Aging & Mental Health, 2015 Vol. 19, No. 8, 723 730, http://dx.doi.org/10.1080/13607863.2014.962009
Leisure, functional disability and depression among older Chinese living in residential care homes Zheng Ouyanga,b, Alice M. L. Chongb*, Ting Kin Ngb and Susu Liub a
China Research Center on Aging, Beijing, China; bDepartment of Applied Social Sciences, City University of Hong Kong, Hong Kong, China (Received 25 May 2014; accepted 19 August 2014)
Objectives: Previous research has rarely examined the intervening and buffering effects of leisure on the relationship between age-related stress and health among institutionalized elders, especially in the Chinese context. This study thus examines the extent to which participation in leisure activities mediates and moderates the impact of functional disability on depression among older adults living in residential care homes in China. Method: A total of 1429 participants (858 men) aged over 60 living in residential care homes, of which 46.1% experienced depression using a cut-off score 5 on the 15-item Geriatric Depression Scale, were selected from a national survey across China by using the probability proportional to size sampling method. Results: The findings showed that depression was positively predicted by functional disability and negatively predicted by participation in leisure activities. The results of the mediation analysis showed that participation in leisure activities partially mediated the relationship between functional disability and depression. Functional disability predicted depression both directly and indirectly through its negative influence on participation in leisure activities. Participation in leisure activities also significantly buffered the relationship between functional disability and depression such that the impact of functional disability was weaker for those who participated in leisure activities more frequently. Conclusion: These results provide support for the mediating and moderating roles of leisure in the stress health relationship among institutionalized elders. To enhance residents’ psychological health, residential care homes are recommended to organize more leisure activities. Keywords: leisure activities; moderation; mediation
Introduction China has the largest elderly population in the world. By 2009, the population aged 60 and above in China (160.25 million) had exceeded that in Europe (158.50 million), and it is estimated to reach 440 million by 2050, accounting for 31% of the total Chinese population (United Nations Population Division, 2009). The aging trend in China is accompanied by potential challenges such as the dramatically increasing need to develop a long-term care system and the emphasis on promoting physical and mental well-being in later life. This study investigates how depression rates are affected by leisure activities and functional disability in the elderly by conducting a cross-sectional survey of Chinese elders living in residential care homes (RCHs) recruited in a national survey. Depression and its prevalence Depression is a syndrome that encompasses a broad set of symptoms characterized by sadness and low mood with diverse behavioral referents (Segal, Williams, & Teasdale, 2012). Previous research has identified a range of risk factors for depression in older adults, including chronic medical conditions, polypharmacy, sleep disturbance, physical or cognitive functional impairment, bereavement and major
*Corresponding author. Email: [email protected] Ó 2014 Taylor & Francis
losses, social isolation, and substance abuse (Cahoon, 2012; Madhusoodanan, Ibrahim, & Malik, 2010). Untreated depression can lead to serious problems, including alcohol abuse, chronic illnesses, suicide and the increased use of health care services (Reynolds, Alexopoulos, & Katz, 2002). Various chronic medical illnesses that are prevalent in the elderly have been associated with untreated depression, such as cardiovascular disease, diabetes, Parkinson’s disease and chronic pain (Cahoon, 2012; Slavich & Irwin, 2014). One key barrier to the effective treatment of depression is medication non-adherence, which occurs when patents do not follow the recommended medication regimens (Pompili et al., 2013). In depressive disorder, reduced adherence has been shown to be associated with less improvement of depressive symptoms, greater relapse and recurrence of depression, hospitalization, functional decline and increased suicide risk (Pompili et al., 2009, 2013). Moreover, depression affects not only the patient, but also the family and society (Pincus & Pettit, 2001). Depressive symptoms among older adults have been associated with increased caregiver burden among family caregivers (Cahoon, 2012). Depression is one of the most prevalent mental disorders and health threats among older people worldwide (Blazer, 2000). Its prevalence rate is even higher among
