwest Journal on Aging: Vol. 19,

17 Primary Stressors After controlling for caregiver gender and care recipient's level of impairment, a key finding was that Hispanic caregivers provided help with more ADL/IADL than did Non-Hispanic White caregivers. They also reported more hours per day caring for the care recipient. Since this difference remained after controlling for the other confounding factors that would affect involvement (caregiver gender and level of parent impairment), one conclusion is that comparatively greater involvement in parent care is a function of ethnicity. Again, however, it is difficult to discern whether this greater involvement in caregiving is more directly a function of co-residence or ethnicity. Co-residence and Delay in Institutionalization The finding that Hispanic caregivers reported both a greater tendency to co-reside with their elderly parents, and a greater tendency to be involved in their day-to-day care, might have result in delaying needed institutionalization in some instances. Mental heath providers working with Hispanic families could help these caregivers to frequently assess or monitor the caregiving situation, and help them decide when the institutionalization of their elder would be the appropriate action to take. Therapists should work with families to overcome internal conflicts in the process of institutionalizing a parent. The decision to institutionalize a parent is painful, difficult, and violates Hispanic caregiving values. Sometimes the decision to place the elder in a facility produces significant conflict and emotional response among family members. Placement in a facility should occur when adequate care could no longer be provided by the family at home. The elder's physician, as well as therapists and counselors, can play an important role in the decision making of the placement (White, Peters, Martin, & Hummel, 2002). Moreover, the fact that Hispanic families tend to keep their parents at home longer is not a guarantee for the parents to have all of their needs covered. Dietz (2002) found that despite the amount of contact, relatively small percentages of Mexican American families help their aging family members with their needs. Several studies have challenged the idea that older member of minority groups tend to underutilize formal social services because their needs are already met by members of their family or community. This is another reason for therapist to help Hispanic families in assessing how they are helping their older members. Secondary Stressors Before controlling for care recipient age, there was a significant difference in the amount of family conflict regarding attitudes toward the care recipient that Hispanic and non-hispanic White caregivers reported. After controlling for care recipient age, this difference disappeared. Wh y this was the case is unclear. After controlling for care recipient age, no differences between the two groups were found in the following areas: amount of work-family conflict, amount of family conflict regarding safety of the care recipient, amount of family conflict regarding attitudes toward the care recipient, amount of family conflict regarding attitudes toward the caregiver, caregiving mastery, and caregiving competence. Mediating Factors The domain of mediators included both caregi ver coping skills, strategies, and economic resources, as well as social support. The general finding was that there are very few ethnic differences in this domain. The only notable exception was in the arena of social support. Initially, the utilization of formal support services appeared to be a product of ethnicity, with Non-Hispanic White caregivers reporting higher utilization than Hispanic caregivers. This finding is congruent with a recent comprehensive review of studies on the effect of race, ethnicity and culture on caregiving wherein the authors observe that the utilization of formal support services is generally lower among minority caregivers than nonminority caregivers (Dilworth-Anderson, Williams, & Gibson, 2002). However, after controlling for care recipient's level of impairment and caregiver income, the difference disappeared. The initial difference in The Southwest Journal on Aging: Vol. 19,

