Article

Social Support, Mastery, and Psychological Distress in Black and White Older Adults

The International Journal of Aging and Human Development 0(0) 1–20 ! The Author(s) 2016 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/0091415015627161 ahd.sagepub.com

Ruth T. Morin1 and Elizabeth Midlarsky1

Abstract Social support and mastery can protect against psychological distress in late life, carrying implications for theory and intervention. However, some groups have not been well studied, with African Americans receiving less empirical attention, especially in regard to their satisfaction with social support. In this study, samples of African American and White American community-dwelling older adults reported their perceived mastery, degree of psychological distress, and social support. A model investigating the separate relationships of these variables by race explained significantly more variance than a model for all participants combined. For both groups, mastery was significantly associated with lower psychological distress. However, among White Americans, social support was significantly associated with lower distress, while among African Americans, there was no relationship between satisfaction with social support and distress. The findings indicate that social support and mastery are important variables to consider in their relationship to psychological distress in later life and that diverse racial groups may display differing relationships among these variables. Keywords social support, mastery, late life, race, mental health

1

Teachers College, Columbia University, New York, NY, USA

Corresponding Author: Elizabeth Midlarsky, Teachers College, Columbia University, 525 West 120th Street Box 303, New York, NY 10027, USA. Email: [email protected]

2

The International Journal of Aging and Human Development 0(0)

With the U.S. population aging at a rapid rate, there is an increasing interest in identifying factors that buffer older adults from psychological distress. In addition to the awareness of the population aging, there will continue to be increasing racial and ethnic diversity in cohorts of aging adults. It is more important than ever to view late life in terms of the experiences and trajectories of diverse groups of aged adults, including African Americans, who are likely to be differentially affected by historical and political treatment. The life course perspective supports this approach, arguing that historical circumstances encountered early in life will be related to experiences and psychological distress in late life, including older adults’ sense of mastery and support (Harevan, 1994). Social support appears to be important for adjustment and healthy coping across the life span, particularly in late life (Steverink & Lindenberg, 2006). Although often associated with cognitive and physical impairments, it has been shown to be salient in the context of both heightened emotional distress (Mikulincer & Florian, 1997) and aging-related stressors (Krause, 2007). Research has demonstrated that an active and social lifestyle is associated with higher levels of physical and mental health for older adults (Blazer, 2005) and that, conversely, social isolation and inadequate social support are associated with depressive symptomology in the elderly (Bruce, 2002). Mastery, like perceived social support, appears to be of central importance in understanding variations in adjustment in late life (Pearlin, Nguyen, Schieman, & Milkie, 2007). Mastery is conceptualized as the extent to which one perceives that one is able to effectively manage life circumstances (Pearlin & Schooler, 1978). The literature on mastery, and on the related construct of self-efficacy, shares the theoretical premise that psychological distress is not necessarily a function of challenging life events but rather of the perceived inability to manage them effectively (Bandura & Locke, 2003). From this perspective, higher levels of mastery and self-efficacy promote flexibility in adapting to the increased functional limitations associated with aging (Blazer, 2002). In attempting to understand how these variables operate in late life, especially as they may protect against or lead to clinically significant psychological distress, theoretical models have emphasized the positive relationship between perceived social support and mastery (Krause, 2005; Lincoln, Chatters, & Taylor, 2003; Lincoln et al., 2010), though the findings related to social support have not been consistent (Antonucci, Akiyama, & Lansford, 1998). There is a heightened probability of needing support in late life, and it has been demonstrated that when one finds that support is unavailable, feels dependent on support, or experiences negative interactions with members of one’s support system, one may experience significant distress (Lincoln et al., 2010). It is necessary, therefore, to clearly define and measure social support, in order to understand how it may relate to psychological distress in late life.