724
Z. Ouyang et al.
the hospitalized and institutionalized elderly, ranging from 30% to 40% (Kerber, Dyck, Culp, & Buckwalter, 2005; Teresi, Abrams, Holmes, Ramirez, & Eimicke, 2001). Recent studies have shown that the prevalence rate of depression among Chinese elders varies from 19% to 45% in Mainland China (Guo & Lai, 2011; Li et al., 2012) and from 29% to 42.5% in Hong Kong (Chow et al., 2004; Woo et al., 1994). However, depression has been astonishingly ignored or unrecognized by staff in RCHs. One study found that fewer than half of depressive cases among older nursing home residents in New York were recognized by staff (Teresi et al., 2001). Another US study found that only 55% of older nursing home residents with significant depressive symptoms received antidepressant therapy and only 68% of those receiving antidepressant therapy received adequate dosages (Brown, Lapane, & Luisi, 2002). Because of the deteriorating physical health of institutionalized elders, clinicians may assume that the depression is a consequence of the medical condition, or consider the concomitant illness as a contraindication to antidepressant therapy and withhold antidepressants (Roose & Dalack, 1992). As depression in the elderly is treatable, researchers have called for improving the management of depression in RCHs (Brown et al., 2002). Thus, there is a great need to examine the prevalence, impacts and correlates of depression among the elderly living in RCHs. Functional disability and depression A large body of studies has examined the stressors associated with depression among older adults (e.g., Ishine, Okumiya, & Matsubayashi, 2007; Koenig & George, 1998; Penninx et al., 1998). Functional disability has consistently been identified as a crucial stressor contributing to poor well-being and depression in later life (Dunne, Wrosch, & Miller, 2011; Yang, 2006). Research has found that measures of physical function in older adults are related to feelings of well-being (Spirduso & Cronin, 2001). Failure to carry out activities of daily living (ADL) has been found to trigger depression among elderly people (Mancini & Bonanno, 2006). Elderly people’s inability to take care of their basic needs may elicit incapable and dependent feelings, which in turn lead to distressed and pessimistic emotions (Bozo, Toksabay, & Kurum, 2009). Furthermore, elderly people living in RCHs who experience many functional problems such as walking impairment, bowel incontinence and bladder incontinence are prone to depression. Leisure, functional disability and depression among older adults Leisure is regarded as an important means of coping with stress and maintaining mental health. Iwasaki and Mannell (2000) suggested that leisure activities may promote a positive mood and enable people to take a temporary break from stressful situations and feel refreshed. Previous studies have established that participation in leisure
activities in the elderly is related to less depressive moods and more positive feelings (Hong, Hasche, & Bowland, 2009; Simone & Haas, 2013). For instance, Dupuis and Smale (1995) found that participation in various kinds of leisure activities such as outdoor pursuits, making crafts, visiting friends and relatives, and swimming was associated with better psychological well-being and less depressive symptoms among older adults. It has also been found that participation in leisure activities reduces depressive symptoms among older widowed women (Janke, Nimrod, & Kleiber, 2008). A plausible explanation for these associations is that participation in leisure activities may enhance the physical health and cognitive functioning of the elderly (Simone & Haas, 2013). Despite ample empirical support for the inverse relationship between leisure and depression in later life, the previous studies have mostly focused on the main effect of leisure. The current research draws upon the stress process model (Lazarus & Folkman, 1984; Pearlin, 1999) to investigate the potential mediating and moderating effects of leisure on the association between functional disability and depression among older adults living in Chinese RCHs. The moderating effect of leisure The stress process model provides a useful framework to explain the underlying process between stressors and mental illnesses (Katerndahl & Parchman, 2002). It highlights the roles of personal coping resources (e.g., personality characteristics, coping strategies) and social coping resources (e.g., social support) in dealing with different stressors. (Lazarus & Folkman, 1984; Pearlin, 1999). Psychosocial coping resources may serve as mediating or moderating factors in the association between stressors and health (Ensel & Lin, 1991). Iwasaki and Mannell (2000) suggested that leisure generates two types of stress coping resources that buffer the negative effect of stress on mental health. First, leisure activities can develop friendships, which offer social support. Second, leisure can generate a sense of autonomy, which refers to people’s beliefs that their actions are selfdetermined. These two components constitute a leisure coping beliefs dimension, which represents beliefs that leisure helps cope with stress (Iwasaki & Mannell, 2000). Coleman (1993) found that leisure-generated self-determination disposition buffered the detrimental effects of life stress on health. Moreover, Iso-Ahola and Park’s (1996) study of a sample of Taekwondo practitioners supported the buffering impact of leisure-based social support on the stress illness relationship. The mediating effect of leisure Ensel and Lin (1991) proposed two rival indirect processes by which coping resources may intervene between stressors and health outcomes. First, in a counteractive model, the experience of stressors mobilizes or increases the use of coping resources, which in turn promotes wellbeing. Second, in a deterioration model, the presence of