18 fonnal service utilization was more a product of functional impairment and caregiver financial resources than ethnicity, although one could argue that fmancial resources are a derivative of minority culture in the United States. There were no differences in how satisfied Hispanic and Non-Hispanic White caregivers were with the amount and quality of infonnal support they received. However, they did differ in tenns of the primary sources of infonnal support. Hispanics caregivers reported a more extended family network and more frequent contact with relatives as a source of informal support than did Non-Hispanic White caregivers. On the other hand, Non-Hispanic White caregivers reported a more extended friendship network as a source of informal support. Phillips, de Ardon, Komnenich, Killeen, and Rusinak (2000) reported that Mexican American caregivers identified significantly fewer networks available to provide informal support than their Non-Hispanic White counterparts. Outcomes Negative outcomes. There were no ethnic differences for any of the measures of negative outcomes: decline in health, comorbidity, depression, subjective burden, or anxiety. The finding that the,amount of burden experienced by these two groups was similar (after controlling for care recipient age and functional ability) was not surprising, since it mirrors fmdings reported in other studies involving ethnic comparisons of caregivers (Cox & Monk, 1996; Valle, Cook-Gait, & Tazbaz, 1993). Positive outcomes. One of the most important ethnic differences in the domain of outcomes was that Hispanic caregivers reported more caregiving satisfaction and greater personal gains from caregiving. This finding is significant because Hispanic caregivers also reported comparatively higher levels of involvement-a situation that might be expected to result in fewer positive outcomes. Different "World Views" One of the implications of the aforementioned combination of findings-and particularly those centered on the positive outcomes and gains derived from fllial caregiving-is that Hispanic and Non-Hispanic White caregivers view caregiving differently and, as a consequence, may either use a different anchor for making judgments or may simply tend to emphasize the positive rather than the negative. Some literature suggests that from a Mexican American "world view," aging and caregiving reflect a positive affmnation of life that is directly related to strengthening cohesion within the family unit (Escobar & Randolph, 1982; Mintzer et at, 1992). In fact, caring for an elderly parent has been reported to be a source of pride and status for Hispanic families (Jenkins & Kamo, 1992). Additionally, for Hispanic adult children caring for elderly parents-more so than for their Non-Hispanic White fllial counterparts-the predominant view of all life events (including caregiving) may be the "sociocentric self' (Landrine, 1992), which orients the individual to embeddedness in relationships and contexts rather than to autonomous needs, desires, and actions. If what a filial caregiver is judging, with regard to caring for an elderly parent, is viewed as "life," not as "infrrmity" or "frailty," then it should not be surprising that when confronted with the aging and increasing frailty of a parent, an Hispanic adult child views involvement in the caregiving experience as strengthening social continuity rather than "interfering with my life." Structural differences In addition to different worldviews, the findings of this study could also reflect structural differences. Two are worth noting here. First, Hispanic adult children may assume a caregiving role earlier than Non-Hispanic Whites, when their elders are actually less functionally impaired. In general, Hispanic children are believed to be very protective of elderly parents, particularly elderly mothers (Markides & Krause, 1985). The finding that daughters assume caregiving roles earlier may be a function of Hispanic children's propensity to protect their aging mothers. Second, although the precipitating factor for fllial caregiving in Non-Hispanic White families may be when an elderly parent is widowed, this may not be needed in order for Hispanic children to initiate the The Southwest Journal on Aging: Vol. 19,

19 caregiving role. Further, the mechanism for designating caregivers may be different for Hispanics. Among Non-Hispanic Whites, the principle of substitution (Brody, 1985) has been documented, with spouses assuming responsibility for caregiving, and adult children taking over when the spouse is no longer able or available to do so. This principle may not be operative in Hispanic families. Conclusion Kleinman (1980) described how cultural beliefs regarding the causes and treatments of disease strongly impact the kinds of problems that are experienced and the kinds of treatments that are viewed as acceptable. Applied to caregiving, Kleinman~s work means that not all cultures experience caregiving as an exclusively stressful experience. In such cultures-and the authors of this article believe this applies to Hispanic familiesinterventions that focus on the negative aspects of caregiving may not be well received, and may even be counterproductive. Those working with minority caregivers will benefit from being reminded that there are several key similarities among minority caregivers in the United States (Lockery, 1991): (a) the value placed on