Differences in the Perceived Well-Being of Wives and Husbands Caring for Persons with Alzheimer's Disease 1

Copyright 1998 by The Cerontological Society of America The Gerontologist Vol. 38, No. 2, 224-230 This study aimed to examine differences in subjective psychological well-being between husband and wife caregivers of persons with Alzheimer's disease in comparison to the psychological states of noncaregiver husbands and wives similar in age and ethnicity. The principal comparison of interest was whether differences between husband and wife caregivers were greater than those between comparison group husbands and wives. Significant interactions for six out of nine psychological measures indicate that genderrelated differences between spouse caregivers were specifically associated with the caregiving role. Possible explanations for the discrepancy between husband and wife caregivers as well as practice implications are discussed. Key Words: Emotional health, Spouse caregivers, Gender differences Differences in the Perceived Well-Being of Wives and Husbands Caring for Persons with Alzheimer's Disease 1 Sharon K. Rose-Rego, 2 Milton E. Strauss, 3 and Kathleen A. Smyth 4 Caring for a family member with dementia can be a source of considerable burden. Not surprisingly, spouses appear to be more adversely affected than adult children caring for a parent (George & Cwyther, 1986; itting, Rabins, Lucas, & Eastham, 1986), particularly in areas such as social activities and physical health (Barber & Pasley, 1995). Caregiver wives seem to be more at risk than caregiver husbands. They report higher levels of depression and nonspecific physical symptoms such as headaches, dizziness, and sleep problems than do husbands caring for demented spouses (Pruchno & Potashnik, 1989; Pruchno & Resch, 1989a). Lower levels of well-being (Borden & Berlin, 1990) and more subjective burden (Pruchno & Resch, 1989a) are also reported by caregiving wives. A metaanalysis of gender differences in the caregiving literature (Miller & Cafasso, 1992) revealed that greater burden reported by caregiving females is robust, albeit the effect size is modest. emales were also found to be more involved in personal care activities and household chores than males. Unfortunately, Miller and Cafasso (1992) were unable to compare roles (spouse or adult child) within gender due to the small number of studies used in the meta-analysis. The psychological effects of caring for persons with dementia were first elaborated through comparisons 1 A portion of this work was presented at the 48th Annual Scientific Meeting of the Cerontological Society of America, Los Angeles, CA, November 1995. This research was supported by NIA Crant #AG08012. The authors acknowledge the help or Stuart Quirk. 2 Address correspondence to Sharon Rose-Rego, Department of Psychology, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, Ohio 44106-7123. E-mail: skr6@po.cwru.edu department of Psychology, Case Western Reserve University and Alzheimer Center of University Hospitals and Case Western Reserve University School of Medicine, Cleveland, OH. 4 Alzheimer Center of University Hospitals and Case Western Reserve University School of Medicine, Department of Epidemiology and Biostatistics, Cleveland, OH. of caregivers with community-based population norms (George & Cwyther, 1986; Pruchno & Potashnik, 1989). Another line of research looked at comparisons among different kinds of caregivers. That is, comparisons were made between subgroups of caregivers classified in terms of role relationship (i.e., spouses compared to adult children) (Barber & Pasley, 1995; George & Gwyther, 1986), residence pattern (i.e., co-resident caregivers compared to caregivers of nursing home residents) (Barber, 1993; Stephens, Kinney, & Ogrocki, 1991), or gender-related role (i.e., caregiver wives and daughters compared to caregiver husbands and sons) (Barber & Pasley, 1995). These studies of variation among caregivers generally did not use comparison groups of persons who were not caregivers. Consequently the extent to which differences between husbands and wives (or spouses and children) in mental health and social participation are specifically a function of caregiving is not known. We are aware of only two studies that compared husbands and wives who were caregivers of impaired spouses with demographically comparable husbands and wives who did not have such responsibilities. In one study, Dura, Stukenberg, and Kiecolt-Glaser (1990) compared current and lifetime rates of depressive disorders in 86 spouse caregivers and 86 matched controls. A greater proportion of caregivers experienced a depressive disorder throughout the years they were providing care compared to the matched controls. However, these researchers did not specifically examine gender differences. In addition, over onethird of the controls were not spouses. In a second preliminary study, Dura, Haywood-Niler, and Kiecolt- Glaser (1990) evaluated the severity of depression of spouse caregivers of persons with Alzheimer's disease (AD), persons with Parkinson's disease (PD), and couples in which neither partner was impaired. Although the focus of the study was on differences 224 The Gerontologist