Aging & Mental Health stressors has a negative indirect effect on well-being by weakening the levels of coping resources. Both the models suggest that coping resources may mediate the relationship between stressors and well-being. However, the counteractive model postulates that stressors have positive impacts on coping resources, while the deterioration model posits that these impacts are negative. Apart from the leisure coping beliefs dimension (leisure-generated friendships and autonomy), Iwasaki and Mannell (2000) also proposed a leisure coping strategies dimension, which refers to actual situation-specific stress coping behaviors through involvements in leisure. Leisure coping strategies represent a process through which a particular stressful situation may elicit the use of a certain type of leisure activity to cope with stress and maintain good health (Iwasaki & Mannell, 2000). This line of thinking is consistent with Ensel and Lin’s (1991) counteractive model of stress coping. On the contrary, Simone and Haas (2013) proposed an alternative mediating process by which age-related frailty may limit the older people’s ability to participate in leisure activities, which in turn results in lower well-being. This idea echoes Ensel and Lin’s (1991) deterioration model of stress coping. However, research on these competing mediation models for leisure has been sparse and inconsistent (Iwasaki, 2003; Qian, Yarnal, & Almeida, 2014). For example, Iwasaki (2003) examined the mediating effect of leisure, but the results did not support either the counteractive or the deterioration models. Research gaps and aims This study aims to address several knowledge gaps in the current literature. Previous research on the moderating and mediating roles of leisure in the stress health relationship has produced inconsistent results (Qian et al., 2014) and needs further investigations. Moreover, most related studies have focused on life stressors involving major life events, school or work problems, and interpersonal difficulties among university students and younger adults (e.g., Coleman, 1993; Iso-Ahola & Park, 1996; Iwasaki, 2003). It is not well understood whether leisure can mediate or moderate the effects of age-related stressors (e.g., functional disability, cognitive impairment) on well-being among older adults. Additionally, research on leisure and depression among institutionalized elders has been scarce, especially in the Chinese context (Wang, Shi, & Gao, 2007). This study thus aims to contribute to the literature by illuminating the moderating and mediating effects of participation in leisure activities on the association between an age-related stressor (namely, functional disability) and depression among Chinese elders living in RCHs. The aims of this study are threefold. The first is to examine the prevalence of depression among institutionalized elders in China. The second is to examine the effects of functional disability and leisure on depression. The third is to examine the mediating and moderating roles of leisure in the relationship between functional disability and depression. Figure 1 presents the conceptual
725 Participation in Leisure Activities
Functional Disability
Depression
Figure 1. The conceptual framework of this study.
framework of this study. Toward these ends, a national survey was carried out in which the first author was involved.
Method Participants The study participants were adults aged over 60 living in RCHs in China, which was obtained through a stratified multistage sampling method conducted by the China Research Center on Aging (CRCA). First, the primary sampling units were all the 31 provinces (excluding Hong Kong and Macau) in China. These provinces were classified into the following four regions according to their geographical location: east (10 provinces/municipalities), center (six provinces), west (12 provinces, municipalities or autonomous regions) and northeast (three provinces). Within each region, the sampled provinces were selected by using the probability proportional to size sampling method, where the measure of size was based on the population aged 60 and over from the fifth National Census (National Bureau of Statistics of China, 2000). Ten sample provinces were selected for this survey, including three provinces each in the eastern region (Jiangsu, Shandong and Guangdong), central region (Anhui, Shanxi and Hubei) and western region (Shaanxi, Sichuan and Yunnan) and one province in the northeastern region (Heilongjiang). Second, within each sampled province, one urban city and one rural county were selected by using a simple random sampling method, such that 10 cities and 10 counties were obtained from each selected province. In the third stage, within each city or county, a sample of 10 RCHs was randomly selected. Then, a list of 200 RCHs was obtained, half from urban areas and half from rural counties. In the fourth stage (individual stage), nine older residents from each urban RCH and eight from each rural RCH were selected through random sampling. Finally, a sample frame of 1700 older residents was obtained. A structured questionnaire was successfully administered to 1636 respondents by trained interviewers from March 2010 to May 2010, with a response rate of 96.24%. The main reasons for non-response were poor health and refusal to participate. The missing data on the demographic variables were handled through listwise deletion. As a result, 207 cases were removed. The missing values for depression, functional disability and participation in leisure activities were imputed by using the expectation-maximization
726
Z. Ouyang et al.
algorithm. Data from 1429 participants were finally used in the subsequent analyses. The adoption of this stratified multistage sampling method was to increase the representativeness of the sample to the population of the older people living in RCHs in China (Babbie, 2013). Data collection The interviewers were social workers and volunteers recruited from the Committee on Aging. They were trained, and a specially designed interview handbook was prepared for them. The interview process was supervised by researchers from the CRCA with extensive fieldwork experience. Prior to participation, all participants were informed about the purpose of the study and they all signed a consent form. Upon the completion of the questionnaire, all participants were thanked and given souvenirs. To ensure accurate responses, the CRCA researchers randomly checked 10% of the questionnaires submitted by each interviewer. Measures
Participation in leisure activities Participation in leisure activities was measured by the following six items developed by the CRCA: making handicrafts, playing games, singing and dancing, attending courses, following rehabilitation therapy and performing mental activities (e.g., chess, jigsaw puzzles). Each item had four options: 1 D the RCH does not organize this activity, 2 D never attend, 3 D participate occasionally and 4 D participate regularly. The total score ranged from 6 to 24. The higher the score, the more leisure activities the individual was involved in. Demographic variables Participants were asked to report their gender, age (in years), education level (illiterate, primary, secondary, college), marital status (never married, married, divorced, widowed) and monthly income (below 82 USD, 82 246 USD, above 246 USD). In the subsequent analyses, age was adopted as a continuous variable, while the other demographic variables were dummy coded. The reference categories for these dummy variables were female, illiterate, never married and below 82 USD.