Morin and Midlarsky

3

Perceived Social Support, Mastery, and Race There is a growing comprehension in the field that the process of aging may differ in accordance with race, ethnicity, and culture. The influence may be most pronounced for older members of minority groups who have been deprived of resources relative to older White Americans (Hayward, Miles, Crimmins, & Yu, 2000). Some studies investigating race and mastery have indicated that compared to White Americans, African Americans report lower self-efficacy and perceived control (Ajrouch, Antonucci, & Janevic, 2001), while others have found that African Americans experience themselves to be more responsible for both causes of problems and their solutions than do White Americans (Nemeroff & Midlarsky, 2000). Some researchers contend that older adults of minority status who have been chronically exposed to stress may believe that life circumstances are not amenable to change (Pearlin et al., 2007). Others argue that minority group members become more resilient and better prepared to handle stress as a result of their experience with hardships (DilworthAnderson & Burton, 1999). In any event, there is an extensive literature documenting the relationship between chronic stress and psychological distress, particularly in African Americans (Bruce, 2002). Studies that examined differences in perceived social support present an equally complex picture (Hinrichsen & Ramirez, 1992), with an increasing number of studies indicating racial differences in the sources of social support among older adults (Chang, Chen, & Alegria, 2014). Research results suggest that the social support networks of African Americans often consist of faithbased communities rather than secular support networks in contrast to White Americans, who more often report larger secular networks (Krause, 2005). African Americans reportedly have smaller networks but with more family members and a higher frequency of contact than the White Americans surveyed (Ajrouch et al., 2001). Furthermore, African Americans are more likely than White Americans to value helping and being helped by extended family members (Dilworth-Anderson & Burton, 1999). Krause (2005) found that a faith-based social network buffered stress reactions for older African Americans but not for their White American counterparts. There is also evidence that patterns of familial social support may differ by race, with a recent study finding that middle-aged African Americans tend to give support to their elderly parents, whereas middle-aged White Americans are more likely to support their grown children, suggesting possible system-level family differences among racial groups (Fingerman, VanderDrift, Dotterer, Birditt, & Zarit, 2011). It is possible that these differences in support networks between White American and African American older adults (specifically in amount of faithbased and family support) help explain the lower levels of reported psychological distress among African Americans, even when disparities in socioeconomic

4

The International Journal of Aging and Human Development 0(0)

status persist (Harris, Edlund, & Larson, 2005). These differences in perceived social support between White American and African American older adults have been found to be related to psychological distress, regardless of the level of mastery (Nemeroff, Midlarsky, & Meyer, 2010). Other studies have found mastery to be directly related to distress among White American older adults, while social support seems to be more related to psychological distress among African American older adults (Lincoln et al., 2003). Indeed, other researchers have looked at mastery as a mediating variable between social support and psychological distress (Krause, 1987). As many of these studies are cross-sectional in nature, it is impossible to determine the direction of causation among social support, mastery, and psychological distress. Although there have been many studies conducted on the relationships among perceived social support, mastery, and psychological distress (both in clinical and community samples), there has been a paucity of studies investigating these relationships in racial and ethnic subgroups of older adults experiencing some degree of psychological distress. Certainly, there have been methodological constraints in the ability to conduct the analyses; for example, the desire to employ a representative random sample, which typically yields too small a number of minority participants for meaningful analysis. It is of prime importance to discover whether variables like perceived social support and mastery function differently among diverse groups of older adults who experience psychological distress. Nevertheless, it is important to study the relationships among the variables in these groups because of the reported disparity in access to mental health services between White Americans and African Americans (Alegria et al., 2002; Harris et al., 2005).

The Current Study This investigation was designed to explore the relationships among mastery, perceived social support, and psychological distress among White American and African American older adults referred for outpatient mental health services. Social support and mastery were examined because of prior findings indicating that they may be inversely associated with psychological distress in late life (e.g., Krause, 2005; Lincoln et al., 2003; Nemeroff & Midlarsky, 2000). The primary aim was to identify the patterns of relationships among these variables for the sample as a whole and to separately investigate the models for White American and African American participants. It was hypothesized that mastery is related to psychological distress similarly across races, such that for both groups, greater perceived mastery is associated with lower levels of psychological distress (Nemeroff et al., 2010). On the other hand, social support was expected to be related to distress for White American older adults, with satisfaction with social support inversely related to psychological distress. This relationship was not expected to be present for African

Morin and Midlarsky

5

American older adults. That is, support tends to be stronger in African American families and faith-based communities, and there is a lack of relationships between emotional support and diagnoses of anxiety and depression in African American older adults (Lincoln et al., 2010). We therefore predicted that satisfaction with social support is not significantly related to psychological distress in later life among African American older adults. However, it is possible that the association between mastery and psychological distress may be mediated by social support, and this relationship may differ by racial group. The current analysis seeks to make sense of findings that social support is associated with psychological distress over and above mastery in both White American and African American older adults (e.g., Nemeroff et al., 2010). It is possible that although mastery has been found to be strongly associated with distress, the strength of this association may vary with perceptions about support. Because of the difficulties associated with the increased need for support and help in late life, it is important to understand how social support is associated with the relationships between mastery and distress in both White American and African American older adults. Thus, an exploratory mediation model was assessed, in order to better understand the associations among mastery, perceived support, and psychological distress, especially in light of the marked differences in social support between racial groups and the indices of social support that are found in the literature. An attempt to better understand the relationships among these variables in later life in general and differentially by racial group in particular is of primary importance, especially for those older adults who are visibly experiencing psychological distress. However, the sample employed in this study consists of reasonably high-functioning older adults with no apparent cognitive impairment, serious mental illness, or physical disability. While generally high functioning within their age groups, these people have encountered recent stressors, including losses and other stresses that have apparently resulted in the need for outpatient services. This group, albeit relatively unique among the elderly, warrant study concerning factors related to psychological distress.