between caregivers of AD and PD patients, caregiver gender was also included in the analysis. These investigators found that differences between husbands and wives of AD or PD patients in Hamilton Depression Rating Scale scores, a measure of clinical depression, were not greater than husband-wife differences in the comparison group. It is well documented that women report higher levels of depressive symptoms than do men, at least in younger populations (Nolen-Hoeksema, 1987). So one might conclude from Dura Haywood-Niler and Kiecolt-Glaser's (1990) study that the greater psychological distress of caregiving wives relative to husbands is mainly a gender effect, not specific to differences in role expectations or demands or support services used by husband and wife caregivers. However, as the authors noted in the title of their report, this was a "preliminary comparison," and the power to detect a gender role by caregiver role interaction was very low. The number of husbands and wives studied in each group (7 and 16, respectively) was small and unequal, both of which adversely affect statistical power. Wives in the caregiver role face more demands than do husbands (Miller & Cafasso, 1992). Wife caregivers also tend, more than male spouse caregivers, to use emotion-focused coping methods (Pruchno & Resch, 1989b; Rose, Strauss, Neundorfer, Smyth, & Stuckey, 1997), which are associated with higher levels of distress (Rose et al., 1997). Women caregivers are also more likely to express negative emotions, perhaps because of their greater attentiveness to them (Lutzky & Knight, 1994). The purpose of the present study was to test two hypotheses drawn from the above considerations. Hypothesis 1. - Caregiver wives would report poorer emotional health, social functioning, life satisfaction, physical health, and lower levels of mastery than caregiver husbands. Hypothesis 2. - Caregiver wives would differ more from caregiver husbands than would noncaregiver wives differ from noncaregiver husbands. Method Subjects The participants in this study were 99 spouses (61 wives, 38 husbands) who were currently living with a spouse with Alzheimer's disease and 113 spouses (74 wives, 39 husbands) who were living with a neuropsychologically and psychiatrically healthy spouse. The caregiver spouses were participants in the Caregiver Core of the Alzheimer Disease Research Center (ADRC) at University Hospitals of Cleveland Case Western Reserve University, as were the noncaregiver spouses. The caregiver participants were all enrollees in the Caregiver Core of the ADRC at the time this study was conducted. All caregivers were residing with a person with Alzheimer's disease. Spouses who were not caregivers were referred from a collateral case-control study at the ADRC to serve as a comparison group in studies conducted by the ADRC Caregiver Core. The comparison group was not providing formal or informal care to family members. Twenty-eight of the 113 persons in the noncaregiving group (25%) were married to each other. Spouses with Alzheimer's disease (AD) were diagnosed according to the DSM-III-R (American Psychiatric Association, 1987) and the NINCDS-ADRDA (McKhann, Drachman, olstein, Katzman, Price, & Stadlan, 1984) criteria for the clinical diagnosis of probable AD. Procedure Spouses of persons with Alzheimer's disease and spouse controls completed a series of questionnaires as part of baseline data collection of the Alzheimer Center Caregiver Core. Instruments describing their emotional health, social functioning, sense of mastery, global life satisfaction, and overall physical health were completed by mail. Measures Emotional health was assessed with two scales. Depression was measured by the Center for Epidemiological Studies-Depression Scale (CES-D; Radloff, 1977), a 20-item scale that asks subjects to rate how often in the past week they have felt a certain way (i.e., "I thought my life had been a failure"). The CES- D has been used previously to measure depression in older adults and has been shown to have adequate reliability and validity in this population (Radloff & Teri, 1986). Internal consistency