Depression Depression was measured by using the 15-item short form Geriatric Depression Scale (GDS-15; Sheikh & Yesavage, 1986) translated into Chinese and validated by Mui (1996). Each item was coded as 0 D no and 1 D yes. The total score ranged from 0 to 15, with a higher score indicating a higher level of depression. GDS-15 has been shown to have high sensitivity and specificity rates in the detection of clinical depression (Wancata, Alexandrowicz, Marquart, Weiss, & Friedrich, 2006) and good reliability among the older adults in China (Boey, 2000; Mui, 1996). Following previous research on Chinese older adults (Boey, 2000; Ho, Niti, Kua, & Ng, 2008; Mui, 1996; Nyunt, Fones, Niti, & Ng, 2009), this study adopted a cut-off score 5 indicating mild to severe depression. Functional disability Functional disability was assessed by using the measures of ADL, instrumental activities of daily living (IADL), and mobility developed by the CRCA. The items were similar to those adopted in the previous research (e.g., Katz, 1983). The ADL measure included the following six items: eating, dressing, bathing, using the toilet, moving about indoors and getting into and out of bed. The IADL measure contained seven items, namely cooking, doing the laundry, shopping, making phone calls, sweeping, taking buses and managing money. Mobility was measured by three items, namely carrying objects weighing 10 kg, walking 1.5 2.0 km and climbing stairs. All items were rated on three-point Likert scales (1 D not difficult at all, 2 D a little bit difficult, 3 D unable to do the task). The functional disability score was thus the unweighted sum of these 16 items. The total score ranged from 16 to 48, with a higher score indicating higher functional disability.
Results Sample characteristics Table 1 presents the demographic characteristics of the sample. The sample was predominantly male (58.8%) and ever married (74.0%), with ages ranging from 60 to 103 years (M D 77.83, SD D 8.47). The majority of participants were of low education and income, with 72.0% reporting primary school level or below and 50.0% reporting a monthly income of less than 82 USD. Table 1. Description of sample characteristics (N D 1429). Variable Gender Male Female Age 60 69 70 79 80 M § SD Education Illiterate Primary Secondary College Marital status Never married Ever married Monthly income Below 82 USD 82 246 USD Above 246 USD
N
%
858 571
60.0 40.0
266 535 628 77.63 § 8.47
18.6 37.4 43.9
622 413 338 56
43.5 28.9 23.7 3.9
371 1058
26.0 74.0
744 363 322
52.1 25.4 22.5
Aging & Mental Health
727
Table 2. Descriptive statistics and correlations of the major variables (N D 1429). Variable
Range
M
SD
1
2
3
1. Depression 2. Functional disability 3. Leisure activities
0 15 16 48 6 24
4.57 30.37 15.05
3.11 10.45 4.19
(.766) .319 ¡.347
(.963) ¡.148
(.873)
Note: Values in the diagonal are Cronbach’s as.
p < .001.
Descriptive statistics Table 2 presents the means, standard deviations, reliabilities and correlations of the major variables in this study. All measures demonstrated good internal reliability. By using a cut-off score 5 on the GDS-15 as suggested in the previous research among Chinese older adults (e.g., Boey, 2000; Ho et al., 2008; Mui, 1996; Nyunt et al., 2009), it was found that nearly half of respondents (46.1%) experienced mild to severe depressive symptoms. Higher depression was correlated with higher functional disability (r D .319, p < .001) and lower participation in leisure activities (r D ¡.347, p < .001). Functional disability was negatively correlated with participation in leisure activities (r D ¡.148, p < .001).