Method Participants Participants in this study were 311 community-dwelling, English-speaking older adults residing in the greater New York metropolitan area. None was a current psychotherapy patient but all had been referred for outpatient psychotherapy, by professionals such as physicians, social workers, or members of the clergy on the basis of observed psychological distress. All were considered by referring professionals to be reasonably healthy—emotionally, cognitively, and physically. None had required frequent or intense psychological services in the

6

The International Journal of Aging and Human Development 0(0)

past; the current distress seemed to be due to distress occasioned by recent losses and stresses of the kind that are not uncommon among older adults (e.g., death of a spouse and decrement of financial resources). Of those who were informed about the study, 96% agreed to participate. Of those initially approached (n ¼ 321), 10 did not meet criteria for inclusion. Those excluded from participating in this study had received mental health treatment within the past 5 years, resided in an institution (including a nursing home), or suffered from a neurodegenerative disorder or psychosis. None of the participants had a prior history of cognitive impairment, and none showed any evidence of cognitive impairment during their interview. The data were collected after the referring professional noted distress and made a referral, and before mental health services were obtained, so that psychiatric diagnoses were not available. The ages of the participants ranged from 65 to 94 years (M ¼ 74.1, SD ¼ 6.95). The majority (65.3%) of the sample were female. A special effort had been made to contact equal numbers of senior facilities primarily serving White Americans or African Americans (especially Black churches). A little more than half of the sample were African American (n ¼ 164), while the remaining were of European American descent (n ¼ 146). Because of the heterogeneity of the African American community in the tristate area, the present study did not include older adults who identified as either Caribbean Americans or as recent African immigrants. Additional demographic information can be found in Table 1.

Measures Demographics. Participants provided information concerning their gender, age, ethnic identity, marital status, living situation, level of education, and occupation. Gender and race were coded as dichotomous variables (1 ¼ male, 2 ¼ female; 1 ¼ White American, 2 ¼ African American). Education was measured on a six-level scale that included junior high school, partial high school, high school, partial college, college/university, and graduate training. Marital status encompassed married, divorced, widowed, separated, and never married, and living situation: living alone, living with spouse, or living with someone else. Perceived financial adequacy was measured on a four-level scale that included not at all adequate, not quite enough, just enough, and more than enough. Mastery. General perceptions of mastery were assessed using the Mastery Scale (Pearlin & Schooler, 1978). This seven-item measure assesses the degree to which individuals perceive themselves as able to influence events in their lives, using statements such as “I can do just about anything I really set my mind to.” Respondents were asked to rate their agreement with such statements on a 4point Likert-type scale, with options ranging from strongly disagree to strongly

Morin and Midlarsky

7

Table 1. Descriptive Statistics. Variable

Age Sex Female Male Education Junior high school Partial high school High school Partial college College/university Graduate training Marital status Married Widowed Divorced/separated Never married Living situation Lives alone Lives with spouse Lives with someone else Perceived health Perceived financial adequacy Not at all adequate Not quite enough Just enough More than adequate Perceived social support Mastery Psychological distress Range

M (SD)/%

White Americans

African Americans

45.70% 76.74 (6.9)

54.30% 71.94 (6.2)

65.30% 34.70%

64.8% (94) 35.2% (51)

65.7% (113) 34.3% (59)

17.10% 21.50% 33.80% 13.90% 8.80% 6.00%

11.1% 20.7% 29% 17.9% 13.8% 7.6%

(16) (30) (42) (26) (20) (11)

20.3% 22.1% 37.8% 10.5% 4.7% 4.7%

29.30% 47.60% 17.90% 5.00%

31.7% 53.1% 3.4% 6.9%

(46) (77) (5) (10)

27.3(47) 43% (74) 21.5% (37) 2.9% (5)

63.10% 27.10% 9.80% 3.46 (.81)

66.2% 29.7% 4.1% 3.23

(96) (43) (6) (.86)

60.5% 25% 14.5% 3.64

74.14 (6.9)

9.50% 25.20% 47.60% 17.40% 5.03 (1.19) 2.75 (.57) .67 (.56) 0–2.36

13.1% (19) 18.6% (27) 49 % (71) 18.6% (27) 4.75 (1.42) 2.58 (.57) .86 (.54) .04–2.36

(35) (38) (65) (18) (8) (8)

(104) (43) (25) (.72)

6.4% (11) 30.8% (53) 46.5% (80) 16.3% (28) 5.27(.87) 2.89 (.54) .50 (.52) 0–2.36

agree (where higher scores indicate stronger perceptions of mastery). The Mastery Scale has been found to have satisfactory psychometric properties (Pearlin & Schooler, 1978) and has been utilized with racially diverse samples (e.g., Pearlin et al., 2007). In the current study, the alpha coefficient for this scale was .67 in the White American sample and .71 in the African American sample.