reliability (Coefficient a) in this study was.91. The second measure, the Positive and Negative Affect Schedule (PANAS: Watson, Clark, & Tellegen, 1988) asks subjects to what extent they have experienced ten positive and ten negative emotions. Reliability for the PA and NA scales in a general adult population were.86 and.87, respectively (Watson, Clark, & Tellegen, 1988). Coefficient a was.91 for both scales. Social functioning was assessed with three measures. Expressive support was measured using an 8- item scale (Pearlin et al., 1990) which asks subjects to indicate on a 4-point Likert scale the degree to which they agree or disagree with statements like, "There is really no one who understands what you are going through" and "You have at least one friend or relative you can really confide in." The measure had reliability in this study (a =.89) comparable to that reported by Pearlin et al. (1990) (a =.87). Social ties were assessed by asking subjects to indicate the number of neighborhood contacts and close friends they had, their degree of group or organizational involvement, and the extent of their social contacts (see Donald & Ware, 1984). Reliability for this measure appears not to have been reported previously and in this study was relatively low (a =.64). Satisfaction with social activities was assessed using one question: "Overall how satisfied are you with the amount of social activities in which you are involved?" Subjects were asked to indicate their satisfaction on Vol. 38, No. 2,1998 225

a 4-point Likert scale. This item has been used previously in caregiver populations (Stuckey, Neundorfer, & Smyth, 1996). Mastery was assessed using Pearlin and Schooler's (1978) 7-point scale which asks subjects to what extent they feel in control of their life circumstances. This measure had good reliability in this study (a =.80) as it has in other studies (Pearlin, Menaghan, Lieberman, & Mullen, 1981; Pearlin & Schooler, 1978). Life satisfaction was measured using the Satisfaction with Life Scale (SWLS: Diener, Emmons, Larsen, & Griffin, 1985). This five-item scale asks subjects to rate their agreement with statements such as "In most ways my life is close to my ideal." High internal consistency reliability for this measure was reported by Diener and colleagues (1985), and found here as well (a =.87). Overall Physical Health was assessed by a one-item, 5-point Likert question: "In general would you say your health is... (poor to excellent)?" Such selfreport assessments of health have been shown to be consistent with a physician's clinical appraisal of health (Stahl, 1984; Bush, Miller, Golden, & Hale, 1989). Self-reports of physical health are also significantly correlated with mortality (Idler & Kasl, 1991). Analysis Multivariate analyses of variance (MANOVA) was used to compare husbands and wives within caregiver groups for the sets of emotional health and social functioning measures separately. Univariate analyses were conducted for individual scales to evaluate significant multivariate effects. Only univariate analyses of variance were conducted for the mastery, life satisfaction, and physical health constructs, each of which was assessed by only one measure. Numbers of subjects vary across analyses because of missing data. Results Characteristics of Caregivers The caregivers averaged approximately 70 years of age and were predominantly Euro-American. The four subgroups did not differ in age or ethnicity (see Table 1). Educational levels ranged from less than high school to some post-baccalaureate work, and was not comparably distributed across the four subject groups (see Table 1). Separate chi-square tests for husbands and wives indicated that caregiver and noncaregiver husbands had similar levels of education, whereas caregiver wives were significantly less educated than noncaregiver wives. The average impairment level of spouse care recipients, assessed by the Clinical Dementia Rating Scale (Hughes, Berg, Danziger, Coben, & Martin, 1982) was between mild and moderate {M = 1.67) and ranged from mild (1) to severe (3). There was no difference in spouse impairment between husband {M = 1.70) and wife (M = 1.65) caregivers, t = 0.29, p =.77. The means and standard deviations for each of Table 1. Characteristics of Participants Caregivers Controls Husbands Wives Husbands Wives (N = 38) (N = 61) (N = 39) (N = 74) Age (Mean, SO) 71.9(6.1) 69.5(7.6) 72.7(5.9) 72.4(7.9) Ethnicity (% Caucasian) 92% 97% 100% 93% Education 3 ' 15 ' 0 High school or less 31% 65% 28% 38% College 53% 27% 39% 36% Postgraduate 16% 8% 33% 26% "Caregivers vs controls, x 2 (2, N = 211) = 11.31, p <.01. b Caregiver husbands vs control husbands, x 