Caregiver depression predicts early discontinuation of care for disabled elderly at home
Correspondence address: Yumiko Arai Research Unit for Nursing Caring Sciences and Psychology, National Institute for Longevity Sciences (NILS), 36-3 Gengo, Morioka-cho, Obu-shi, Aichi 474-8522, Japan. Email: firstname.lastname@example.org
Abstract This longitudinal study investigates the caregiving experiences among Japanese caregivers who provided informal care at home for disabled elderly between 1998 and 1999. Forty-seven caregivers of the impaired elderly continued caregiving at home in Matsuyama Town, a rural area of northern Japan, while 18 caregivers discontinued it. The mean score of the Center for Epidemiologic Studies Depression Scale in 1998 among those who gave up caregiving was significantly higher than that of those who continued caregiving, indicating that depression predicts early discontinuation of care in the home. This is one of the few studies in Japan to suggest that initial caregiver depression is a factor in the decision to terminate care for the disabled elderly at home.
This longitudinal study investigates factors associated with Japanese caregivers' decision to discontinue care for disabled elderly at home after more than 1 year of having provided care at home. There has been an abundance of studies in the USA in relation to factors associated with caregivers' decision to stop providing care for the demented elderly (e.g. by institutionalizing them at long-term care (LTC) facilities). These studies have suggested that caregivers' burden and behavioral disturbances are two of the most important factors predicting this discontinuation in the USA.1–4
The scarcity of studies in Japan led us to conduct a case control study to investigate differences between those who began care at home but rather soon institutionalize the elderly in their care and those who continued caregiving at home. It was found that caregivers' daughters-in-law and moderately limited activities of daily living are the two factors related to caregivers' decision to institutionalize the elderly in a semisuburban area in southern Japan.5 However, that study was retrospective; thus, we were unable to conclude that the above two factors were causes or effects.
Therefore, the present prospective, longitudinal study was conducted in order to investigate predictors for caregivers terminating caregiving to the disabled elderly.
This longitudinal study commenced in June 1998 in Matsuyama Town (population 7126), located in Miyagi Prefecture, northern Japan. The baseline data (T1) were collected by questionnaire June and August of 1998 from 70 ‘certifiably registered disabled elderly’ who lived at home with his/her caregivers. Details of the baseline study, which mainly investigated the factors which deterred caregivers from service utilization, have been reported elsewhere.6 Between June 1999 and August 1999 (T2), these subjects were located where they were being cared for (e.g. at home, in an institution), excluding the deceased.
Among the 70 elderly followed, 12 were institutionalized, two were deceased, three refused to answer, and six of the caregivers in the baseline study eventually gave up caregiving, asking someone else to look after the elderly, leaving 47 eligible subjects. Thus, the number of caregivers who continued caregiving more than 1 year was 47, while 18 caregivers gave up caregiving, including those who institutionalized those in care (n = 12) or asked someone else to look after them (n = 6).
The caregivers of the elderly (T1) were asked to complete the following questionnaires in relation to their burden and caregiving situation: (i) the Japanese version of the CES-D (Center for Epidemiologic Studies Depression Scale);7,8 (ii) the Japanese version of the Zarit Burden Interview (ZBI);9,10 and (iii) questions regarding the demographic variables of the caregivers and patients.
The CES-D is a widely used 20-item self-report scale to identify individuals at risk for depression with the total score ranging from 0 to 60. Individuals scoring 16 or more are generally considered to be at risk for clinical depression. The reliability and the validity of the Japanese version of the CES-D were confirmed.7,8
Cognitive impairment of the elderly was rated by the revised Hasegawa Dementia Rating scale (HDS-R), which is equivalent to the Mini-Mental State Examination and has been widely used in Japan. Hasegawa Dementia Rating scale is a Japanese screening test for dementia that measures overall cognitive function, including verbal orientation and memory, with scores ranging from 0 to 30.11 The diagnosis of the dementia was made by the principal investigator (YA) employing the DSM-III-R.
Disability of the elderly was assessed by Activities of Daily Living (ADL) of the patients (Barthel Index),12 and their behavioral disturbances if any.
χ2 tests and t-tests were conducted to assess differences between those who continued caregiving at home (n = 47) and those who discontinued it (n = 18). The Statistical Package for Social Science (version 9.01; SPSS, Chicago, IL, USA) was used for the above statistical analyses.13
As shown in Table 1, a t-test was conducted to compare the following groups regarding variables concerned: those who continued caregiving at home (n = 47) and those who discontinued it (n = 18). As a result, the mean score of the CES-D of the former group (n = 47) was found to be 11.2 (SD 9.7), which was significantly lower than that of the latter group (n = 18) (i.e. 17.8 (14.0), t = 2.16, P = 0.04). In addition, the mean score of the ZBI in the former group (n = 47) was 34.0 (SD 17.2), and tended to be lower than that of the latter group (n = 18) (i.e. 43.9 (22.3), t = 1.90, P = 0.06).