Mediating effect of participation in leisure activities We followed Baron and Kenny’s (1986) approach to examine the hypothesized mediating effect of participation in leisure activities on the relationship between functional disability and depression. First, it is required that the independent variable affects the dependent variable. After controlling for the influences of the demographic variables (gender, age, education, marital status, income), functional disability was found to be positively related to depression (b D .353, p < .001). Second, it is required that the independent variable affects the mediator. Functional disability was shown to be negatively related to participation in leisure activities after the effects of demographic variables had been controlled for (b D ¡.194, p < .001). Third, it is required that the mediator affects the dependent variable while controlling for the independent variable. Participation in leisure activities was negatively related to depression when the influences of functional disability and the demographic variables were controlled for (b D ¡.289, p
Leisure, functional disability and depression among older Chinese living in residential care homes Zheng Ouyanga,b, Alice M. L. Chongb*, Ting Kin Ngb and Susu Liub a
China Research Center on Aging, Beijing, China; bDepartment of Applied Social Sciences, City University of Hong Kong, Hong Kong, China (Received 25 May 2014; accepted 19 August 2014)
Objectives: Previous research has rarely examined the intervening and buffering effects of leisure on the relationship between age-related stress and health among institutionalized elders, especially in the Chinese context. This study thus examines the extent to which participation in leisure activities mediates and moderates the impact of functional disability on depression among older adults living in residential care homes in China. Method: A total of 1429 participants (858 men) aged over 60 living in residential care homes, of which 46.1% experienced depression using a cut-off score 5 on the 15-item Geriatric Depression Scale, were selected from a national survey across China by using the probability proportional to size sampling method. Results: The findings showed that depression was positively predicted by functional disability and negatively predicted by participation in leisure activities. The results of the mediation analysis showed that participation in leisure activities partially mediated the relationship between functional disability and depression. Functional disability predicted depression both directly and indirectly through its negative influence on participation in leisure activities. Participation in leisure activities also significantly buffered the relationship between functional disability and depression such that the impact of functional disability was weaker for those who participated in leisure activities more frequently. Conclusion: These results provide support for the mediating and moderating roles of leisure in the stress health relationship among institutionalized elders. To enhance residents’ psychological health, residential care homes are recommended to organize more leisure activities. Keywords: leisure activities; moderation; mediation
Introduction China has the largest elderly population in the world. By 2009, the population aged 60 and above in China (160.25 million) had exceeded that in Europe (158.50 million), and it is estimated to reach 440 million by 2050, accounting for 31% of the total Chinese population (United Nations Population Division, 2009). The aging trend in China is accompanied by potential challenges such as the dramatically increasing need to develop a long-term care system and the emphasis on promoting physical and mental well-being in later life. This study investigates how depression rates are affected by leisure activities and functional disability in the elderly by conducting a cross-sectional survey of Chinese elders living in residential care homes (RCHs) recruited in a national survey. Depression and its prevalence Depression is a syndrome that encompasses a broad set of symptoms characterized by sadness and low mood with diverse behavioral referents (Segal, Williams, & Teasdale, 2012). Previous research has identified a range of risk factors for depression in older adults, including chronic medical conditions, polypharmacy, sleep disturbance, physical or cognitive functional impairment, bereavement and major
*Corresponding author. Email: [email protected] Ó 2014 Taylor & Francis
losses, social isolation, and substance abuse (Cahoon, 2012; Madhusoodanan, Ibrahim, & Malik, 2010). Untreated depression can lead to serious problems, including alcohol abuse, chronic illnesses, suicide and the increased use of health care services (Reynolds, Alexopoulos, & Katz, 2002). Various chronic medical illnesses that are prevalent in the elderly have been associated with untreated depression, such as cardiovascular disease, diabetes, Parkinson’s disease and chronic pain (Cahoon, 2012; Slavich & Irwin, 2014). One key barrier to the effective treatment of depression is medication non-adherence, which occurs when patents do not follow the recommended medication regimens (Pompili et al., 2013). In depressive disorder, reduced adherence has been shown to be associated with less improvement of depressive symptoms, greater relapse and recurrence of depression, hospitalization, functional decline and increased suicide risk (Pompili et al., 2009, 2013). Moreover, depression affects not only the patient, but also the family and society (Pincus & Pettit, 2001). Depressive symptoms among older adults have been associated with increased caregiver burden among family caregivers (Cahoon, 2012). Depression is one of the most prevalent mental disorders and health threats among older people worldwide (Blazer, 2000). Its prevalence rate is even higher among