2 (2, N = 77) = 3.32, p >.10. c Caregiver wives vs control wives, x 2 (2, N = 134) = 11.45, p <.01. the dependent variables are shown in Table 2. This table also presents the effect sizes (Cohen's d) for the gender role differences for caregivers and noncaregivers. Conventionally, effect sizes of.20,.50, and.80 are considered small, medium, and large, respectively (Rosenthal & Rosnow, 1991). In all instances, the scores of caregiver wives are in the direction of less wellbeing than the scores of husbands. The differences between spouse caregivers range from small to very large for all measures but positive affect and social ties. The differences between husbands and wives without caregiving responsibilities are small to very small with the exception of positive affect (which is a medium effect) and satisfaction with social activities (which was a small to medium effect size). or both of these measures, wives have more positive scores. The pattern of means in Table 2 suggests an interaction between gender role and caregiver role, as was predicted. This impression was confirmed for six out of nine measures by the results of multivariate and univariate tests. Specifically, the MANOVA of emotional health measures indicated a significant interaction of gender role (husbands vs wives) by caregiver role (caregivers vs noncaregivers), (3, 183) = 5.33, p <.01, as well as significant main effects for both of these roles (gender role, [3, 183] = 6.89, p <.001, and caregiver role, [3, 183] = 17.57, p <.001). ollow-up univariate analyses were conducted to identify main effects and interactions for each variable. In cases where both the main effects and the interaction were significant, both were reported but only the interaction was interpreted. As shown in Table 3, a significant interaction of gender role by caregiver role was revealed for negative affect, as were significant main effects of both gender role and caregiver role. A significant gender role by caregiver role interaction was also found for depression in addition to significant main effects of gender role and caregiver role. Caregivers acknowledged significantly less positive affect than the comparison group. or social functioning (see Table 3), the MANOVA indicated a significant interaction for gender role by 226 The Gerontologist

Table 2. Means, Standard Deviations, and Effect Sizes for Emotional Health, Social unctioning, Mastery, Life Satisfaction and Physical Health Caregivers Controls Husbands Wives d Husbands Wives d Emotional health (N = 189) Positive affect Negative affect Depression Social functioning (N = 203) Expressive support Social ties Satisfaction with social activities Mastery (N = 204) Life satisfaction (N = 204) Physical health (N = 212) 29.22 (8.25) 15.18(4.36) 9.31 (8.09) 25.53(3.15) 14.08(4.10) 2.81 (0.65) 20.50(3.13) 13.97(1.67) 2.68(0.81) 28.95 (8.00) 21.87 (8.98) 16.64 (11.52) 27.13 (4.19) 13.55 (5.12) 2.43 (0.84) 18.37 (3.10) 10.98 (2.87) 2.20 (1.04).03 -.82 -.68 -.41.11.47.65 1.04.40 31.83(7.45) 13.33 (3.50) 4.30 (4.03) 28.31 (3.97) 18.73 (4.71) 3.00 (0.93) 22.95 (3.09) 15.05 (2.00) 2.82 (0.72) 35.59 (6.20) 12.90(3.57) 4.31 (4.84) 28.66 (3.74) 18.87 (4.14) 3.31 (0.72) 22.68 (3.07) 14.90 (2.56) 2.95 (0.72) -.55.12.00 -.09 -.03 -.38.09.06 -.18 caregiver role, (3, 197) = 4.91, p <.01, as well as a main effect for caregiver role, (3, 197) = 20.92, p <.001. The follow-up ANOVAs revealed a significant interaction of gender role by caregiver role, and a significant main effect of caregiver role for satisfaction with social activities. Other significant effects at the univariate level indicated that caregivers reported significantly fewer social ties and lower levels of expressive support than the comparison eroup. There were no significant interactions for either of these measures, smallest p =.21. The ANOVA of the mastery score showed significant main effects for gender role and caregiver role which were not interpreted due a significant gender role by caregiver role interaction (Table 4). The same pattern of results was found for life satisfaction; there was a significant gender role by caregiver role interaction, as well as main effects for gender role and caregiver role. Univariate analyses for overall physical health indicated a significant main effect for caregiver role which was not interpreted due to the presence of a significant gender role by caregiver role interaction. To more clearly characterize the differences between men and women who are in caregiving roles and those who are not, we present in igure 1 the interactions for these comparisons. Each of the six scales for which there were interactions were standardized to eliminate