8

The International Journal of Aging and Human Development 0(0)

Perceived social support. Perceived social support was assessed using the satisfaction subscale of the Social Support Questionnaire—Self Report (SSQSR; Sarason, Sarason, Shearin, & Pierce, 1987). The SSQSR is a 6-item index that was derived by means of factor analysis from the 27-item SSQ (Sarason, Levine, Basham, & Sarason, 1983) and includes such items as “whom can you count on to really be dependable when you need help?” The scale measures two dimensions: the individuals who can be counted on for support (for each question a respondent may list up to nine supporters, identifying them as either friends or family members) and the individual’s degree of satisfaction with the support. The satisfaction with their reported support network is measured on a 6-point Likert-type scale, with options ranging from very dissatisfied to very satisfied. The SSQSR has been used widely, with a diverse range of participants (e.g., Anan & Barnett, 1999). Because a disparity has been reported in the literature around support network size between White American and African American older adults, the satisfaction subscale, rather than size of the support network, was used in this study (Ajrouch et al., 2001). In the current investigation, the alpha coefficient for satisfaction with social support was .90 for the White American sample and .86 for the African American sample. Psychological distress. Psychological distress was assessed using the Brief Symptom Inventory (BSI; Derogatis, 1975). The BSI is a 53-item self-report inventory designed to reflect typical symptomology of people suffering from psychiatric problems. Each item is scored on a 5-point Likert-type scale (0–4), ranging from not at all to extremely in response to the experience of such problems as feeling hopeless about the future. The BSI has been subject to extensive reliability and validity analyses in noncognitively disabled adult populations (Derogatis, 1993). For the purposes of this study, the Global Severity Index was used as a general measure of current psychopathology, because it combines data on both the number of symptoms and the intensity of distress into an effective single summary score (with a T score of 63 indicating clinical significance of symptoms). As this sample was not seeking mental health services and was functioning within the community, the intensity and severity of recently experienced symptoms, rather than particular diagnostic criteria, were used to represent psychological distress. Cronbach’s alphas on the GSI were .88 for White Americans and .95 for African Americans in the current sample.

Procedure Project staff recruited participants through social workers associated with senior centers and residential sites, religious clergy, and physicians and mental health professionals who have extensive contact with older adults. All participants had been referred for outpatient mental health services and were notified about the

Morin and Midlarsky

9

study by the staff at the referral site. Those who had agreed to participate were contacted by phone and invited to take part. Study participants were individually interviewed in their homes (with no family members or others present) or in offices designated for study use. All interviewers had either masters or doctoral degrees in psychology, sociology, or social work. Participants were not matched to interviewers by race. Before the interview, written consent was obtained from each participant, at which time they were informed that the study was a survey about their health, social relationships, and current life situation. Participants were informed about the steps taken to ensure confidentiality. After giving consent, participants were administered a structured interview that included the measures described earlier, in addition to other measures such as current physical health, religious beliefs, and attitudes toward mental health help seeking. Questions were read aloud to the participants, who responded verbally, a desired protocol for older adult participants (Kane & Kane, 1981). Interviewers made a written record of each response as the interview was being conducted. The protocol and procedures used in the investigation were approved by the institutional review board of Teachers College, Columbia University.

Data Analysis Means, standard deviations, and correlations between demographic variables, perceived social support, mastery, and psychological distress were generated prior to testing the hypotheses of the study. Comparisons of the two racial groups on key study variables were also conducted. To explore the hypotheses concerning the direct and indirect effects among mastery, perceived social support, and psychological distress, a path analysis based in structural equation modeling was used (Muthen, 1984). As it was of interest to compare the hypothesized models for best fit while using appropriate estimation techniques for the variables of interest, path analyses were modeled using MPlus statistical software. This allowed for the objective assessment of which model better explained the relationships in the data, a level of analysis which would not be possible using multiple regression-based path analyses. The model was first assessed for the whole sample, and fit indices were compared to a subsequent analysis categorizing the sample by race (a two-group solution). To investigate the links among the constructs measured, including a potential mediating model, path analyses were utilized to better account for measurement error and possible structural invariance by group (Cook & Campbell, 1979; Lei & Wu, 2007). However, because of the cross-sectional nature of the data, claims of temporal precedence and directionality cannot be established with certainty, with possibilities of bidirectional or third variable effects impossible to rule out (Gelfand, Mensinger, & Tenhave, 2009).