Table 1. Comparison of those continuing (n = 47) and discontinuing (n = 18) caregiving (t-tests)
|Age||80.0 (7.9)||80.3 (8.8)||0.16||0.88|
|Barthel Index||10.5 (6.5)||8.9 (6.1)||− 0.86||0.39|
|HDS-R||14.5 (9.0)||14.1 (8.5)||− 0.16||0.87|
|Age||58.5 (13.0)||58.3 (15.2)||− 0.06||0.96|
|Hours of Caregiving||6.3 (5.1)||7.4 (5.4)||0.71||0.48|
|ZBI score||34.0 (17.2)||43.9 (22.3)||1.90||0.06|
|CES-D||11.2 (9.7)||17.8 (14.0)||2.16||0.04|
Table 2 gives the results of a comparison between those who continued caregiving and those who discontinued it regarding other variables concerned using χ2 tests (or Fisher's exact tests where appropriate). First, the proportion of the caregivers who were depressed (i.e. those scoring higher than 16), at baseline study was significantly lower among those who continued caregiving than in those who discontinued it (P = 0.02). Second, the proportion of caregivers who received assistance from his/her immediate family or others tended to be higher among those who continued caregiving than among those who discontinued caregiving, although the difference was not statistically significant.
Table 2. Comparison of those continuing (n = 47) and discontinuing (n = 18) caregiving (χ2-tests)
|Behavioral disturbances (+)||10||5||0.31||0.58|
|Use of formal services (+)||25||10||0.03||0.86|
|Help from others (+)||31||8||2.51||0.11|
|Spouse caregiver (+)||7||16||1.34||0.72|
|Daughter-in-law caregiver (+)||16||7||1.34||0.72|
|Child caregiver (+)||7||4|| ||0.48*|
None of the following variables significantly differed between the two groups: elderly person's age; the score of Barthel Index; the proportion of the elderly with behavioral disturbances; caregiver's age; the proportion of caregivers who used formal services; the proportion of the caregiver children; the proportion of caregivers' spouses; and the proportion of daughter-in-law caregivers.
The present study has shown that the mean score of the CES-D among those who gave up caregiving (n = 18) was higher than in those who continued caregiving (n = 47), and that the proportion of caregivers who were depressed at baseline study was significantly higher among those who gave up caregiving than in those who continued it. In addition, although the difference was not significant, the mean score of the ZBI among those who gave up caregiving tended to be higher than in those who continued caregiving. These results indicated that those who gave up caregiving after more than a year were likely to be depressed at the very outset. This implies that those who gave up caregiving, either by asking someone else to look after the elderly or by institutionalizing them, were highly distressed. This is consistent with previous studies in the USA.1–4
The proportion of caregivers who received assistance from immediate family and/or others tended to be higher among those who continued caregiving. This is also consistent with previous findings. Schulz et al., for example, showed in their 1993 study that those able to continue caregiving had greater social support.14
In the present study, the behavioral disturbance of the person being cared for was not a significant factor. A possible explanation may be that the study population of the previous studies in the USA was mainly demented elderly, while our study population consisted of disabled elderly including those with both physical and mental disabilities. Thus, behavioral disturbance was not as manifest as in previous studies, which resulted in the above observation. Neither limited ADL nor the fact that a caregiver was a daughter-in-law, albeit significant in our case-control study, proved to be a significant factor in the present study, probably due to differences in the study design as well as the study population. However, the design of our study is prospective, which is more valid when a cause–effect is investigated.
The present study demonstrated that caregivers' depression is a factor predicting the eventual discontinuation of care for the disabled elderly at home by the caregiver. The importance of caregivers' depression demonstrated in the present study has significant implications for the targeting of services to the disabled elderly and their caregivers. In the new LTC scheme recently implemented in Japan in April, 2000, the provision of services is merely based on the functional levels of the disabled elderly, and caregivers' burden is not taken into account.15,16 Although the coordinators of social services (care-managers) are called upon to monitor the well-being of the caregivers, this monitoring has not been done using valid assessments. Such a lack of valid assessments can result in inappropriate service provision. Thus, a vigorous monitoring system which enables care-managers to assess caregivers' burden together with an insurance scheme which takes this burden into account is necessary if the premature institutionalization of the disabled elderly is to be avoided. Closer coordination between care-managers and psychiatrists is also needed.
The authors would like to thank the participants in the present study, which was in part supported by a grant for Comprehensive Research on Ageing and Health (No. H12-036) provided by the Ministry of Health and Welfare, Japan. The authors are grateful to Ms Yoko Mizuno for her editorial assistance.