scale differences and the differences in scores between husband and wife participants in both groups are expressed as z-scores. Differences between husband and wife noncaregivers are close to zero whereas differences between caregiver spouses are much larger in all measures, except satisfaction with social ties. In each case, caregiver wives had significantly less favorable scores. The planned comparisons indicated that for negative affect, depression, mastery, life satisfaction, and overall physical health, the difference between caregiver husbands and wives was greater than the difference between noncaregiving husbands and wives (see igure 1 and Table 5). Satisfaction with social ties was the only variable for which differences existed between husband and wife controls. Caregiving wives {M = 2.42) reported less satisfaction than caregiver husbands {M = 2.81), t = 2.57, p <.05; however, wives who were not caring for an infirm spouse (M = 3.31) reported greater satisfaction than noncaregiving husbands (M = 3.00), t= 1.97, p =.05. As previously noted, caregiver wives were less edu- Table 3. Values for Univariate ollow-up Tests for Significant MANOVA Effects 3 Gender Role Caregiver Role Gender x Caregiver Role Interaction Emotional health (N = 189) Positive affect Negative affect Depression Social functioning (N = 203) Expressive support Social ties Satisfaction with social activities 2.49 12.01*** 9.27** 16.80*** 34.80*** 51.10*** 17.46*** 55.99*** 22.15*** (1, 199) 3.26 14.86*** 9.25** 1.55.24 10.83*** "Univariate gender role follow-up tests for social functioning are not reported because the multivariate effect was not significant. *p <.05; **p <.01; ***p <.001. Vol. 38, No. 2,1998 227

Table 4. Values for Univariate Tests for Mastery, Life Satisfaction, and Overall Physical Health Gender Role Caregiver Role Gender x Caregiver Role Interaction Mastery (N = 204) Life satisfaction (N = 204) Overall physical health (N = 212) 7.14** (1,200) 20.00*** (1,200) 57.04*** (1,200) 50.58*** (1,200) 10.75*** (1,208) 4.35* 16.33*** 4.59* (1, 200) (1,200) (1, 208) *p <.05; **p <.01; ***p <.001. cated than noncaregiver wives. Husbands were also more educated than wives, r =.17, p <.05. Education was significantly correlated with depression among noncaregivers, r = -.24, p <.05, and with overall health among both caregivers, r =.24, p <.05, and noncaregivers, r =.27, p <.01. Thus, it was necessary to examine the potential influence of education on the results reported above. Interpretation and validity of covariance adjustment is questionable when a covariate is correlated with the independent variables (caregiver group and gender) as well as the dependent variable (Lord, 1967; Reichardts, 1979). Because of this, we evaluated the possible confounding role of education by comparing caregivers and noncaregivers at each education level. These levels were high school graduate or less (low education, n = 90), college degree (medium education, n = 78), or post-graduate education (high education, n = 43). Caregiver role by gender role ANOVAs were performed at each level of education. The ANOVAs for depression indicated that education had no role in the group differences. There was a main effect of caregiver role at every level of education, low, (1, 78) = 21.65, p <.001, me- in 3 X 0) CO V u I Caregivers (Wives Husbands) o Controls (Wives - Husbands) o o o Negaff Depression Social Mastery Life Sat. Dependent Variables 0 Health igure 1. Mean z-score differences between husband and wife caregivers and controls. Table 5. Planned Comparisons Negative Affect Depression Social Mastery t 14.22 df (3,185) p.001 10.87 (3,185).001 0.54 9.39 (3,199) (3,200) ns.001 Life Satisfaction Health 16.59 4.31 (3, 200) (3, 208).001.001 dium, (1, 70) = 14.36, p <.001, and high, (1, 36) = 25.18, p <.001. However, gender role by caregiver role interactions were also evident at both the medium, (1, 70) = 7.50, p <.05, and high levels of education, (1, 36) = 8.90, p <.01. Consistent with the approach previously used, the difference between depression scores Tor husband and wife caregivers was compared to that of noncaregiver husbands and wives. The difference between caregiver husbands and wives was greater than the difference between noncaregiving husbands and wives at both the medium, t = 9.45, p <.001, and high levels of education, t = 12.77, p <.001. The ANOVAs for overall health revealed a main effect for caregiver role at the low, (1, 86) = 4.36, p <.05, and high level of education, (1, 39) = 4.79, p <.05, but not at the medium level, (1, 74) = 0.25, p =.62. Wives and husbands who were not caring for an impaired spouse reported better overall health than caregiver