10

The International Journal of Aging and Human Development 0(0)

Paths employed to tests relationships among variables were first tested for the total sample and then separately for White American and African American older adults. The analyses were of the direct relationship of mastery and perceived social support to psychological distress and the indirect relationship of mastery to psychological distress through perceived social support.

Results Descriptive Statistics Means and standard deviations of all variables, categorized by race, are presented in Table 1. In order to assess differences by racial group, we conducted one-way analyses of variance on the key variables, wherein the independent variable was race (White American and African American), and the dependent variables were the demographic variables and the primary study variables (social support, mastery, and psychological distress). Significant differences in the study groups found with age, F(1, 311) ¼ 42.65, p < .01, education, F(1, 311) ¼ 15.05, p < .01, perceived health, F(1, 311) ¼ 22.55, p < .01, mastery, F(1, 311) ¼ 27.15, p < .01, satisfaction with social support, F(1, 311) ¼ 17.16, p < .01, and psychological distress, F(1, 311) ¼ 40.43, p < .01. White American participants were older and had more years of education and higher levels of distress, whereas African American participants reported higher levels of perceived health, mastery, and social support. Correlation coefficients among all study variables for the two separate racial groups were in expected directions (see Table 2). When preliminary regression analyses were conducted by group, the study variables of interest (perceived social support and mastery) were significantly related to psychological distress over and above the effects of sex, education, and perceived health for both White American and African American participants, with no interactive effects found between sex and race. Sex, education, and health were therefore excluded from further analyses.

Path Model for the Total Sample A structural equation-based path model was employed to assess the fit of the hypothesized model for the total sample (see Figure 1). Because the model was precisely identified (with zero degrees of freedom), the meaningful fit indices consisted of the Akaike Information Criteria (AIC), the Bayesian Information Criteria (BIC), and the sample-size adjusted BIC (ssBIC). In this model, the AIC yielded a value of 1408.07, the BIC a value of 1434.23, and the ssBIC a value of 1412.02. Decomposition effects for this model, as shown in Table 3, yielded significant effects for all predicted relationships. Mastery was inversely related to

Morin and Midlarsky

11

Table 2. Bivariate Pearson Correlation Matrix for All Variables. Variables 1. 2. 3. 4. 5. 6. 7.

Sex Education Perceived health Marital status Living situation Financial adequacy Perceived social support 8. Mastery 9. Psychological distress

1

2

.5 .1 .11 0.16 .01 .09 .40** .05 .1 .1 0 .1 .08 .07

3 .19* .50** .03 .11 .09 .12

4

5

6

7

8

.09 .20** .28** .03 .2 .08 .06 .32** .07 .31** .28** .09 .26** .2 .54** .24 .29** .01 .21 .41** .19* .25** .19* .03 .14 .02 .16* .12 .12 .12 .23**

.24** .34** .13 .17* .50** .17*

.05 .06

.16 .02

9 .04 .23** .44** .04 .06 .09 .09

.14 .43** .37** .46**

Note. Correlations for White Americans are listed below the diagonal and for African Americans are listed above the diagonal. *p < .05. **p < .01.

Figure 1. Path analysis for the total sample.

psychological distress for the total sample ( ¼ .457, p < .001), as was social support ( ¼ .071, p < .01). Additionally, mastery was positively related to satisfaction with social support ( ¼ .430, p < .001). In the model for the total sample, the hypothesis of a mediating effect of perceived social support on the relationship between mastery and psychological distress was supported, as evidenced by a significant indirect effect between mastery and psychological distress ( ¼ .031, p < .05).

12

The International Journal of Aging and Human Development 0(0)

Table 3. Decomposition of Effects From the Path Analysis for Total Sample. Effect

Parameter estimate

On psychological distress Of mastery 0.457** Of perceived social support 0.071** On perceived social support Of mastery 0.430** From mastery to psychological distress Total 0.487** Total indirect 0.031*

Standardized estimate

Two-tailed p value

0.048 0.023

0.473** 0.155**

Social Support, Mastery, and Psychological Distress in Black and White Older Adults.

Social support and mastery can protect against psychological distress in late life, carrying implications for theory and intervention. However, some g...
238KB Sizes 0 Downloads 9 Views