724
Z. Ouyang et al.
the hospitalized and institutionalized elderly, ranging from 30% to 40% (Kerber, Dyck, Culp, & Buckwalter, 2005; Teresi, Abrams, Holmes, Ramirez, & Eimicke, 2001). Recent studies have shown that the prevalence rate of depression among Chinese elders varies from 19% to 45% in Mainland China (Guo & Lai, 2011; Li et al., 2012) and from 29% to 42.5% in Hong Kong (Chow et al., 2004; Woo et al., 1994). However, depression has been astonishingly ignored or unrecognized by staff in RCHs. One study found that fewer than half of depressive cases among older nursing home residents in New York were recognized by staff (Teresi et al., 2001). Another US study found that only 55% of older nursing home residents with significant depressive symptoms received antidepressant therapy and only 68% of those receiving antidepressant therapy received adequate dosages (Brown, Lapane, & Luisi, 2002). Because of the deteriorating physical health of institutionalized elders, clinicians may assume that the depression is a consequence of the medical condition, or consider the concomitant illness as a contraindication to antidepressant therapy and withhold antidepressants (Roose & Dalack, 1992). As depression in the elderly is treatable, researchers have called for improving the management of depression in RCHs (Brown et al., 2002). Thus, there is a great need to examine the prevalence, impacts and correlates of depression among the elderly living in RCHs. Functional disability and depression A large body of studies has examined the stressors associated with depression among older adults (e.g., Ishine, Okumiya, & Matsubayashi, 2007; Koenig & George, 1998; Penninx et al., 1998). Functional disability has consistently been identified as a crucial stressor contributing to poor well-being and depression in later life (Dunne, Wrosch, & Miller, 2011; Yang, 2006). Research has found that measures of physical function in older adults are related to feelings of well-being (Spirduso & Cronin, 2001). Failure to carry out activities of daily living (ADL) has been found to trigger depression among elderly people (Mancini & Bonanno, 2006). Elderly people’s inability to take care of their basic needs may elicit incapable and dependent feelings, which in turn lead to distressed and pessimistic emotions (Bozo, Toksabay, & Kurum, 2009). Furthermore, elderly people living in RCHs who experience many functional problems such as walking impairment, bowel incontinence and bladder incontinence are prone to depression. Leisure, functional disability and depression among older adults Leisure is regarded as an important means of coping with stress and maintaining mental health. Iwasaki and Mannell (2000) suggested that leisure activities may promote a positive mood and enable people to take a temporary break from stressful situations and feel refreshed. Previous studies have established that participation in leisure
activities in the elderly is related to less depressive moods and more positive feelings (Hong, Hasche, & Bowland, 2009; Simone & Haas, 2013). For instance, Dupuis and Smale (1995) found that participation in various kinds of leisure activities such as outdoor pursuits, making crafts, visiting friends and relatives, and swimming was associated with better psychological well-being and less depressive symptoms among older adults. It has also been found that participation in leisure activities reduces depressive symptoms among older widowed women (Janke, Nimrod, & Kleiber, 2008). A plausible explanation for these associations is that participation in leisure activities may enhance the physical health and cognitive functioning of the elderly (Simone & Haas, 2013). Despite ample empirical support for the inverse relationship between leisure and depression in later life, the previous studies have mostly focused on the main effect of leisure. The current research draws upon the stress process model (Lazarus & Folkman, 1984; Pearlin, 1999) to investigate the potential mediating and moderating effects of leisure on the association between functional disability and depression among older adults living in Chinese RCHs. The moderating effect of leisure The stress process model provides a useful framework to explain the underlying process between stressors and mental illnesses (Katerndahl & Parchman, 2002). It highlights the roles of personal coping resources (e.g., personality characteristics, coping strategies) and social coping resources (e.g., social support) in dealing with different stressors. (Lazarus & Folkman, 1984; Pearlin, 1999). Psychosocial coping resources may serve as mediating or moderating factors in the association between stressors and health (Ensel & Lin, 1991). Iwasaki and Mannell (2000) suggested that leisure generates two types of stress coping resources that buffer the negative effect of stress on mental health. First, leisure activities can develop friendships, which offer social support. Second, leisure can generate a sense of autonomy, which refers to people’s beliefs that their actions are selfdetermined. These two components constitute a leisure coping beliefs dimension, which represents beliefs that leisure helps cope with stress (Iwasaki & Mannell, 2000). Coleman (1993) found that leisure-generated self-determination disposition buffered the detrimental effects of life stress on health. Moreover, Iso-Ahola and Park’s (1996) study of a sample of Taekwondo practitioners supported the buffering impact of leisure-based social support on the stress illness relationship. The mediating effect of leisure Ensel and Lin (1991) proposed two rival indirect processes by which coping resources may intervene between stressors and health outcomes. First, in a counteractive model, the experience of stressors mobilizes or increases the use of coping resources, which in turn promotes wellbeing. Second, in a deterioration model, the presence of