wives and husbands at the low and high levels of education. No significant main effects of gender nor significant interactions of caregiver role by gender role were present at any of the educational levels. Discussion The results of this study suggest that gender differences in psychological, social and physical well-being are specifically associated with the caregiver role. In this study as in previous studies, caregiver wives reported more negative psychological states than did caregiver husbands. However, differences in wellbeing were not generally found between husbands and wives who did not have an infirm spouse. Scores of spouses without caregiving responsibilities differed appreciably only for satisfaction with social activities. This suggests that there is something about the caregiving role that leads caregiver wives to report more 228 The Gerontologist

detrimental effects than husbands. There are at least three classes of factors that may contribute to the more negative mental health outcomes for caregiver wives. Emotional attentiveness. - One reason caregiving wives may experience and/or report more negative effects than caregiving husbands is that wives are more attentive to their emotions and thus are more likely to report negative effects such as depression. Caregiving husbands may be experiencing the same level of distress but may not recognize it and thus would be less likely to report it. Lutzky and Knight (1994) have found that gender was a significant predictor of depression, psychological symptoms, and burden in spouse caregivers of persons with dementia. However, once neuroticism (used to measure emotional attentiveness) was entered into the regresssion equation, gender was no longer a significant predictor of depression or psychological symptoms. The emotional attentiveness hypothesis suggests that it is not the caregiving experience itself, but rather the wives' recognition of this distress that produces negative psychological effects. However, this explanation raises two measurement issues: one, are the measures of distress in this study not assessing "real" burden, only attentiveness; or two, are the measures missing real burden in males because they are not attending to their emotions? Both of these issues need to be addressed in future studies. Coping styles. - Differences in coping styles between husband and wife caregivers also may contribute to the differences in psychological effects of caregiving. Compared to husbands, caregiving wives have been found to use more emotion-foe used coping strategies which are associated with higher levels of distress (Pruchno & Resch, 1989b; Rose et al., 1997). Lutzky and Knight (1994) also found some support for differences in coping styles: escape-avoidance and seeking social support coping styles explained a significant amount of variance in gender differences in depression and psychological symptoms. Zarit, Todd, and Zarit (1986) found caregiver wives reported greater burden during an initial interview, yet there were no gender differences at a two-year follow-up interview. The interviewers reported that caregivers wives appeared to use the same type of coping strategies as the caregiver husbands at the follow-up interview; however, this was based on interviewer impression rather than formal assessment of coping strategies. Caregiving tasks. - Differences in caregiving tasks might also explain the greater dissatisfaction of wives as compared to husbands. Miller and Cafasso's (1992) meta-analysis found that female caregivers perform more personal care and household chores than male caregivers do. These tasks may be more time consuming, demanding, and ongoing than traditionally male tasks, such as money management. Noelker and Wallace (1985) examined gender differences among caregivers who were currently living with the care recipient. They found that male caregivers receive more help from family members and other informal services than female caregivers do (Noelker & Wallace, 1985). This may leave caregiver wives feeling overwhelmed and thus likely to report more distress. It should be noted that there is an association between education level and overall health. Differences between caregivers and noncaregivers in perceived overall health were found for only respondents with less than college experience and more than college graduation. Caregivers with some college or a college degree did not differ from noncaregivers. This nonlinear relationship was unanticipated and does not have an obvious interpretation. It is necessary to examine further the possibly moderating effects of education on the perceived health of caregivers. The present study adds