intergenerational contact, (b) the tradition of respect toward the elderly~ and (c) the clear family obligation for caregiving in the home, rather than in an institutional setting. Further, clinicians should not overlook the fact that positive appraisal (as a cognitive strategy) contributes to minority caregivers' abilities to overcome added hardships imposed by low income and limited access to health care (Haley et al., 1996). We close by echoing the observation and challenge issued by Gallagher-Thompson, Coon, Rivera, Powers, and Zeiss (1998): minority caregivers' values and beliefs, along with their life experiences, create a powerful variable which impacts appraisal and coping, so that for many, caregiving is almost automatically less stressful than for Anglo counterparts. Thus, they may not need or want the same kinds of services that have been developed to serve the more well-characterized needs of Anglo caregivers... future research must be culturally sensitive to the way it asks about caregiving, and in the types of interventions and services that are proposed to assist minority families with the caregiving process. (p. 484) Acknowledgments Funding for this project was provided by a grant from the Agricultural Experiment Station through the College of Applied Human Sciences at Colorado State University. The authors express appreciation to Dr. Richard Schulz for permission to incorporate into the interview several measures developed for use in REACH (Resources for Enhancing Alzheimer's Caregiver Health), a national initiative established in 1995 by the National Institutes of Health. Appreciation is also expressed to Dr. Dolores Gallagher-Thompson for her assistance in arranging training for project interviewers at the REACH site in Palo Alto, and for sharing Spanish versions of REACH instruments. Lastly, the authors gratefully acknowledge the diligent efforts all persons who worked on this project, and especially thank caregivers who willingly and candidly shared their caregiving experiences. Notes I Clifton E. Barber is a professor in the Department of Human Development and Family Studies~ Colorado State University. 2 L. Diego Vega is a doctoral candidate in the School of Education, Colorado State University. References Aneshensel, C., Pearlin, L., & Schuler, R. (1993). Stress, role captivity and the cessation of caregiving. Jounlal of Health and Social Behavior, 34, Baker, F.M., Velli, B.A., Friedman, J., & Wiley, C. (1995). Screening tests for depression in older Black vs. White patients. The American Jounlal of Geriatric Psychiatry, 3( I), The Southwest Journal on Aging: Vol. 19,

20 Barber, C. E. (1989). Burden and family care of the elderly. In S. J. Bahr, & E. T. Peterson (Eds.), Aging and the family (pp ). Lexington, MA: Lexington Books. Barrera, M., Sandler, I & Ramsey, T. (1981). Preliminary development of a scale of social supports of college students. American Journal of Community Psychology, 9 (4), Brody, E. M. (1985). Parent care as normative family stress. The Gerontologist, 25(1), Clark, M.C., & Huttlinger, K. (1998). Elder Care among Mexican American Families. Clinical Nursing Research, 7( 1), Cox, C. & Monk, A. (1996). Strain among caregivers: Comparing the experiences of African American and Hispanic caregivers of Alzheimer's relatives. International Journal of Aging and Human Development, 43, Dietz, T. (2002). Intergenerational assistance within the Mexican American family. In N. Benokraitis (Ed.), Cotemporary Ethnic Families in the United States: Characteristics, variations and dynamics (pp ). Upper Saddle River, NJ: Prentice Hall. Dilworth-Anderson, P., Williams, I. C., & Gibson, B. E. (2002). Issues of race, ethnicity, and culture in caregiving research: A 20-year review ( ). The Gerontologist, 42(2), Escobar, J. & Randolph, E. (1982). The Hispanic and social networks. In R. Becerra, J. Kamo & 1. Escobar (Eds.), Mental Health and Hispanic Americans: Clinical perspectives (pp ). New York: Grune & Stratton. Gallagher-Thompson, D., Coon, D. W., Rivera, P., Powers, D., & Zeiss, A. M. (1998). Family caregiving: Stress, coping, and intervention. In M. Hersen & V. B. Van Hasselt (eds.), Handbook of Clinical Geropsychology (pp ). New York: Plenum Publishing Corporation. Gouldner, A. W. (1960). The Norm of Reciprocity: A Preliminary Statement. American Sociological Review 25, Guralnik, J.M. & Simonsick, E. M. (1993). Physical disability in older Americans. Journal of Gerontology, 8 (Special Issue): Haley, W.E., Roth, D.L., Coleton, M.I., Ford, G.R., West, C.A.C., Collins, R.P., & Isobe, T.L. (1996). Appraisal, coping, and social support as mediators of well-being in Black and White family caregivers of patients with Alzheimer's disease. Journal o.fconsulting and Clinical Psychology, 64, Jenkins, J. & Karno, M. (1992). The meaning of expressed emotion: Theoretical issues raised by cross-cultural research. American Journal of Psychiatry, 149( 1), Katz, S., Ford, A., Moskowitz, R., Jackson, B., & Jaffee, M. (1963). Studies of illness in the aged. The index of ADL, a standardized measure of biological and psychosocial function. Journal of the American Medical Association. 