Aging & Mental Health stressors has a negative indirect effect on well-being by weakening the levels of coping resources. Both the models suggest that coping resources may mediate the relationship between stressors and well-being. However, the counteractive model postulates that stressors have positive impacts on coping resources, while the deterioration model posits that these impacts are negative. Apart from the leisure coping beliefs dimension (leisure-generated friendships and autonomy), Iwasaki and Mannell (2000) also proposed a leisure coping strategies dimension, which refers to actual situation-specific stress coping behaviors through involvements in leisure. Leisure coping strategies represent a process through which a particular stressful situation may elicit the use of a certain type of leisure activity to cope with stress and maintain good health (Iwasaki & Mannell, 2000). This line of thinking is consistent with Ensel and Lin’s (1991) counteractive model of stress coping. On the contrary, Simone and Haas (2013) proposed an alternative mediating process by which age-related frailty may limit the older people’s ability to participate in leisure activities, which in turn results in lower well-being. This idea echoes Ensel and Lin’s (1991) deterioration model of stress coping. However, research on these competing mediation models for leisure has been sparse and inconsistent (Iwasaki, 2003; Qian, Yarnal, & Almeida, 2014). For example, Iwasaki (2003) examined the mediating effect of leisure, but the results did not support either the counteractive or the deterioration models. Research gaps and aims This study aims to address several knowledge gaps in the current literature. Previous research on the moderating and mediating roles of leisure in the stress health relationship has produced inconsistent results (Qian et al., 2014) and needs further investigations. Moreover, most related studies have focused on life stressors involving major life events, school or work problems, and interpersonal difficulties among university students and younger adults (e.g., Coleman, 1993; Iso-Ahola & Park, 1996; Iwasaki, 2003). It is not well understood whether leisure can mediate or moderate the effects of age-related stressors (e.g., functional disability, cognitive impairment) on well-being among older adults. Additionally, research on leisure and depression among institutionalized elders has been scarce, especially in the Chinese context (Wang, Shi, & Gao, 2007). This study thus aims to contribute to the literature by illuminating the moderating and mediating effects of participation in leisure activities on the association between an age-related stressor (namely, functional disability) and depression among Chinese elders living in RCHs. The aims of this study are threefold. The first is to examine the prevalence of depression among institutionalized elders in China. The second is to examine the effects of functional disability and leisure on depression. The third is to examine the mediating and moderating roles of leisure in the relationship between functional disability and depression. Figure 1 presents the conceptual
725 Participation in Leisure Activities
Functional Disability
Depression
Figure 1. The conceptual framework of this study.
framework of this study. Toward these ends, a national survey was carried out in which the first author was involved.
Method Participants The study participants were adults aged over 60 living in RCHs in China, which was obtained through a stratified multistage sampling method conducted by the China Research Center on Aging (CRCA). First, the primary sampling units were all the 31 provinces (excluding Hong Kong and Macau) in China. These provinces were classified into the following four regions according to their geographical location: east (10 provinces/municipalities), center (six provinces), west (12 provinces, municipalities or autonomous regions) and northeast (three provinces). Within each region, the sampled provinces were selected by using the probability proportional to size sampling method, where the measure of size was based on the population aged 60 and over from the fifth National Census (National Bureau of Statistics of China, 2000). Ten sample provinces were selected for this survey, including three provinces each in the eastern region (Jiangsu, Shandong and Guangdong), central region (Anhui, Shanxi and Hubei) and western region (Shaanxi, Sichuan and Yunnan) and one province in the northeastern region (Heilongjiang). Second, within each sampled province, one urban city and one rural county were selected by using a simple random sampling method, such that 10 cities and 10 counties were obtained from each selected province. In the third stage, within each city or county, a sample of 10 RCHs was randomly selected. Then, a list of 200 RCHs was obtained, half from urban areas and half from rural counties. In the fourth stage (individual stage), nine older residents from each urban RCH and eight from each rural RCH were selected through random sampling. Finally, a sample frame of 1700 older residents was obtained. A structured questionnaire was successfully administered to 1636 respondents by trained interviewers from March 2010 to May 2010, with a response rate of 96.24%. The main reasons for non-response were poor health and refusal to participate. The missing data on the demographic variables were handled through listwise deletion. As a result, 207 cases were removed. The missing values for depression, functional disability and participation in leisure activities were imputed by using the expectation-maximization
726
Z. Ouyang et al.
algorithm. Data from 1429 participants were finally used in the subsequent analyses. The adoption of this stratified multistage sampling method was to increase the representativeness of the sample to the population of the older people living in RCHs in China (Babbie, 2013). Data collection The interviewers were social workers and volunteers recruited from the Committee on Aging. They were trained, and a specially designed interview handbook was prepared for them. The interview process was supervised by researchers from the CRCA with extensive fieldwork experience. Prior to participation, all participants were informed about the purpose of the study and they all signed a consent form. Upon the completion of the questionnaire, all participants were thanked and given souvenirs. To ensure accurate responses, the CRCA researchers randomly checked 10% of the questionnaires submitted by each interviewer. Measures
Participation in leisure activities Participation in leisure activities was measured by the following six items developed by the CRCA: making handicrafts, playing games, singing and dancing, attending courses, following rehabilitation therapy and performing mental activities (e.g., chess, jigsaw puzzles). Each item had four options: 1 D the RCH does not organize this activity, 2 D never attend, 3 D participate occasionally and 4 D participate regularly. The total score ranged from 6 to 24. The higher the score, the more leisure activities the individual was involved in. Demographic variables Participants were asked to report their gender, age (in years), education level (illiterate, primary, secondary, college), marital status (never married, married, divorced, widowed) and monthly income (below 82 USD, 82 246 USD, above 246 USD). In the subsequent analyses, age was adopted as a continuous variable, while the other demographic variables were dummy coded. The reference categories for these dummy variables were female, illiterate, never married and below 82 USD.