to the caregiver literature by demonstrating that gender differences in wellbeing among spouse caregivers are specifically associated with the caregiving role. However, a number of questions remain. The duration and specific activities of caregiving were not assessed in this study, nor were relations between well-being and the cognitive and functional capacities of the care recipients. Zarit, Todd, and Zarit's (1986) preliminary work suggests gender differences in burden may disappear over time, and this question merits additional inquiry. The negative impact of caregiving on wives emphasizes the necessity of services tailored to meet their needs. Linking caregiver wives to support groups would provide an outlet for emotionally distressed caregivers. Easily accessible and affordable respite services could help decrease the daily, demanding workload of caregiving wives. inally, educational programs could alert wives to the negative impacts of caregiving and encourage their use of services to help prevent additional burden. Coping styles are modifiable (Keefe, 1988; Lovett, utterman, Gallagher, & Rose, 1989), and so such educational programs might be used to assist caregiving wives to rely more on problemfocused coping strategies which are associated with less distress (Rose et at, 1997). Two limitations of the present study warrant specific mention. The noncaregiving group included 14 couples; the data were treated as though from 28 individuals. Because of assortative mating and the influences of marriage on spouses, the data from each person in a couple are not truly independent. The inclusion of couples may have reduced differences between husband and wives. The dfs for the numerator in the MANOVAs are also somewhat inflated by the inclusion of some scores that are not fully independent of others. However, reducing the dfs to eliminate such pairs does not change the p-values for the significance tests. Nonetheless, replication of these results with caregiving samples that do not include married couples would be useful. uture studies also need to include more representative samples of both caregivers and noncaregivers. Spouses in this study were highly educated and mainly Caucasian which limits the generalizability of the findings. Connell and Gibson's (1997) recent re- Vol. 38, No. 2,1998 229

view suggests that minority caregivers may report lower levels of stress, burden, and depression. Minority caregivers were also found to rely more on their religious faith in coping with caregiving, a set of mechanisms associated with fewer depressive symptoms in other older adult populations (Koenig, Conen, Blazer, et al., 1992). These findings suggest that the pattern of greater differences in well-being between caregiving husbands and wives and the lower reported wellbeing among caregivers may be attenuated in non- Euro-American cultural groups, particularly those in which faith-based coping strategies are common. References American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author. Barber, C. E. (1993). Spousal care of Alzheimer's disease patients in nursing home versus in-home settings: Patient impairment and caregiver impacts. Clinical Gerontologist, 12, 3-30. Barber, C. E., & Pasley, B. K. (1995). amily care of Alzheimer's patients: The role of gender and generational relationship on caregiver outcomes. Journal of Applied Gerontology, 14, 172-192. Borden, W., & Berlin, S. (1990). Gender, coping, and psychological wellbeing in spouses of older adults with chronic dementia. American Journal of Orthopsychiatry, 60, 603-610. Bush, T. L, Miller, S. R., Golden, A. L, & Hale, W. E. (1989). Self-report and medical record report agreement of selected medical conditions in the elderly. American Journal of Public Health, 79, 1554-1556. Connell, C. M., & Gibson, G. D. (1997). Racial, ethnic, and cultural differences in dementia caregiving: Review and analysis. Cerontologist, 37, 355-364. Diener, E., Emmons, R. A., Larsen, R. J., & Griffen, S. (1985). The Satisfaction with Life Scale. Journal of Personality Assessment, 49, 71-75. Dura, J. R., Haywood-Niler, E., & Kiecolt-Glaser, J. K. (1990). Spousal caregivers of persons with Alzheimer's and Parkinson's disease dementia: A preliminary comparison. Cerontologist, 30, 332-336. Dura, J. R., Stukenberg, K. W., & Kiecolt-Claser, J. K. (1990). Chronic stress and depressive disorders in older adults. Journal of Abnormal Psychology, 99, 284-290. itting, M., Rabins, P., Lucas, M. J., & Eastham, J. (1986). Caregivers for dementia patients: A comparison of husbands and wives. Cerontologist, 26, 248-252. Donald, C. A., & Ware, J. E. (1984). The measurement of social support. Research