185, Kleinman, A. (1980). Patients and healers in the context of culture. Berkley, CA: University of California Press. Krause, N. (1995). Negative interaction and satisfaction with social support among older adults. Journal of Gerontology: Psychological Sciences, 50B, Kulis, S. S. (1992). Social class and the locus of reciprocity in relationships with adult children. Journal of Family Issues. 13(4), Landrine, H. (1992). Clinical implications of cultural differences: The referential versus the indexical self. Clinical Psychology Review, 12(4), Lawton, M. P., & Brody, E. (1969). Assessment of older people: Self-maintaining and instrumental activities ofdaily living. The Gerontologist. 9, Lawton, M. P., Kleban, M. H., Moss, M., Rovine, M., & Glicksman, A. (1989). Measuring caregiving appraisal. Journal o/gerontology. 44, The Southwest Journal 011 Aging: Vol. 19,

21 Lawton, M. P., Rajagopal, D., Brody, E., & Kleban, M. H. (1992). The dynamics of care giving for a demented elder among black and white families. Journal of Gerontology, 47, SI56-S164. Lawton, M. P., Moss, M., Kleban, M.H., Glicksman, A., & Rovine, M. (1991). A two-factor model of caregiving stress and psychological well-being. Journal of Gerontology: Psychological Sciences 46, Lockery S. A. (1991). Caregiving among racial and ethnic minority elders. Generations, 15(4), Lubben J. E. (1988). Assessing social network among elderly populations. Family and Community Health, J J, Markides. K. S., & Krause, N. (1985). Intergenerational solidarity and psychological well-being among older Mexican-Americans: A three generation study. Journal of Gerontology; 40(3), [Not cited in the text]mintzer, 1. E., Rubert, M.P., Loewenstein, D., Gamez, E., Millor, A., Quinteros, R., Flores, L., Miller, M., Rainerman, A., & Eisdorfer C. (1992). Daughters caregiving for Hispanic and non-hispanic Alzheimer patients: Does ethnicity make a difference? Community Mental Health Journal, 28 (4), Mutran, E. (1985). Intergenerational Family Support Among Blacks and Whites: Response to Culture or to Socioeconomic Differences. Journal of Gerontology, 40, Neal, M., Ingersoll-Dayton, B., & Starrels, M. E. (1997). Gender and relationship differences in caregiving patterns and consequences among employed caregivers. The Gerontologist, 37, Newsom, J. T., & Schulz, R. (1996). Social support as a mediator in the relation between functional status and quality of life in older adults. Psychology and Aging, 11, Pearlin, L. I., Mullan, J., Semple, S., & Skaff, M. M. (1990). Caregiving and the stress process: an overview of concepts and their measures. The Gerontologist, 30, Pearlin, L.I., & Schooler, C. (1978). The structure of coping. Journal of Health and Social Behavior, 19, Pearlin, L. L & Zarit, S. H. (1993). Research into informal caregiving: Current perspectives and future directions. In S. Zarit and L. Pearlin (eds.), Caregiving systems: Formal and informal helpers (pp ). Hillsdale, New Jersey: Lawrence Erlbaum Associates. Phillips, L. R., de Ardon, E. T., Komnenich, P., Killeen, M., Rusinak, R. (2000). The Mexican American caregiving experience. Hispanic Journal of Behavioral Sciences, 22(3), Posner, B. M., Jette, A. M., Smith, K. W., & Miller, D. R. (1993). Nutrition and health risks in the elderly: The Nutrition Screening Initiative. American Journal of Public Health, 83, Radloff, L. S. (1977). The Center for Epidemiological Studies-Depression Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 3, Rubert, M. & Dry, M. (1998, November). Four different interventions with minority caregivers: Clinical issues from the Resources for Alzheimer's Caregiver Health (REACH) project. Symposium conducted at the Gerontological Society of America, Philadelphia, PA. Schulz, R., Newsom, J., Mittelmark, M., Burton, L. Hirsch, C., & Jackson, S. (1997). Health effects of caregiving: The Caregiver Health Effects Study: An ancillary study of the Cardiovascular Health StUdy. Annals of Behavioral Medicine, 19, Skaff, M. M., Pearlin, L. L, & Mullan, 1. T. (1996). Transitions in the caregiving career: Effects on sense of mastery. Psychology of Aging, 11(2), Spector, W.D. (1990). Functional disability scales. In B. Spilker, (Ed.)~ Quality of l~fe assessments in clinical trials (pp ). New York: Raven Press~ Ltd. Spielberger, C. D.~ Gorsuch, R. L., Lushene, P. R., Vagg, P. R.~ & Jacobs, G. A. (1983). State-Trait Anxiety Inventory for Adults (STAIS-AD) Manual. Palo Alto, CA: Consulting Psychologists Press, Inc The Southwest Journal on Aging: Vol. 19,