Depression Depression was measured by using the 15-item short form Geriatric Depression Scale (GDS-15; Sheikh & Yesavage, 1986) translated into Chinese and validated by Mui (1996). Each item was coded as 0 D no and 1 D yes. The total score ranged from 0 to 15, with a higher score indicating a higher level of depression. GDS-15 has been shown to have high sensitivity and specificity rates in the detection of clinical depression (Wancata, Alexandrowicz, Marquart, Weiss, & Friedrich, 2006) and good reliability among the older adults in China (Boey, 2000; Mui, 1996). Following previous research on Chinese older adults (Boey, 2000; Ho, Niti, Kua, & Ng, 2008; Mui, 1996; Nyunt, Fones, Niti, & Ng, 2009), this study adopted a cut-off score 5 indicating mild to severe depression. Functional disability Functional disability was assessed by using the measures of ADL, instrumental activities of daily living (IADL), and mobility developed by the CRCA. The items were similar to those adopted in the previous research (e.g., Katz, 1983). The ADL measure included the following six items: eating, dressing, bathing, using the toilet, moving about indoors and getting into and out of bed. The IADL measure contained seven items, namely cooking, doing the laundry, shopping, making phone calls, sweeping, taking buses and managing money. Mobility was measured by three items, namely carrying objects weighing 10 kg, walking 1.5 2.0 km and climbing stairs. All items were rated on three-point Likert scales (1 D not difficult at all, 2 D a little bit difficult, 3 D unable to do the task). The functional disability score was thus the unweighted sum of these 16 items. The total score ranged from 16 to 48, with a higher score indicating higher functional disability.
Results Sample characteristics Table 1 presents the demographic characteristics of the sample. The sample was predominantly male (58.8%) and ever married (74.0%), with ages ranging from 60 to 103 years (M D 77.83, SD D 8.47). The majority of participants were of low education and income, with 72.0% reporting primary school level or below and 50.0% reporting a monthly income of less than 82 USD. Table 1. Description of sample characteristics (N D 1429). Variable Gender Male Female Age 60 69 70 79 80 M § SD Education Illiterate Primary Secondary College Marital status Never married Ever married Monthly income Below 82 USD 82 246 USD Above 246 USD
N
%
858 571
60.0 40.0
266 535 628 77.63 § 8.47
18.6 37.4 43.9
622 413 338 56
43.5 28.9 23.7 3.9
371 1058
26.0 74.0
744 363 322
52.1 25.4 22.5
Aging & Mental Health
727
Table 2. Descriptive statistics and correlations of the major variables (N D 1429). Variable
Range
M
SD
1
2
3
1. Depression 2. Functional disability 3. Leisure activities
0 15 16 48 6 24
4.57 30.37 15.05
3.11 10.45 4.19
(.766) .319 ¡.347
(.963) ¡.148
(.873)
Note: Values in the diagonal are Cronbach’s as.
p < .001.
Descriptive statistics Table 2 presents the means, standard deviations, reliabilities and correlations of the major variables in this study. All measures demonstrated good internal reliability. By using a cut-off score 5 on the GDS-15 as suggested in the previous research among Chinese older adults (e.g., Boey, 2000; Ho et al., 2008; Mui, 1996; Nyunt et al., 2009), it was found that nearly half of respondents (46.1%) experienced mild to severe depressive symptoms. Higher depression was correlated with higher functional disability (r D .319, p < .001) and lower participation in leisure activities (r D ¡.347, p < .001). Functional disability was negatively correlated with participation in leisure activities (r D ¡.148, p < .001).
Mediating effect of participation in leisure activities We followed Baron and Kenny’s (1986) approach to examine the hypothesized mediating effect of participation in leisure activities on the relationship between functional disability and depression. First, it is required that the independent variable affects the dependent variable. After controlling for the influences of the demographic variables (gender, age, education, marital status, income), functional disability was found to be positively related to depression (b D .353, p < .001). Second, it is required that the independent variable affects the mediator. Functional disability was shown to be negatively related to participation in leisure activities after the effects of demographic variables had been controlled for (b D ¡.194, p < .001). Third, it is required that the mediator affects the dependent variable while controlling for the independent variable. Participation in leisure activities was negatively related to depression when the influences of functional disability and the demographic variables were controlled for (b D ¡.289, p