in Community and Mental Health, 4, 325-370. George, L. K., & Gwyther, L. P. (1986). Caregiver well-being: A multidimensional examination of family caregivers of demented adults. Cerontologist, 26, 253-259. Hughes, C. P., Berg, L., Danziger, W. L., Coben, L. A., & Martin, R. L. (1982). A new clinical scale for the staging of dementia. British Journal of Psychiatry, 140, 566-572. Idler, E. L., & Kasl, S. (1991). Health perceptions and survival: Do global evaluations of health status really predict mortality? Journal of Gerontology: Social Sciences, 46, S55-65. Keefe, T. (1988). Stress-coping skills: An ounce of prevention in direct practice. Social Casework, October, 475-482. Koenig, H. G., Cohen, H. J., Blazer, D. G., Pieper, C, Meador, K. C., Shelp,., Coli, V., & DiPasquale, B. (1992). Religious coping and depression among elderly, hospitalized medically ill men. American Journal of Psychiatry, 149, 1693-1700. Lord,. M. (1967). A paradox in the interpretation of group comparisons. Psychological Bulletin, 68, 304-305. Lovett, S., utterman, A., Gallagher, D., & Rose, J. (1989). Changes in coping behavior following psychoeducational interventions for caregivers. Paper presented at the 42nd Annual Scientific Meeting of the Gerontological Society of America, Minneapolis, MN. Lutzky, S. M., & Knight, B. G. (1994). Explaining gender differences in caregiver distress: The roles of emotional attentiveness and coping styles. Psychology and Aging, 9, 513-519. McKhann, G., Drachman, D., olstein, M., Katzmen, R., Price, D., & Stadlan, M. (1984). Clinical diagnosis of Alzheimer's disease: Report of the NINCDS-ADRDA work group under the auspices of Department of Health and Human Services Task orce on Alzheimer's Disease. Neurology, 34, 939-944. Miller, B., & Cafasso, L. (1992). Gender differences in caregiving: act or artifact? Cerontologist, 32, 498-507. Noelker, L. S., & Wallace, R. W. (1985). The organization of family care for impaired elderly. Journal of amily Issues, 6, 23-44. Nolan-Hoeksema, S. (1987). Sex differences in unipolar depression: Evidence and theory. Psychological Bulletin, 101, 259-282. Pearlin, L. I., Menaghan, E. G., Lieberman, M. A., & Mullen, J. T. (1981). The stress process. Journal of Health and Social Behavior, 22, 337-356. Pearlin, L. I., Mullan, J. T., Semple, S. J., & Skaff, M. M. (1990). Caregiving and the stress process: An overview of concepts and their measures. Cerontologist, 30, 583-594. Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journal of Health and Social Behavior, 19, 2-21. Pruchno, R. A., & Potashnik, S. L. (1989). Caregiving spouses: Physical and mental health in perspective. Journal of the American Geriatrics Society, 37, 697-705. Pruchno, R. A., & Resch, N. L. (1989a). Husbands and wives as caregivers: Antecedents of depression and burden. Gerontologist, 29, 159-165. Pruchno, R. A., & Resch, N. L. (1989b). Mental health of caregiving spouses: Coping as a mediator, moderator, or main effect? Psychology and Aging, 4, 454-463. Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 3, 385-401. Radloff, L. S., & Teri, L. (1986). Use of the Center for Epidemiological Studies-Depression Scale with older adults. Clinical Cerontologist, 5, 119-136. Reichardts, C. S. (1979). The statistical analysis of data from nonequivalent group designs. In T. D. Cook, & D. T. Campbell (Eds.), Quasiexperimentation (pp. 147-170). Chicago: Rand McNally. Rose, S. K., Strauss, M. E., Neundorfer, M. M., Smyth, K. A., & Stuckey, J. C. (1997). The relationship of self-restraint and distress to coping among spouses caring for persons with Alzheimer's disease. Journal of Applied Gerontology, 16, 91-103. Rosentnal, R., & Rosnow, R. L. (1991). Essentials of behavioral research: Methods and data analysis. New York: McGraw Hill. Stahl, S. (1984). Health. In D. Mangen & W. Peterson (Eds.), Research instruments in social gerontology: Volume 3, Health program evaluation and demography. Minneapolis: University of Minnesota Press. Stephens, M. A. P., Kinney, J. M., & Ogrocki, P. K. (1991). Stressors and well-being among caregivers to older adults with dementia: The inhome versus nursing home experience. Gerontologist, 31, IM-HZ. Stuckey, J. C, Neundorfer, M. M., & Smyth, K. A. (1996). Burden and well-being: The same coin or related currency? Gerontologist, 36, 686-693. Watson, D. Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS Scales. Journal of Personality and Social Psychology, 54, 1063-1070. Zarit, S. H., Todd, P. A., & Zarit, J. M. (1986). Subjective burden of husbands and wives as caregivers: A longitudinal study. Cerontologist, 26, 260-266. Received October 16, 1996 Accepted October 30, 1997 230 The Gerontologist